diff options
| author | Andrew xia <junluan.xia@intel.com> | 2013-07-31 19:36:36 +0800 |
|---|---|---|
| committer | Andrew xia <junluan.xia@intel.com> | 2013-07-31 19:36:36 +0800 |
| commit | 5670c96f29f9f5ea929f802b0f4e6460cd2c54af (patch) | |
| tree | 27d75a3498d4bd5047f51b6ce5f42beac43cd0d8 /core | |
| parent | 5406013997d6b9e9e3f43c09cd3f53ec7b815efb (diff) | |
| parent | e87de037d608231610352512b8f66d7232398374 (diff) | |
| download | spark-5670c96f29f9f5ea929f802b0f4e6460cd2c54af.tar.gz spark-5670c96f29f9f5ea929f802b0f4e6460cd2c54af.tar.bz2 spark-5670c96f29f9f5ea929f802b0f4e6460cd2c54af.zip | |
Merge branch 'master' into Pool_UI
Conflicts:
core/src/main/scala/spark/SparkContext.scala
core/src/main/scala/spark/scheduler/DAGScheduler.scala
core/src/main/scala/spark/scheduler/SparkListener.scala
core/src/main/scala/spark/scheduler/cluster/ClusterTaskSetManager.scala
core/src/main/scala/spark/scheduler/cluster/TaskSetManager.scala
core/src/main/scala/spark/scheduler/local/LocalTaskSetManager.scala
core/src/main/scala/spark/ui/jobs/IndexPage.scala
core/src/main/scala/spark/ui/jobs/JobProgressUI.scala
Diffstat (limited to 'core')
322 files changed, 7453 insertions, 963 deletions
diff --git a/core/pom.xml b/core/pom.xml index 5edafb3706..f0c936c86a 100644 --- a/core/pom.xml +++ b/core/pom.xml @@ -1,4 +1,21 @@ <?xml version="1.0" encoding="UTF-8"?> +<!-- + ~ Licensed to the Apache Software Foundation (ASF) under one or more + ~ contributor license agreements. See the NOTICE file distributed with + ~ this work for additional information regarding copyright ownership. + ~ The ASF licenses this file to You under the Apache License, Version 2.0 + ~ (the "License"); you may not use this file except in compliance with + ~ the License. You may obtain a copy of the License at + ~ + ~ http://www.apache.org/licenses/LICENSE-2.0 + ~ + ~ Unless required by applicable law or agreed to in writing, software + ~ distributed under the License is distributed on an "AS IS" BASIS, + ~ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + ~ See the License for the specific language governing permissions and + ~ limitations under the License. + --> + <project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd"> <modelVersion>4.0.0</modelVersion> <parent> @@ -40,8 +57,14 @@ <artifactId>protobuf-java</artifactId> </dependency> <dependency> - <groupId>de.javakaffee</groupId> - <artifactId>kryo-serializers</artifactId> + <groupId>com.twitter</groupId> + <artifactId>chill_2.9.3</artifactId> + <version>0.3.0</version> + </dependency> + <dependency> + <groupId>com.twitter</groupId> + <artifactId>chill-java</artifactId> + <version>0.3.0</version> </dependency> <dependency> <groupId>com.typesafe.akka</groupId> @@ -56,6 +79,14 @@ <artifactId>akka-slf4j</artifactId> </dependency> <dependency> + <groupId>org.scala-lang</groupId> + <artifactId>scalap</artifactId> + </dependency> + <dependency> + <groupId>org.scala-lang</groupId> + <artifactId>scala-library</artifactId> + </dependency> + <dependency> <groupId>net.liftweb</groupId> <artifactId>lift-json_2.9.2</artifactId> </dependency> @@ -83,6 +114,14 @@ <groupId>log4j</groupId> <artifactId>log4j</artifactId> </dependency> + <dependency> + <groupId>com.codahale.metrics</groupId> + <artifactId>metrics-core</artifactId> + </dependency> + <dependency> + <groupId>com.codahale.metrics</groupId> + <artifactId>metrics-jvm</artifactId> + </dependency> <dependency> <groupId>org.apache.derby</groupId> diff --git a/core/src/hadoop1/scala/org/apache/hadoop/mapred/HadoopMapRedUtil.scala b/core/src/hadoop1/scala/org/apache/hadoop/mapred/HadoopMapRedUtil.scala index f286f2cf9c..25386b2796 100644 --- a/core/src/hadoop1/scala/org/apache/hadoop/mapred/HadoopMapRedUtil.scala +++ b/core/src/hadoop1/scala/org/apache/hadoop/mapred/HadoopMapRedUtil.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package org.apache.hadoop.mapred trait HadoopMapRedUtil { diff --git a/core/src/hadoop1/scala/org/apache/hadoop/mapreduce/HadoopMapReduceUtil.scala b/core/src/hadoop1/scala/org/apache/hadoop/mapreduce/HadoopMapReduceUtil.scala index 264d421d14..b1002e0cac 100644 --- a/core/src/hadoop1/scala/org/apache/hadoop/mapreduce/HadoopMapReduceUtil.scala +++ b/core/src/hadoop1/scala/org/apache/hadoop/mapreduce/HadoopMapReduceUtil.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package org.apache.hadoop.mapreduce import org.apache.hadoop.conf.Configuration diff --git a/core/src/hadoop1/scala/spark/deploy/SparkHadoopUtil.scala b/core/src/hadoop1/scala/spark/deploy/SparkHadoopUtil.scala index f1c86de4cc..617954cb98 100644 --- a/core/src/hadoop1/scala/spark/deploy/SparkHadoopUtil.scala +++ b/core/src/hadoop1/scala/spark/deploy/SparkHadoopUtil.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy import org.apache.hadoop.conf.Configuration import org.apache.hadoop.mapred.JobConf @@ -24,4 +41,7 @@ object SparkHadoopUtil { // add any user credentials to the job conf which are necessary for running on a secure Hadoop cluster def addCredentials(conf: JobConf) {} + + def isYarnMode(): Boolean = { false } + } diff --git a/core/src/hadoop2-yarn/scala/org/apache/hadoop/mapred/HadoopMapRedUtil.scala b/core/src/hadoop2-yarn/scala/org/apache/hadoop/mapred/HadoopMapRedUtil.scala index 875c0a220b..0f972b7a0b 100644 --- a/core/src/hadoop2-yarn/scala/org/apache/hadoop/mapred/HadoopMapRedUtil.scala +++ b/core/src/hadoop2-yarn/scala/org/apache/hadoop/mapred/HadoopMapRedUtil.scala @@ -1,4 +1,21 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package org.apache.hadoop.mapred import org.apache.hadoop.mapreduce.TaskType diff --git a/core/src/hadoop2-yarn/scala/org/apache/hadoop/mapreduce/HadoopMapReduceUtil.scala b/core/src/hadoop2-yarn/scala/org/apache/hadoop/mapreduce/HadoopMapReduceUtil.scala index 8bc6fb6dea..1a7cdf4788 100644 --- a/core/src/hadoop2-yarn/scala/org/apache/hadoop/mapreduce/HadoopMapReduceUtil.scala +++ b/core/src/hadoop2-yarn/scala/org/apache/hadoop/mapreduce/HadoopMapReduceUtil.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package org.apache.hadoop.mapreduce import org.apache.hadoop.conf.Configuration diff --git a/core/src/hadoop2-yarn/scala/spark/deploy/SparkHadoopUtil.scala b/core/src/hadoop2-yarn/scala/spark/deploy/SparkHadoopUtil.scala index 301a57fffa..6122fdced0 100644 --- a/core/src/hadoop2-yarn/scala/spark/deploy/SparkHadoopUtil.scala +++ b/core/src/hadoop2-yarn/scala/spark/deploy/SparkHadoopUtil.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy import collection.mutable.HashMap diff --git a/core/src/hadoop2-yarn/scala/spark/deploy/yarn/ApplicationMaster.scala b/core/src/hadoop2-yarn/scala/spark/deploy/yarn/ApplicationMaster.scala index f19648ec68..1b06169739 100644 --- a/core/src/hadoop2-yarn/scala/spark/deploy/yarn/ApplicationMaster.scala +++ b/core/src/hadoop2-yarn/scala/spark/deploy/yarn/ApplicationMaster.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.yarn import java.net.Socket @@ -27,6 +44,7 @@ class ApplicationMaster(args: ApplicationMasterArguments, conf: Configuration) e private val yarnConf: YarnConfiguration = new YarnConfiguration(conf) private var yarnAllocator: YarnAllocationHandler = null + private var isFinished:Boolean = false def run() { @@ -68,10 +86,7 @@ class ApplicationMaster(args: ApplicationMasterArguments, conf: Configuration) e // Wait for the user class to Finish userThread.join() - - // Finish the ApplicationMaster - finishApplicationMaster() - // TODO: Exit based on success/failure + System.exit(0) } @@ -124,17 +139,30 @@ class ApplicationMaster(args: ApplicationMasterArguments, conf: Configuration) e } } } - + private def startUserClass(): Thread = { logInfo("Starting the user JAR in a separate Thread") val mainMethod = Class.forName(args.userClass, false, Thread.currentThread.getContextClassLoader) .getMethod("main", classOf[Array[String]]) val t = new Thread { override def run() { - // Copy - var mainArgs: Array[String] = new Array[String](args.userArgs.size()) - args.userArgs.copyToArray(mainArgs, 0, args.userArgs.size()) - mainMethod.invoke(null, mainArgs) + var successed = false + try { + // Copy + var mainArgs: Array[String] = new Array[String](args.userArgs.size()) + args.userArgs.copyToArray(mainArgs, 0, args.userArgs.size()) + mainMethod.invoke(null, mainArgs) + // some job script has "System.exit(0)" at the end, for example SparkPi, SparkLR + // userThread will stop here unless it has uncaught exception thrown out + // It need shutdown hook to set SUCCEEDED + successed = true + } finally { + if (successed) { + ApplicationMaster.this.finishApplicationMaster(FinalApplicationStatus.SUCCEEDED) + } else { + ApplicationMaster.this.finishApplicationMaster(FinalApplicationStatus.FAILED) + } + } } } t.start() @@ -179,7 +207,7 @@ class ApplicationMaster(args: ApplicationMasterArguments, conf: Configuration) e logInfo("All workers have launched.") // Launch a progress reporter thread, else app will get killed after expiration (def: 10mins) timeout - if (userThread.isAlive){ + if (userThread.isAlive) { // ensure that progress is sent before YarnConfiguration.RM_AM_EXPIRY_INTERVAL_MS elapse. val timeoutInterval = yarnConf.getInt(YarnConfiguration.RM_AM_EXPIRY_INTERVAL_MS, 120000) @@ -197,7 +225,7 @@ class ApplicationMaster(args: ApplicationMasterArguments, conf: Configuration) e val t = new Thread { override def run() { - while (userThread.isAlive){ + while (userThread.isAlive) { val missingWorkerCount = args.numWorkers - yarnAllocator.getNumWorkersRunning if (missingWorkerCount > 0) { logInfo("Allocating " + missingWorkerCount + " containers to make up for (potentially ?) lost containers") @@ -235,14 +263,23 @@ class ApplicationMaster(args: ApplicationMasterArguments, conf: Configuration) e } } */ - - def finishApplicationMaster() { + + def finishApplicationMaster(status: FinalApplicationStatus) { + + synchronized { + if (isFinished) { + return + } + isFinished = true + } + + logInfo("finishApplicationMaster with " + status) val finishReq = Records.newRecord(classOf[FinishApplicationMasterRequest]) .asInstanceOf[FinishApplicationMasterRequest] finishReq.setAppAttemptId(appAttemptId) - // TODO: Check if the application has failed or succeeded - finishReq.setFinishApplicationStatus(FinalApplicationStatus.SUCCEEDED) + finishReq.setFinishApplicationStatus(status) resourceManager.finishApplicationMaster(finishReq) + } } @@ -256,7 +293,7 @@ object ApplicationMaster { private val ALLOCATOR_LOOP_WAIT_COUNT = 30 def incrementAllocatorLoop(by: Int) { val count = yarnAllocatorLoop.getAndAdd(by) - if (count >= ALLOCATOR_LOOP_WAIT_COUNT){ + if (count >= ALLOCATOR_LOOP_WAIT_COUNT) { yarnAllocatorLoop.synchronized { // to wake threads off wait ... yarnAllocatorLoop.notifyAll() @@ -291,14 +328,16 @@ object ApplicationMaster { logInfo("Invoking sc stop from shutdown hook") sc.stop() // best case ... - for (master <- applicationMasters) master.finishApplicationMaster + for (master <- applicationMasters) { + master.finishApplicationMaster(FinalApplicationStatus.SUCCEEDED) + } } } ) } // Wait for initialization to complete and atleast 'some' nodes can get allocated yarnAllocatorLoop.synchronized { - while (yarnAllocatorLoop.get() <= ALLOCATOR_LOOP_WAIT_COUNT){ + while (yarnAllocatorLoop.get() <= ALLOCATOR_LOOP_WAIT_COUNT) { yarnAllocatorLoop.wait(1000L) } } diff --git a/core/src/hadoop2-yarn/scala/spark/deploy/yarn/ApplicationMasterArguments.scala b/core/src/hadoop2-yarn/scala/spark/deploy/yarn/ApplicationMasterArguments.scala index 1b00208511..8de44b1f66 100644 --- a/core/src/hadoop2-yarn/scala/spark/deploy/yarn/ApplicationMasterArguments.scala +++ b/core/src/hadoop2-yarn/scala/spark/deploy/yarn/ApplicationMasterArguments.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.yarn import spark.util.IntParam diff --git a/core/src/hadoop2-yarn/scala/spark/deploy/yarn/Client.scala b/core/src/hadoop2-yarn/scala/spark/deploy/yarn/Client.scala index 514c17f241..8bcbfc2735 100644 --- a/core/src/hadoop2-yarn/scala/spark/deploy/yarn/Client.scala +++ b/core/src/hadoop2-yarn/scala/spark/deploy/yarn/Client.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.yarn import java.net.{InetSocketAddress, URI} diff --git a/core/src/hadoop2-yarn/scala/spark/deploy/yarn/ClientArguments.scala b/core/src/hadoop2-yarn/scala/spark/deploy/yarn/ClientArguments.scala index 07e7edea36..67aff03781 100644 --- a/core/src/hadoop2-yarn/scala/spark/deploy/yarn/ClientArguments.scala +++ b/core/src/hadoop2-yarn/scala/spark/deploy/yarn/ClientArguments.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.yarn import spark.util.MemoryParam diff --git a/core/src/hadoop2-yarn/scala/spark/deploy/yarn/WorkerRunnable.scala b/core/src/hadoop2-yarn/scala/spark/deploy/yarn/WorkerRunnable.scala index cc6f3344a1..f458f2f6a1 100644 --- a/core/src/hadoop2-yarn/scala/spark/deploy/yarn/WorkerRunnable.scala +++ b/core/src/hadoop2-yarn/scala/spark/deploy/yarn/WorkerRunnable.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.yarn import java.net.URI diff --git a/core/src/hadoop2-yarn/scala/spark/deploy/yarn/YarnAllocationHandler.scala b/core/src/hadoop2-yarn/scala/spark/deploy/yarn/YarnAllocationHandler.scala index 61dd72a651..b0af8baf08 100644 --- a/core/src/hadoop2-yarn/scala/spark/deploy/yarn/YarnAllocationHandler.scala +++ b/core/src/hadoop2-yarn/scala/spark/deploy/yarn/YarnAllocationHandler.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.yarn import spark.{Logging, Utils} diff --git a/core/src/hadoop2-yarn/scala/spark/scheduler/cluster/YarnClusterScheduler.scala b/core/src/hadoop2-yarn/scala/spark/scheduler/cluster/YarnClusterScheduler.scala index ed732d36bf..307d96111c 100644 --- a/core/src/hadoop2-yarn/scala/spark/scheduler/cluster/YarnClusterScheduler.scala +++ b/core/src/hadoop2-yarn/scala/spark/scheduler/cluster/YarnClusterScheduler.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler.cluster import spark._ diff --git a/core/src/hadoop2/scala/org/apache/hadoop/mapred/HadoopMapRedUtil.scala b/core/src/hadoop2/scala/org/apache/hadoop/mapred/HadoopMapRedUtil.scala index a0652d7fc7..4b3d84670c 100644 --- a/core/src/hadoop2/scala/org/apache/hadoop/mapred/HadoopMapRedUtil.scala +++ b/core/src/hadoop2/scala/org/apache/hadoop/mapred/HadoopMapRedUtil.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package org.apache.hadoop.mapred trait HadoopMapRedUtil { diff --git a/core/src/hadoop2/scala/org/apache/hadoop/mapreduce/HadoopMapReduceUtil.scala b/core/src/hadoop2/scala/org/apache/hadoop/mapreduce/HadoopMapReduceUtil.scala index 7fdbe322fd..aa3b1ed3a5 100644 --- a/core/src/hadoop2/scala/org/apache/hadoop/mapreduce/HadoopMapReduceUtil.scala +++ b/core/src/hadoop2/scala/org/apache/hadoop/mapreduce/HadoopMapReduceUtil.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package org.apache.hadoop.mapreduce import org.apache.hadoop.conf.Configuration diff --git a/core/src/hadoop2/scala/spark/deploy/SparkHadoopUtil.scala b/core/src/hadoop2/scala/spark/deploy/SparkHadoopUtil.scala index f1c86de4cc..617954cb98 100644 --- a/core/src/hadoop2/scala/spark/deploy/SparkHadoopUtil.scala +++ b/core/src/hadoop2/scala/spark/deploy/SparkHadoopUtil.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy import org.apache.hadoop.conf.Configuration import org.apache.hadoop.mapred.JobConf @@ -24,4 +41,7 @@ object SparkHadoopUtil { // add any user credentials to the job conf which are necessary for running on a secure Hadoop cluster def addCredentials(conf: JobConf) {} + + def isYarnMode(): Boolean = { false } + } diff --git a/core/src/main/java/spark/network/netty/FileClient.java b/core/src/main/java/spark/network/netty/FileClient.java index a4bb4bc701..0625a6d502 100644 --- a/core/src/main/java/spark/network/netty/FileClient.java +++ b/core/src/main/java/spark/network/netty/FileClient.java @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.network.netty; import io.netty.bootstrap.Bootstrap; diff --git a/core/src/main/java/spark/network/netty/FileClientChannelInitializer.java b/core/src/main/java/spark/network/netty/FileClientChannelInitializer.java index af25baf641..05ad4b61d7 100644 --- a/core/src/main/java/spark/network/netty/FileClientChannelInitializer.java +++ b/core/src/main/java/spark/network/netty/FileClientChannelInitializer.java @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.network.netty; import io.netty.buffer.BufType; diff --git a/core/src/main/java/spark/network/netty/FileClientHandler.java b/core/src/main/java/spark/network/netty/FileClientHandler.java index 9fc9449827..e8cd9801f6 100644 --- a/core/src/main/java/spark/network/netty/FileClientHandler.java +++ b/core/src/main/java/spark/network/netty/FileClientHandler.java @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.network.netty; import io.netty.buffer.ByteBuf; diff --git a/core/src/main/java/spark/network/netty/FileServer.java b/core/src/main/java/spark/network/netty/FileServer.java index dd3a557ae5..9f009a61d5 100644 --- a/core/src/main/java/spark/network/netty/FileServer.java +++ b/core/src/main/java/spark/network/netty/FileServer.java @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.network.netty; import java.net.InetSocketAddress; diff --git a/core/src/main/java/spark/network/netty/FileServerChannelInitializer.java b/core/src/main/java/spark/network/netty/FileServerChannelInitializer.java index 8f1f5c65cd..50c57a81a3 100644 --- a/core/src/main/java/spark/network/netty/FileServerChannelInitializer.java +++ b/core/src/main/java/spark/network/netty/FileServerChannelInitializer.java @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.network.netty; import io.netty.channel.ChannelInitializer; diff --git a/core/src/main/java/spark/network/netty/FileServerHandler.java b/core/src/main/java/spark/network/netty/FileServerHandler.java index a78eddb1b5..176ba8da49 100644 --- a/core/src/main/java/spark/network/netty/FileServerHandler.java +++ b/core/src/main/java/spark/network/netty/FileServerHandler.java @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.network.netty; import java.io.File; diff --git a/core/src/main/java/spark/network/netty/PathResolver.java b/core/src/main/java/spark/network/netty/PathResolver.java index 302411672c..f446c55b19 100755 --- a/core/src/main/java/spark/network/netty/PathResolver.java +++ b/core/src/main/java/spark/network/netty/PathResolver.java @@ -1,3 +1,20 @@ +/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
package spark.network.netty;
diff --git a/core/src/main/scala/spark/Accumulators.scala b/core/src/main/scala/spark/Accumulators.scala index 57c6df35be..6ff92ce833 100644 --- a/core/src/main/scala/spark/Accumulators.scala +++ b/core/src/main/scala/spark/Accumulators.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import java.io._ diff --git a/core/src/main/scala/spark/Aggregator.scala b/core/src/main/scala/spark/Aggregator.scala index df8ce9c054..136b4da61e 100644 --- a/core/src/main/scala/spark/Aggregator.scala +++ b/core/src/main/scala/spark/Aggregator.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import java.util.{HashMap => JHashMap} diff --git a/core/src/main/scala/spark/BlockStoreShuffleFetcher.scala b/core/src/main/scala/spark/BlockStoreShuffleFetcher.scala index 3239f4c385..8f6953b1f5 100644 --- a/core/src/main/scala/spark/BlockStoreShuffleFetcher.scala +++ b/core/src/main/scala/spark/BlockStoreShuffleFetcher.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import scala.collection.mutable.ArrayBuffer diff --git a/core/src/main/scala/spark/Cache.scala b/core/src/main/scala/spark/Cache.scala deleted file mode 100644 index 20d677a854..0000000000 --- a/core/src/main/scala/spark/Cache.scala +++ /dev/null @@ -1,63 +0,0 @@ -package spark - -import java.util.concurrent.atomic.AtomicInteger - -private[spark] sealed trait CachePutResponse -private[spark] case class CachePutSuccess(size: Long) extends CachePutResponse -private[spark] case class CachePutFailure() extends CachePutResponse - -/** - * An interface for caches in Spark, to allow for multiple implementations. Caches are used to store - * both partitions of cached RDDs and broadcast variables on Spark executors. Caches are also aware - * of which entries are part of the same dataset (for example, partitions in the same RDD). The key - * for each value in a cache is a (datasetID, partition) pair. - * - * A single Cache instance gets created on each machine and is shared by all caches (i.e. both the - * RDD split cache and the broadcast variable cache), to enable global replacement policies. - * However, because these several independent modules all perform caching, it is important to give - * them separate key namespaces, so that an RDD and a broadcast variable (for example) do not use - * the same key. For this purpose, Cache has the notion of KeySpaces. Each client module must first - * ask for a KeySpace, and then call get() and put() on that space using its own keys. - * - * This abstract class handles the creation of key spaces, so that subclasses need only deal with - * keys that are unique across modules. - */ -private[spark] abstract class Cache { - private val nextKeySpaceId = new AtomicInteger(0) - private def newKeySpaceId() = nextKeySpaceId.getAndIncrement() - - def newKeySpace() = new KeySpace(this, newKeySpaceId()) - - /** - * Get the value for a given (datasetId, partition), or null if it is not - * found. - */ - def get(datasetId: Any, partition: Int): Any - - /** - * Attempt to put a value in the cache; returns CachePutFailure if this was - * not successful (e.g. because the cache replacement policy forbids it), and - * CachePutSuccess if successful. If size estimation is available, the cache - * implementation should set the size field in CachePutSuccess. - */ - def put(datasetId: Any, partition: Int, value: Any): CachePutResponse - - /** - * Report the capacity of the cache partition. By default this just reports - * zero. Specific implementations can choose to provide the capacity number. - */ - def getCapacity: Long = 0L -} - -/** - * A key namespace in a Cache. - */ -private[spark] class KeySpace(cache: Cache, val keySpaceId: Int) { - def get(datasetId: Any, partition: Int): Any = - cache.get((keySpaceId, datasetId), partition) - - def put(datasetId: Any, partition: Int, value: Any): CachePutResponse = - cache.put((keySpaceId, datasetId), partition, value) - - def getCapacity: Long = cache.getCapacity -} diff --git a/core/src/main/scala/spark/CacheManager.scala b/core/src/main/scala/spark/CacheManager.scala index f7a2b7e802..81314805a9 100644 --- a/core/src/main/scala/spark/CacheManager.scala +++ b/core/src/main/scala/spark/CacheManager.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import scala.collection.mutable.{ArrayBuffer, HashSet} diff --git a/core/src/main/scala/spark/ClosureCleaner.scala b/core/src/main/scala/spark/ClosureCleaner.scala index d5e7132ff9..8b39241095 100644 --- a/core/src/main/scala/spark/ClosureCleaner.scala +++ b/core/src/main/scala/spark/ClosureCleaner.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import java.lang.reflect.Field diff --git a/core/src/main/scala/spark/Dependency.scala b/core/src/main/scala/spark/Dependency.scala index 2af44aa383..d17e70a4fa 100644 --- a/core/src/main/scala/spark/Dependency.scala +++ b/core/src/main/scala/spark/Dependency.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark /** diff --git a/core/src/main/scala/spark/DoubleRDDFunctions.scala b/core/src/main/scala/spark/DoubleRDDFunctions.scala index 178d31a73b..104168e61c 100644 --- a/core/src/main/scala/spark/DoubleRDDFunctions.scala +++ b/core/src/main/scala/spark/DoubleRDDFunctions.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import spark.partial.BoundedDouble @@ -37,7 +54,13 @@ class DoubleRDDFunctions(self: RDD[Double]) extends Logging with Serializable { * Compute the sample standard deviation of this RDD's elements (which corrects for bias in * estimating the standard deviation by dividing by N-1 instead of N). */ - def sampleStdev(): Double = stats().stdev + def sampleStdev(): Double = stats().sampleStdev + + /** + * Compute the sample variance of this RDD's elements (which corrects for bias in + * estimating the variance by dividing by N-1 instead of N). + */ + def sampleVariance(): Double = stats().sampleVariance /** (Experimental) Approximate operation to return the mean within a timeout. */ def meanApprox(timeout: Long, confidence: Double = 0.95): PartialResult[BoundedDouble] = { diff --git a/core/src/main/scala/spark/FetchFailedException.scala b/core/src/main/scala/spark/FetchFailedException.scala index 40b0193f19..a2dae6cae9 100644 --- a/core/src/main/scala/spark/FetchFailedException.scala +++ b/core/src/main/scala/spark/FetchFailedException.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import spark.storage.BlockManagerId diff --git a/core/src/main/scala/spark/HadoopWriter.scala b/core/src/main/scala/spark/HadoopWriter.scala index 5e8396edb9..b1fe0075a3 100644 --- a/core/src/main/scala/spark/HadoopWriter.scala +++ b/core/src/main/scala/spark/HadoopWriter.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package org.apache.hadoop.mapred import org.apache.hadoop.fs.FileSystem diff --git a/core/src/main/scala/spark/HttpFileServer.scala b/core/src/main/scala/spark/HttpFileServer.scala index 00901d95e2..a13a7a2859 100644 --- a/core/src/main/scala/spark/HttpFileServer.scala +++ b/core/src/main/scala/spark/HttpFileServer.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import java.io.{File} @@ -42,4 +59,4 @@ private[spark] class HttpFileServer extends Logging { return dir + "/" + file.getName } -}
\ No newline at end of file +} diff --git a/core/src/main/scala/spark/HttpServer.scala b/core/src/main/scala/spark/HttpServer.scala index 4e0507c080..c9dffbc631 100644 --- a/core/src/main/scala/spark/HttpServer.scala +++ b/core/src/main/scala/spark/HttpServer.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import java.io.File diff --git a/core/src/main/scala/spark/JavaSerializer.scala b/core/src/main/scala/spark/JavaSerializer.scala index b04a27d073..04c5f44e6b 100644 --- a/core/src/main/scala/spark/JavaSerializer.scala +++ b/core/src/main/scala/spark/JavaSerializer.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import java.io._ diff --git a/core/src/main/scala/spark/KryoSerializer.scala b/core/src/main/scala/spark/KryoSerializer.scala index d723ab7b1e..eeb2993d8a 100644 --- a/core/src/main/scala/spark/KryoSerializer.scala +++ b/core/src/main/scala/spark/KryoSerializer.scala @@ -1,25 +1,34 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import java.io._ import java.nio.ByteBuffer -import java.nio.channels.Channels - -import scala.collection.immutable -import scala.collection.mutable - -import com.esotericsoftware.kryo._ -import com.esotericsoftware.kryo.{Serializer => KSerializer} +import com.esotericsoftware.kryo.{Kryo, KryoException} import com.esotericsoftware.kryo.io.{Input => KryoInput, Output => KryoOutput} import com.esotericsoftware.kryo.serializers.{JavaSerializer => KryoJavaSerializer} -import de.javakaffee.kryoserializers.KryoReflectionFactorySupport - +import com.twitter.chill.ScalaKryoInstantiator import serializer.{SerializerInstance, DeserializationStream, SerializationStream} import spark.broadcast._ import spark.storage._ private[spark] class KryoSerializationStream(kryo: Kryo, outStream: OutputStream) extends SerializationStream { - val output = new KryoOutput(outStream) def writeObject[T](t: T): SerializationStream = { @@ -33,7 +42,6 @@ class KryoSerializationStream(kryo: Kryo, outStream: OutputStream) extends Seria private[spark] class KryoDeserializationStream(kryo: Kryo, inStream: InputStream) extends DeserializationStream { - val input = new KryoInput(inStream) def readObject[T](): T = { @@ -41,7 +49,7 @@ class KryoDeserializationStream(kryo: Kryo, inStream: InputStream) extends Deser kryo.readClassAndObject(input).asInstanceOf[T] } catch { // DeserializationStream uses the EOF exception to indicate stopping condition. - case e: com.esotericsoftware.kryo.KryoException => throw new java.io.EOFException + case _: KryoException => throw new EOFException } } @@ -52,10 +60,9 @@ class KryoDeserializationStream(kryo: Kryo, inStream: InputStream) extends Deser } private[spark] class KryoSerializerInstance(ks: KryoSerializer) extends SerializerInstance { - - val kryo = ks.kryo.get() - val output = ks.output.get() - val input = ks.input.get() + val kryo = ks.newKryo() + val output = ks.newKryoOutput() + val input = ks.newKryoInput() def serialize[T](t: T): ByteBuffer = { output.clear() @@ -91,130 +98,59 @@ private[spark] class KryoSerializerInstance(ks: KryoSerializer) extends Serializ * serialization. */ trait KryoRegistrator { - def registerClasses(kryo: Kryo): Unit + def registerClasses(kryo: Kryo) } /** * A Spark serializer that uses the [[http://code.google.com/p/kryo/wiki/V1Documentation Kryo 1.x library]]. */ class KryoSerializer extends spark.serializer.Serializer with Logging { + private val bufferSize = System.getProperty("spark.kryoserializer.buffer.mb", "2").toInt * 1024 * 1024 - val bufferSize = System.getProperty("spark.kryoserializer.buffer.mb", "2").toInt * 1024 * 1024 - - val kryo = new ThreadLocal[Kryo] { - override def initialValue = createKryo() - } - - val output = new ThreadLocal[KryoOutput] { - override def initialValue = new KryoOutput(bufferSize) - } + def newKryoOutput() = new KryoOutput(bufferSize) - val input = new ThreadLocal[KryoInput] { - override def initialValue = new KryoInput(bufferSize) - } + def newKryoInput() = new KryoInput(bufferSize) - def createKryo(): Kryo = { - val kryo = new KryoReflectionFactorySupport() + def newKryo(): Kryo = { + val instantiator = new ScalaKryoInstantiator + val kryo = instantiator.newKryo() + val classLoader = Thread.currentThread.getContextClassLoader // Register some commonly used classes val toRegister: Seq[AnyRef] = Seq( - // Arrays - Array(1), Array(1.0), Array(1.0f), Array(1L), Array(""), Array(("", "")), - Array(new java.lang.Object), Array(1.toByte), Array(true), Array('c'), - // Specialized Tuple2s - ("", ""), ("", 1), (1, 1), (1.0, 1.0), (1L, 1L), - (1, 1.0), (1.0, 1), (1L, 1.0), (1.0, 1L), (1, 1L), (1L, 1), - // Scala collections - List(1), mutable.ArrayBuffer(1), - // Options and Either - Some(1), Left(1), Right(1), - // Higher-dimensional tuples - (1, 1, 1), (1, 1, 1, 1), (1, 1, 1, 1, 1), - None, ByteBuffer.allocate(1), StorageLevel.MEMORY_ONLY, PutBlock("1", ByteBuffer.allocate(1), StorageLevel.MEMORY_ONLY), GotBlock("1", ByteBuffer.allocate(1)), GetBlock("1") ) - for (obj <- toRegister) { - kryo.register(obj.getClass) - } + + for (obj <- toRegister) kryo.register(obj.getClass) // Allow sending SerializableWritable kryo.register(classOf[SerializableWritable[_]], new KryoJavaSerializer()) kryo.register(classOf[HttpBroadcast[_]], new KryoJavaSerializer()) - // Register some commonly used Scala singleton objects. Because these - // are singletons, we must return the exact same local object when we - // deserialize rather than returning a clone as FieldSerializer would. - class SingletonSerializer[T](obj: T) extends KSerializer[T] { - override def write(kryo: Kryo, output: KryoOutput, obj: T) {} - override def read(kryo: Kryo, input: KryoInput, cls: java.lang.Class[T]): T = obj - } - kryo.register(None.getClass, new SingletonSerializer[AnyRef](None)) - kryo.register(Nil.getClass, new SingletonSerializer[AnyRef](Nil)) - - // Register maps with a special serializer since they have complex internal structure - class ScalaMapSerializer(buildMap: Array[(Any, Any)] => scala.collection.Map[Any, Any]) - extends KSerializer[Array[(Any, Any)] => scala.collection.Map[Any, Any]] { - - //hack, look at https://groups.google.com/forum/#!msg/kryo-users/Eu5V4bxCfws/k-8UQ22y59AJ - private final val FAKE_REFERENCE = new Object() - override def write( - kryo: Kryo, - output: KryoOutput, - obj: Array[(Any, Any)] => scala.collection.Map[Any, Any]) { - val map = obj.asInstanceOf[scala.collection.Map[Any, Any]] - output.writeInt(map.size) - for ((k, v) <- map) { - kryo.writeClassAndObject(output, k) - kryo.writeClassAndObject(output, v) - } - } - override def read ( - kryo: Kryo, - input: KryoInput, - cls: Class[Array[(Any, Any)] => scala.collection.Map[Any, Any]]) - : Array[(Any, Any)] => scala.collection.Map[Any, Any] = { - kryo.reference(FAKE_REFERENCE) - val size = input.readInt() - val elems = new Array[(Any, Any)](size) - for (i <- 0 until size) { - val k = kryo.readClassAndObject(input) - val v = kryo.readClassAndObject(input) - elems(i)=(k,v) - } - buildMap(elems).asInstanceOf[Array[(Any, Any)] => scala.collection.Map[Any, Any]] + // Allow the user to register their own classes by setting spark.kryo.registrator + try { + Option(System.getProperty("spark.kryo.registrator")).foreach { regCls => + logDebug("Running user registrator: " + regCls) + val reg = Class.forName(regCls, true, classLoader).newInstance().asInstanceOf[KryoRegistrator] + reg.registerClasses(kryo) } + } catch { + case _: Exception => println("Failed to register spark.kryo.registrator") } - kryo.register(mutable.HashMap().getClass, new ScalaMapSerializer(mutable.HashMap() ++ _)) - // TODO: add support for immutable maps too; this is more annoying because there are many - // subclasses of immutable.Map for small maps (with <= 4 entries) - val map1 = Map[Any, Any](1 -> 1) - val map2 = Map[Any, Any](1 -> 1, 2 -> 2) - val map3 = Map[Any, Any](1 -> 1, 2 -> 2, 3 -> 3) - val map4 = Map[Any, Any](1 -> 1, 2 -> 2, 3 -> 3, 4 -> 4) - val map5 = Map[Any, Any](1 -> 1, 2 -> 2, 3 -> 3, 4 -> 4, 5 -> 5) - kryo.register(map1.getClass, new ScalaMapSerializer(mutable.HashMap() ++ _ toMap)) - kryo.register(map2.getClass, new ScalaMapSerializer(mutable.HashMap() ++ _ toMap)) - kryo.register(map3.getClass, new ScalaMapSerializer(mutable.HashMap() ++ _ toMap)) - kryo.register(map4.getClass, new ScalaMapSerializer(mutable.HashMap() ++ _ toMap)) - kryo.register(map5.getClass, new ScalaMapSerializer(mutable.HashMap() ++ _ toMap)) - // Allow the user to register their own classes by setting spark.kryo.registrator - val regCls = System.getProperty("spark.kryo.registrator") - if (regCls != null) { - logInfo("Running user registrator: " + regCls) - val classLoader = Thread.currentThread.getContextClassLoader - val reg = Class.forName(regCls, true, classLoader).newInstance().asInstanceOf[KryoRegistrator] - reg.registerClasses(kryo) - } + kryo.setClassLoader(classLoader) + + // Allow disabling Kryo reference tracking if user knows their object graphs don't have loops + kryo.setReferences(System.getProperty("spark.kryo.referenceTracking", "true").toBoolean) + kryo } def newInstance(): SerializerInstance = { - this.kryo.get().setClassLoader(Thread.currentThread().getContextClassLoader) new KryoSerializerInstance(this) } -} +}
\ No newline at end of file diff --git a/core/src/main/scala/spark/Logging.scala b/core/src/main/scala/spark/Logging.scala index 0fc8c31463..79b0362830 100644 --- a/core/src/main/scala/spark/Logging.scala +++ b/core/src/main/scala/spark/Logging.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import org.slf4j.Logger diff --git a/core/src/main/scala/spark/MapOutputTracker.scala b/core/src/main/scala/spark/MapOutputTracker.scala index fde597ffd1..2c417e31db 100644 --- a/core/src/main/scala/spark/MapOutputTracker.scala +++ b/core/src/main/scala/spark/MapOutputTracker.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import java.io._ diff --git a/core/src/main/scala/spark/PairRDDFunctions.scala b/core/src/main/scala/spark/PairRDDFunctions.scala index 8b313c645f..6b0cc2fbf1 100644 --- a/core/src/main/scala/spark/PairRDDFunctions.scala +++ b/core/src/main/scala/spark/PairRDDFunctions.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import java.nio.ByteBuffer diff --git a/core/src/main/scala/spark/Partition.scala b/core/src/main/scala/spark/Partition.scala index e384308ef6..2a4edcec98 100644 --- a/core/src/main/scala/spark/Partition.scala +++ b/core/src/main/scala/spark/Partition.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark /** diff --git a/core/src/main/scala/spark/Partitioner.scala b/core/src/main/scala/spark/Partitioner.scala index 6f8cd17c88..660af70d52 100644 --- a/core/src/main/scala/spark/Partitioner.scala +++ b/core/src/main/scala/spark/Partitioner.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark /** diff --git a/core/src/main/scala/spark/RDD.scala b/core/src/main/scala/spark/RDD.scala index 106fb2960f..ca7cdd622a 100644 --- a/core/src/main/scala/spark/RDD.scala +++ b/core/src/main/scala/spark/RDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import java.util.Random @@ -778,10 +795,20 @@ abstract class RDD[T: ClassManifest]( }.reduce { (queue1, queue2) => queue1 ++= queue2 queue1 - }.toArray + }.toArray.sorted(ord.reverse) } /** + * Returns the first K elements from this RDD as defined by + * the specified implicit Ordering[T] and maintains the + * ordering. + * @param num the number of top elements to return + * @param ord the implicit ordering for T + * @return an array of top elements + */ + def takeOrdered(num: Int)(implicit ord: Ordering[T]): Array[T] = top(num)(ord.reverse) + + /** * Save this RDD as a text file, using string representations of elements. */ def saveAsTextFile(path: String) { diff --git a/core/src/main/scala/spark/RDDCheckpointData.scala b/core/src/main/scala/spark/RDDCheckpointData.scala index 57e0405fb4..b615f820eb 100644 --- a/core/src/main/scala/spark/RDDCheckpointData.scala +++ b/core/src/main/scala/spark/RDDCheckpointData.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import org.apache.hadoop.fs.Path diff --git a/core/src/main/scala/spark/SequenceFileRDDFunctions.scala b/core/src/main/scala/spark/SequenceFileRDDFunctions.scala index 2911f9036e..9f30b7f22f 100644 --- a/core/src/main/scala/spark/SequenceFileRDDFunctions.scala +++ b/core/src/main/scala/spark/SequenceFileRDDFunctions.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import java.io.EOFException diff --git a/core/src/main/scala/spark/SerializableWritable.scala b/core/src/main/scala/spark/SerializableWritable.scala index 8306fbf570..0236611ef9 100644 --- a/core/src/main/scala/spark/SerializableWritable.scala +++ b/core/src/main/scala/spark/SerializableWritable.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import java.io._ diff --git a/core/src/main/scala/spark/ShuffleFetcher.scala b/core/src/main/scala/spark/ShuffleFetcher.scala index 9513a00126..dcced035e7 100644 --- a/core/src/main/scala/spark/ShuffleFetcher.scala +++ b/core/src/main/scala/spark/ShuffleFetcher.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import spark.executor.TaskMetrics diff --git a/core/src/main/scala/spark/SizeEstimator.scala b/core/src/main/scala/spark/SizeEstimator.scala index f8a4c4e489..6cc57566d7 100644 --- a/core/src/main/scala/spark/SizeEstimator.scala +++ b/core/src/main/scala/spark/SizeEstimator.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import java.lang.reflect.Field diff --git a/core/src/main/scala/spark/SoftReferenceCache.scala b/core/src/main/scala/spark/SoftReferenceCache.scala deleted file mode 100644 index 3dd0a4b1f9..0000000000 --- a/core/src/main/scala/spark/SoftReferenceCache.scala +++ /dev/null @@ -1,18 +0,0 @@ -package spark - -import com.google.common.collect.MapMaker - -/** - * An implementation of Cache that uses soft references. - */ -private[spark] class SoftReferenceCache extends Cache { - val map = new MapMaker().softValues().makeMap[Any, Any]() - - override def get(datasetId: Any, partition: Int): Any = - map.get((datasetId, partition)) - - override def put(datasetId: Any, partition: Int, value: Any): CachePutResponse = { - map.put((datasetId, partition), value) - return CachePutSuccess(0) - } -} diff --git a/core/src/main/scala/spark/SparkContext.scala b/core/src/main/scala/spark/SparkContext.scala index 7747160290..0d1f9fa8d4 100644 --- a/core/src/main/scala/spark/SparkContext.scala +++ b/core/src/main/scala/spark/SparkContext.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import java.io._ @@ -44,15 +61,16 @@ import org.apache.mesos.MesosNativeLibrary import spark.deploy.{LocalSparkCluster, SparkHadoopUtil} import spark.partial.{ApproximateEvaluator, PartialResult} import spark.rdd.{CheckpointRDD, HadoopRDD, NewHadoopRDD, UnionRDD, ParallelCollectionRDD} -import spark.scheduler.{DAGScheduler, ResultTask, ShuffleMapTask, SparkListener, +import spark.scheduler.{DAGScheduler, DAGSchedulerSource, ResultTask, ShuffleMapTask, SparkListener, SplitInfo, Stage, StageInfo, TaskScheduler, ActiveJob} import spark.scheduler.cluster.{StandaloneSchedulerBackend, SparkDeploySchedulerBackend, ClusterScheduler, Schedulable, SchedulingMode} import spark.scheduler.local.LocalScheduler import spark.scheduler.mesos.{CoarseMesosSchedulerBackend, MesosSchedulerBackend} -import spark.storage.{StorageStatus, StorageUtils, RDDInfo} +import spark.storage.{StorageStatus, StorageUtils, RDDInfo, BlockManagerSource} import spark.util.{MetadataCleaner, TimeStampedHashMap} import ui.{SparkUI} +import spark.metrics._ /** * Main entry point for Spark functionality. A SparkContext represents the connection to a Spark @@ -256,6 +274,16 @@ class SparkContext( // Post init taskScheduler.postStartHook() + val dagSchedulerSource = new DAGSchedulerSource(this.dagScheduler) + val blockManagerSource = new BlockManagerSource(SparkEnv.get.blockManager) + + def initDriverMetrics() { + SparkEnv.get.metricsSystem.registerSource(dagSchedulerSource) + SparkEnv.get.metricsSystem.registerSource(blockManagerSource) + } + + initDriverMetrics() + // Methods for creating RDDs /** Distribute a local Scala collection to form an RDD. */ @@ -532,6 +560,12 @@ class SparkContext( StorageUtils.rddInfoFromStorageStatus(getExecutorStorageStatus, this) } + /** + * Returns an immutable map of RDDs that have marked themselves as persistent via cache() call. + * Note that this does not necessarily mean the caching or computation was successful. + */ + def getPersistentRDDs: Map[Int, RDD[_]] = persistentRdds.toMap + def getStageInfo: Map[Stage,StageInfo] = { dagScheduler.stageToInfos } @@ -545,7 +579,7 @@ class SparkContext( /** * Return pools for fair scheduler - * TODO:now, we have not taken nested pools into account + * TODO(xiajunluan):now, we have not taken nested pools into account */ def getPools: ArrayBuffer[Schedulable] = { taskScheduler.rootPool.schedulableQueue @@ -582,7 +616,12 @@ class SparkContext( } else { val uri = new URI(path) val key = uri.getScheme match { - case null | "file" => env.httpFileServer.addJar(new File(uri.getPath)) + case null | "file" => + if (SparkHadoopUtil.isYarnMode()) { + logWarning("local jar specified as parameter to addJar under Yarn mode") + return + } + env.httpFileServer.addJar(new File(uri.getPath)) case _ => path } addedJars(key) = System.currentTimeMillis diff --git a/core/src/main/scala/spark/SparkEnv.scala b/core/src/main/scala/spark/SparkEnv.scala index ec59b4f48f..4a1d341f5d 100644 --- a/core/src/main/scala/spark/SparkEnv.scala +++ b/core/src/main/scala/spark/SparkEnv.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import collection.mutable @@ -7,6 +24,7 @@ import akka.actor.{Actor, ActorRef, Props, ActorSystemImpl, ActorSystem} import akka.remote.RemoteActorRefProvider import spark.broadcast.BroadcastManager +import spark.metrics.MetricsSystem import spark.storage.BlockManager import spark.storage.BlockManagerMaster import spark.network.ConnectionManager @@ -36,6 +54,7 @@ class SparkEnv ( val connectionManager: ConnectionManager, val httpFileServer: HttpFileServer, val sparkFilesDir: String, + val metricsSystem: MetricsSystem, // To be set only as part of initialization of SparkContext. // (executorId, defaultHostPort) => executorHostPort // If executorId is NOT found, return defaultHostPort @@ -51,6 +70,7 @@ class SparkEnv ( broadcastManager.stop() blockManager.stop() blockManager.master.stop() + metricsSystem.stop() actorSystem.shutdown() // Unfortunately Akka's awaitTermination doesn't actually wait for the Netty server to shut // down, but let's call it anyway in case it gets fixed in a later release @@ -167,6 +187,13 @@ object SparkEnv extends Logging { httpFileServer.initialize() System.setProperty("spark.fileserver.uri", httpFileServer.serverUri) + val metricsSystem = if (isDriver) { + MetricsSystem.createMetricsSystem("driver") + } else { + MetricsSystem.createMetricsSystem("executor") + } + metricsSystem.start() + // Set the sparkFiles directory, used when downloading dependencies. In local mode, // this is a temporary directory; in distributed mode, this is the executor's current working // directory. @@ -196,6 +223,7 @@ object SparkEnv extends Logging { connectionManager, httpFileServer, sparkFilesDir, + metricsSystem, None) } } diff --git a/core/src/main/scala/spark/SparkException.scala b/core/src/main/scala/spark/SparkException.scala index aa7a16d7dd..b7045eea63 100644 --- a/core/src/main/scala/spark/SparkException.scala +++ b/core/src/main/scala/spark/SparkException.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark class SparkException(message: String, cause: Throwable) diff --git a/core/src/main/scala/spark/SparkFiles.java b/core/src/main/scala/spark/SparkFiles.java index 566aec622c..f9b3f7965e 100644 --- a/core/src/main/scala/spark/SparkFiles.java +++ b/core/src/main/scala/spark/SparkFiles.java @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark; import java.io.File; diff --git a/core/src/main/scala/spark/TaskContext.scala b/core/src/main/scala/spark/TaskContext.scala index dd0609026a..b79f4ca813 100644 --- a/core/src/main/scala/spark/TaskContext.scala +++ b/core/src/main/scala/spark/TaskContext.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import executor.TaskMetrics diff --git a/core/src/main/scala/spark/TaskEndReason.scala b/core/src/main/scala/spark/TaskEndReason.scala index bb75ec208c..3ad665da34 100644 --- a/core/src/main/scala/spark/TaskEndReason.scala +++ b/core/src/main/scala/spark/TaskEndReason.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import spark.executor.TaskMetrics diff --git a/core/src/main/scala/spark/TaskState.scala b/core/src/main/scala/spark/TaskState.scala index 78eb33a628..9df7d8277b 100644 --- a/core/src/main/scala/spark/TaskState.scala +++ b/core/src/main/scala/spark/TaskState.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import org.apache.mesos.Protos.{TaskState => MesosTaskState} diff --git a/core/src/main/scala/spark/Utils.scala b/core/src/main/scala/spark/Utils.scala index 64547bbdcd..e6a96a5ec1 100644 --- a/core/src/main/scala/spark/Utils.scala +++ b/core/src/main/scala/spark/Utils.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import java.io._ @@ -6,6 +23,7 @@ import java.util.{Locale, Random, UUID} import java.util.concurrent.{ConcurrentHashMap, Executors, ThreadFactory, ThreadPoolExecutor} import java.util.regex.Pattern +import scala.collection.Map import scala.collection.mutable.{ArrayBuffer, HashMap} import scala.collection.JavaConversions._ import scala.io.Source @@ -545,10 +563,15 @@ private object Utils extends Logging { /** * Execute a command and get its output, throwing an exception if it yields a code other than 0. */ - def executeAndGetOutput(command: Seq[String], workingDir: File = new File(".")): String = { - val process = new ProcessBuilder(command: _*) + def executeAndGetOutput(command: Seq[String], workingDir: File = new File("."), + extraEnvironment: Map[String, String] = Map.empty): String = { + val builder = new ProcessBuilder(command: _*) .directory(workingDir) - .start() + val environment = builder.environment() + for ((key, value) <- extraEnvironment) { + environment.put(key, value) + } + val process = builder.start() new Thread("read stderr for " + command(0)) { override def run() { for (line <- Source.fromInputStream(process.getErrorStream).getLines) { @@ -573,8 +596,15 @@ private object Utils extends Logging { output.toString } + /** + * A regular expression to match classes of the "core" Spark API that we want to skip when + * finding the call site of a method. + */ + private val SPARK_CLASS_REGEX = """^spark(\.api\.java)?(\.rdd)?\.[A-Z]""".r + private[spark] class CallSiteInfo(val lastSparkMethod: String, val firstUserFile: String, val firstUserLine: Int, val firstUserClass: String) + /** * When called inside a class in the spark package, returns the name of the user code class * (outside the spark package) that called into Spark, as well as which Spark method they called. @@ -596,7 +626,7 @@ private object Utils extends Logging { for (el <- trace) { if (!finished) { - if (el.getClassName.startsWith("spark.") && !el.getClassName.startsWith("spark.examples.")) { + if (SPARK_CLASS_REGEX.findFirstIn(el.getClassName) != None) { lastSparkMethod = if (el.getMethodName == "<init>") { // Spark method is a constructor; get its class name el.getClassName.substring(el.getClassName.lastIndexOf('.') + 1) @@ -621,15 +651,16 @@ private object Utils extends Logging { callSiteInfo.firstUserLine) } - /** Return a string containing the last `n` bytes of a file. */ - def lastNBytes(path: String, n: Int): String = { + /** Return a string containing part of a file from byte 'start' to 'end'. */ + def offsetBytes(path: String, start: Long, end: Long): String = { val file = new File(path) val length = file.length() - val buff = new Array[Byte](math.min(n, length.toInt)) - val skip = math.max(0, length - n) + val effectiveEnd = math.min(length, end) + val effectiveStart = math.max(0, start) + val buff = new Array[Byte]((effectiveEnd-effectiveStart).toInt) val stream = new FileInputStream(file) - stream.skip(skip) + stream.skip(effectiveStart) stream.read(buff) stream.close() Source.fromBytes(buff).mkString diff --git a/core/src/main/scala/spark/api/java/JavaDoubleRDD.scala b/core/src/main/scala/spark/api/java/JavaDoubleRDD.scala index 16692c0440..8ce7df6213 100644 --- a/core/src/main/scala/spark/api/java/JavaDoubleRDD.scala +++ b/core/src/main/scala/spark/api/java/JavaDoubleRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.api.java import spark.RDD @@ -98,33 +115,48 @@ class JavaDoubleRDD(val srdd: RDD[scala.Double]) extends JavaRDDLike[Double, Jav // Double RDD functions - /** Return the sum of the elements in this RDD. */ + /** Add up the elements in this RDD. */ def sum(): Double = srdd.sum() - /** Return a [[spark.StatCounter]] describing the elements in this RDD. */ + /** + * Return a [[spark.util.StatCounter]] object that captures the mean, variance and count + * of the RDD's elements in one operation. + */ def stats(): StatCounter = srdd.stats() - /** Return the mean of the elements in this RDD. */ + /** Compute the mean of this RDD's elements. */ def mean(): Double = srdd.mean() - /** Return the variance of the elements in this RDD. */ + /** Compute the variance of this RDD's elements. */ def variance(): Double = srdd.variance() - /** Return the standard deviation of the elements in this RDD. */ + /** Compute the standard deviation of this RDD's elements. */ def stdev(): Double = srdd.stdev() + /** + * Compute the sample standard deviation of this RDD's elements (which corrects for bias in + * estimating the standard deviation by dividing by N-1 instead of N). + */ + def sampleStdev(): Double = srdd.sampleStdev() + + /** + * Compute the sample variance of this RDD's elements (which corrects for bias in + * estimating the standard variance by dividing by N-1 instead of N). + */ + def sampleVariance(): Double = srdd.sampleVariance() + /** Return the approximate mean of the elements in this RDD. */ def meanApprox(timeout: Long, confidence: Double): PartialResult[BoundedDouble] = srdd.meanApprox(timeout, confidence) - /** Return the approximate mean of the elements in this RDD. */ + /** (Experimental) Approximate operation to return the mean within a timeout. */ def meanApprox(timeout: Long): PartialResult[BoundedDouble] = srdd.meanApprox(timeout) - /** Return the approximate sum of the elements in this RDD. */ + /** (Experimental) Approximate operation to return the sum within a timeout. */ def sumApprox(timeout: Long, confidence: Double): PartialResult[BoundedDouble] = srdd.sumApprox(timeout, confidence) - - /** Return the approximate sum of the elements in this RDD. */ + + /** (Experimental) Approximate operation to return the sum within a timeout. */ def sumApprox(timeout: Long): PartialResult[BoundedDouble] = srdd.sumApprox(timeout) } diff --git a/core/src/main/scala/spark/api/java/JavaPairRDD.scala b/core/src/main/scala/spark/api/java/JavaPairRDD.scala index 76051597b6..ccc511dc5f 100644 --- a/core/src/main/scala/spark/api/java/JavaPairRDD.scala +++ b/core/src/main/scala/spark/api/java/JavaPairRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.api.java import java.util.{List => JList} diff --git a/core/src/main/scala/spark/api/java/JavaRDD.scala b/core/src/main/scala/spark/api/java/JavaRDD.scala index 626b499454..c0bf2cf568 100644 --- a/core/src/main/scala/spark/api/java/JavaRDD.scala +++ b/core/src/main/scala/spark/api/java/JavaRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.api.java import spark._ diff --git a/core/src/main/scala/spark/api/java/JavaRDDLike.scala b/core/src/main/scala/spark/api/java/JavaRDDLike.scala index b555f2030a..21b5abf053 100644 --- a/core/src/main/scala/spark/api/java/JavaRDDLike.scala +++ b/core/src/main/scala/spark/api/java/JavaRDDLike.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.api.java import java.util.{List => JList, Comparator} @@ -384,4 +401,29 @@ trait JavaRDDLike[T, This <: JavaRDDLike[T, This]] extends Serializable { val comp = com.google.common.collect.Ordering.natural().asInstanceOf[Comparator[T]] top(num, comp) } + + /** + * Returns the first K elements from this RDD as defined by + * the specified Comparator[T] and maintains the order. + * @param num the number of top elements to return + * @param comp the comparator that defines the order + * @return an array of top elements + */ + def takeOrdered(num: Int, comp: Comparator[T]): JList[T] = { + import scala.collection.JavaConversions._ + val topElems = rdd.takeOrdered(num)(Ordering.comparatorToOrdering(comp)) + val arr: java.util.Collection[T] = topElems.toSeq + new java.util.ArrayList(arr) + } + + /** + * Returns the first K elements from this RDD using the + * natural ordering for T while maintain the order. + * @param num the number of top elements to return + * @return an array of top elements + */ + def takeOrdered(num: Int): JList[T] = { + val comp = com.google.common.collect.Ordering.natural().asInstanceOf[Comparator[T]] + takeOrdered(num, comp) + } } diff --git a/core/src/main/scala/spark/api/java/JavaSparkContext.scala b/core/src/main/scala/spark/api/java/JavaSparkContext.scala index 5f18b1e15b..fe182e7ab6 100644 --- a/core/src/main/scala/spark/api/java/JavaSparkContext.scala +++ b/core/src/main/scala/spark/api/java/JavaSparkContext.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.api.java import java.util.{Map => JMap} diff --git a/core/src/main/scala/spark/api/java/JavaSparkContextVarargsWorkaround.java b/core/src/main/scala/spark/api/java/JavaSparkContextVarargsWorkaround.java index 97344e73da..42b1de01b1 100644 --- a/core/src/main/scala/spark/api/java/JavaSparkContextVarargsWorkaround.java +++ b/core/src/main/scala/spark/api/java/JavaSparkContextVarargsWorkaround.java @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.api.java; import java.util.Arrays; diff --git a/core/src/main/scala/spark/api/java/StorageLevels.java b/core/src/main/scala/spark/api/java/StorageLevels.java index 5e5845ac3a..f385636e83 100644 --- a/core/src/main/scala/spark/api/java/StorageLevels.java +++ b/core/src/main/scala/spark/api/java/StorageLevels.java @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.api.java; import spark.storage.StorageLevel; diff --git a/core/src/main/scala/spark/api/java/function/DoubleFlatMapFunction.java b/core/src/main/scala/spark/api/java/function/DoubleFlatMapFunction.java index 3a8192be3a..8bc88d757f 100644 --- a/core/src/main/scala/spark/api/java/function/DoubleFlatMapFunction.java +++ b/core/src/main/scala/spark/api/java/function/DoubleFlatMapFunction.java @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.api.java.function; diff --git a/core/src/main/scala/spark/api/java/function/DoubleFunction.java b/core/src/main/scala/spark/api/java/function/DoubleFunction.java index c6ef76d088..1aa1e5dae0 100644 --- a/core/src/main/scala/spark/api/java/function/DoubleFunction.java +++ b/core/src/main/scala/spark/api/java/function/DoubleFunction.java @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.api.java.function; diff --git a/core/src/main/scala/spark/api/java/function/FlatMapFunction.scala b/core/src/main/scala/spark/api/java/function/FlatMapFunction.scala index e027cdacd3..9eb0cfe3f9 100644 --- a/core/src/main/scala/spark/api/java/function/FlatMapFunction.scala +++ b/core/src/main/scala/spark/api/java/function/FlatMapFunction.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.api.java.function /** diff --git a/core/src/main/scala/spark/api/java/function/FlatMapFunction2.scala b/core/src/main/scala/spark/api/java/function/FlatMapFunction2.scala index 6044043add..dda98710c2 100644 --- a/core/src/main/scala/spark/api/java/function/FlatMapFunction2.scala +++ b/core/src/main/scala/spark/api/java/function/FlatMapFunction2.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.api.java.function /** diff --git a/core/src/main/scala/spark/api/java/function/Function.java b/core/src/main/scala/spark/api/java/function/Function.java index dae8295f21..2a2ea0aacf 100644 --- a/core/src/main/scala/spark/api/java/function/Function.java +++ b/core/src/main/scala/spark/api/java/function/Function.java @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.api.java.function; import scala.reflect.ClassManifest; diff --git a/core/src/main/scala/spark/api/java/function/Function2.java b/core/src/main/scala/spark/api/java/function/Function2.java index 69bf12c8c9..952d31ece4 100644 --- a/core/src/main/scala/spark/api/java/function/Function2.java +++ b/core/src/main/scala/spark/api/java/function/Function2.java @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.api.java.function; import scala.reflect.ClassManifest; diff --git a/core/src/main/scala/spark/api/java/function/PairFlatMapFunction.java b/core/src/main/scala/spark/api/java/function/PairFlatMapFunction.java index b3cc4df6aa..4aad602da3 100644 --- a/core/src/main/scala/spark/api/java/function/PairFlatMapFunction.java +++ b/core/src/main/scala/spark/api/java/function/PairFlatMapFunction.java @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.api.java.function; import scala.Tuple2; diff --git a/core/src/main/scala/spark/api/java/function/PairFunction.java b/core/src/main/scala/spark/api/java/function/PairFunction.java index 9fc6df4b88..ccfe64ecf1 100644 --- a/core/src/main/scala/spark/api/java/function/PairFunction.java +++ b/core/src/main/scala/spark/api/java/function/PairFunction.java @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.api.java.function; import scala.Tuple2; diff --git a/core/src/main/scala/spark/api/java/function/VoidFunction.scala b/core/src/main/scala/spark/api/java/function/VoidFunction.scala index b0096cf2bf..f6fc0b0f7d 100644 --- a/core/src/main/scala/spark/api/java/function/VoidFunction.scala +++ b/core/src/main/scala/spark/api/java/function/VoidFunction.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.api.java.function /** @@ -13,4 +30,4 @@ abstract class VoidFunction[T] extends Serializable { // return Unit), so it is implicitly converted to a Function1[T, Unit]: object VoidFunction { implicit def toFunction[T](f: VoidFunction[T]) : Function1[T, Unit] = ((x : T) => f.call(x)) -}
\ No newline at end of file +} diff --git a/core/src/main/scala/spark/api/java/function/WrappedFunction1.scala b/core/src/main/scala/spark/api/java/function/WrappedFunction1.scala index 923f5cdf4f..1758a38c4e 100644 --- a/core/src/main/scala/spark/api/java/function/WrappedFunction1.scala +++ b/core/src/main/scala/spark/api/java/function/WrappedFunction1.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.api.java.function import scala.runtime.AbstractFunction1 diff --git a/core/src/main/scala/spark/api/java/function/WrappedFunction2.scala b/core/src/main/scala/spark/api/java/function/WrappedFunction2.scala index 2c6e9b1571..b093567d2c 100644 --- a/core/src/main/scala/spark/api/java/function/WrappedFunction2.scala +++ b/core/src/main/scala/spark/api/java/function/WrappedFunction2.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.api.java.function import scala.runtime.AbstractFunction2 diff --git a/core/src/main/scala/spark/api/python/PythonPartitioner.scala b/core/src/main/scala/spark/api/python/PythonPartitioner.scala index d618c098c2..31a719fbff 100644 --- a/core/src/main/scala/spark/api/python/PythonPartitioner.scala +++ b/core/src/main/scala/spark/api/python/PythonPartitioner.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.api.python import spark.Partitioner diff --git a/core/src/main/scala/spark/api/python/PythonRDD.scala b/core/src/main/scala/spark/api/python/PythonRDD.scala index 31d8ea89d4..2dd79f7100 100644 --- a/core/src/main/scala/spark/api/python/PythonRDD.scala +++ b/core/src/main/scala/spark/api/python/PythonRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.api.python import java.io._ @@ -46,34 +63,42 @@ private[spark] class PythonRDD[T: ClassManifest]( // Start a thread to feed the process input from our parent's iterator new Thread("stdin writer for " + pythonExec) { override def run() { - SparkEnv.set(env) - val stream = new BufferedOutputStream(worker.getOutputStream, bufferSize) - val dataOut = new DataOutputStream(stream) - val printOut = new PrintWriter(stream) - // Partition index - dataOut.writeInt(split.index) - // sparkFilesDir - PythonRDD.writeAsPickle(SparkFiles.getRootDirectory, dataOut) - // Broadcast variables - dataOut.writeInt(broadcastVars.length) - for (broadcast <- broadcastVars) { - dataOut.writeLong(broadcast.id) - dataOut.writeInt(broadcast.value.length) - dataOut.write(broadcast.value) - } - dataOut.flush() - // Serialized user code - for (elem <- command) { - printOut.println(elem) - } - printOut.flush() - // Data values - for (elem <- parent.iterator(split, context)) { - PythonRDD.writeAsPickle(elem, dataOut) + try { + SparkEnv.set(env) + val stream = new BufferedOutputStream(worker.getOutputStream, bufferSize) + val dataOut = new DataOutputStream(stream) + val printOut = new PrintWriter(stream) + // Partition index + dataOut.writeInt(split.index) + // sparkFilesDir + PythonRDD.writeAsPickle(SparkFiles.getRootDirectory, dataOut) + // Broadcast variables + dataOut.writeInt(broadcastVars.length) + for (broadcast <- broadcastVars) { + dataOut.writeLong(broadcast.id) + dataOut.writeInt(broadcast.value.length) + dataOut.write(broadcast.value) + } + dataOut.flush() + // Serialized user code + for (elem <- command) { + printOut.println(elem) + } + printOut.flush() + // Data values + for (elem <- parent.iterator(split, context)) { + PythonRDD.writeAsPickle(elem, dataOut) + } + dataOut.flush() + printOut.flush() + worker.shutdownOutput() + } catch { + case e: IOException => + // This can happen for legitimate reasons if the Python code stops returning data before we are done + // passing elements through, e.g., for take(). Just log a message to say it happened. + logInfo("stdin writer to Python finished early") + logDebug("stdin writer to Python finished early", e) } - dataOut.flush() - printOut.flush() - worker.shutdownOutput() } }.start() @@ -280,7 +305,7 @@ class PythonAccumulatorParam(@transient serverHost: String, serverPort: Int) Utils.checkHost(serverHost, "Expected hostname") val bufferSize = System.getProperty("spark.buffer.size", "65536").toInt - + override def zero(value: JList[Array[Byte]]): JList[Array[Byte]] = new JArrayList override def addInPlace(val1: JList[Array[Byte]], val2: JList[Array[Byte]]) diff --git a/core/src/main/scala/spark/api/python/PythonWorkerFactory.scala b/core/src/main/scala/spark/api/python/PythonWorkerFactory.scala index 85d1dfeac8..14f8320678 100644 --- a/core/src/main/scala/spark/api/python/PythonWorkerFactory.scala +++ b/core/src/main/scala/spark/api/python/PythonWorkerFactory.scala @@ -1,6 +1,23 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.api.python -import java.io.{DataInputStream, IOException} +import java.io.{File, DataInputStream, IOException} import java.net.{Socket, SocketException, InetAddress} import scala.collection.JavaConversions._ @@ -50,6 +67,8 @@ private[spark] class PythonWorkerFactory(pythonExec: String, envVars: Map[String val pb = new ProcessBuilder(Seq(pythonExec, sparkHome + "/python/pyspark/daemon.py")) val workerEnv = pb.environment() workerEnv.putAll(envVars) + val pythonPath = sparkHome + "/python/" + File.pathSeparator + workerEnv.get("PYTHONPATH") + workerEnv.put("PYTHONPATH", pythonPath) daemon = pb.start() // Redirect the stderr to ours diff --git a/core/src/main/scala/spark/broadcast/BitTorrentBroadcast.scala b/core/src/main/scala/spark/broadcast/BitTorrentBroadcast.scala index adcb2d2415..6f7d385379 100644 --- a/core/src/main/scala/spark/broadcast/BitTorrentBroadcast.scala +++ b/core/src/main/scala/spark/broadcast/BitTorrentBroadcast.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.broadcast import java.io._ diff --git a/core/src/main/scala/spark/broadcast/Broadcast.scala b/core/src/main/scala/spark/broadcast/Broadcast.scala index 415bde5d67..aba56a60ca 100644 --- a/core/src/main/scala/spark/broadcast/Broadcast.scala +++ b/core/src/main/scala/spark/broadcast/Broadcast.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.broadcast import java.io._ diff --git a/core/src/main/scala/spark/broadcast/BroadcastFactory.scala b/core/src/main/scala/spark/broadcast/BroadcastFactory.scala index 5c6184c3c7..d33d95c7d9 100644 --- a/core/src/main/scala/spark/broadcast/BroadcastFactory.scala +++ b/core/src/main/scala/spark/broadcast/BroadcastFactory.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.broadcast /** diff --git a/core/src/main/scala/spark/broadcast/HttpBroadcast.scala b/core/src/main/scala/spark/broadcast/HttpBroadcast.scala index 7e30b8f7d2..c565876950 100644 --- a/core/src/main/scala/spark/broadcast/HttpBroadcast.scala +++ b/core/src/main/scala/spark/broadcast/HttpBroadcast.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.broadcast import com.ning.compress.lzf.{LZFInputStream, LZFOutputStream} diff --git a/core/src/main/scala/spark/broadcast/MultiTracker.scala b/core/src/main/scala/spark/broadcast/MultiTracker.scala index 3fd77af73f..7855d44e9b 100644 --- a/core/src/main/scala/spark/broadcast/MultiTracker.scala +++ b/core/src/main/scala/spark/broadcast/MultiTracker.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.broadcast import java.io._ diff --git a/core/src/main/scala/spark/broadcast/SourceInfo.scala b/core/src/main/scala/spark/broadcast/SourceInfo.scala index c79bb93c38..b17ae63b5c 100644 --- a/core/src/main/scala/spark/broadcast/SourceInfo.scala +++ b/core/src/main/scala/spark/broadcast/SourceInfo.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.broadcast import java.util.BitSet diff --git a/core/src/main/scala/spark/broadcast/TreeBroadcast.scala b/core/src/main/scala/spark/broadcast/TreeBroadcast.scala index c55c476117..ea1e9a12c1 100644 --- a/core/src/main/scala/spark/broadcast/TreeBroadcast.scala +++ b/core/src/main/scala/spark/broadcast/TreeBroadcast.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.broadcast import java.io._ diff --git a/core/src/main/scala/spark/deploy/ApplicationDescription.scala b/core/src/main/scala/spark/deploy/ApplicationDescription.scala index 02193c7008..a8b22fbef8 100644 --- a/core/src/main/scala/spark/deploy/ApplicationDescription.scala +++ b/core/src/main/scala/spark/deploy/ApplicationDescription.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy private[spark] class ApplicationDescription( diff --git a/core/src/main/scala/spark/deploy/Command.scala b/core/src/main/scala/spark/deploy/Command.scala index 577101e3c3..bad629e965 100644 --- a/core/src/main/scala/spark/deploy/Command.scala +++ b/core/src/main/scala/spark/deploy/Command.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy import scala.collection.Map diff --git a/core/src/main/scala/spark/deploy/DeployMessage.scala b/core/src/main/scala/spark/deploy/DeployMessage.scala index 51274acb1e..7c37a16615 100644 --- a/core/src/main/scala/spark/deploy/DeployMessage.scala +++ b/core/src/main/scala/spark/deploy/DeployMessage.scala @@ -1,108 +1,123 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy +import scala.collection.immutable.List + +import spark.Utils import spark.deploy.ExecutorState.ExecutorState import spark.deploy.master.{WorkerInfo, ApplicationInfo} import spark.deploy.worker.ExecutorRunner -import scala.collection.immutable.List -import spark.Utils -private[spark] sealed trait DeployMessage extends Serializable +private[deploy] sealed trait DeployMessage extends Serializable -// Worker to Master +private[deploy] object DeployMessages { -private[spark] -case class RegisterWorker( - id: String, - host: String, - port: Int, - cores: Int, - memory: Int, - webUiPort: Int, - publicAddress: String) - extends DeployMessage { - Utils.checkHost(host, "Required hostname") - assert (port > 0) -} + // Worker to Master -private[spark] -case class ExecutorStateChanged( - appId: String, - execId: Int, - state: ExecutorState, - message: Option[String], - exitStatus: Option[Int]) - extends DeployMessage + case class RegisterWorker( + id: String, + host: String, + port: Int, + cores: Int, + memory: Int, + webUiPort: Int, + publicAddress: String) + extends DeployMessage { + Utils.checkHost(host, "Required hostname") + assert (port > 0) + } -private[spark] case class Heartbeat(workerId: String) extends DeployMessage + case class ExecutorStateChanged( + appId: String, + execId: Int, + state: ExecutorState, + message: Option[String], + exitStatus: Option[Int]) + extends DeployMessage -// Master to Worker + case class Heartbeat(workerId: String) extends DeployMessage -private[spark] case class RegisteredWorker(masterWebUiUrl: String) extends DeployMessage -private[spark] case class RegisterWorkerFailed(message: String) extends DeployMessage -private[spark] case class KillExecutor(appId: String, execId: Int) extends DeployMessage + // Master to Worker -private[spark] case class LaunchExecutor( - appId: String, - execId: Int, - appDesc: ApplicationDescription, - cores: Int, - memory: Int, - sparkHome: String) - extends DeployMessage + case class RegisteredWorker(masterWebUiUrl: String) extends DeployMessage -// Client to Master + case class RegisterWorkerFailed(message: String) extends DeployMessage -private[spark] case class RegisterApplication(appDescription: ApplicationDescription) - extends DeployMessage + case class KillExecutor(appId: String, execId: Int) extends DeployMessage -// Master to Client + case class LaunchExecutor( + appId: String, + execId: Int, + appDesc: ApplicationDescription, + cores: Int, + memory: Int, + sparkHome: String) + extends DeployMessage -private[spark] -case class RegisteredApplication(appId: String) extends DeployMessage + // Client to Master -private[spark] -case class ExecutorAdded(id: Int, workerId: String, hostPort: String, cores: Int, memory: Int) { - Utils.checkHostPort(hostPort, "Required hostport") -} + case class RegisterApplication(appDescription: ApplicationDescription) + extends DeployMessage -private[spark] -case class ExecutorUpdated(id: Int, state: ExecutorState, message: Option[String], - exitStatus: Option[Int]) + // Master to Client -private[spark] -case class ApplicationRemoved(message: String) + case class RegisteredApplication(appId: String) extends DeployMessage -// Internal message in Client + case class ExecutorAdded(id: Int, workerId: String, hostPort: String, cores: Int, memory: Int) { + Utils.checkHostPort(hostPort, "Required hostport") + } -private[spark] case object StopClient + case class ExecutorUpdated(id: Int, state: ExecutorState, message: Option[String], + exitStatus: Option[Int]) -// MasterWebUI To Master + case class ApplicationRemoved(message: String) -private[spark] case object RequestMasterState + // Internal message in Client -// Master to MasterWebUI + case object StopClient -private[spark] -case class MasterState(host: String, port: Int, workers: Array[WorkerInfo], - activeApps: Array[ApplicationInfo], completedApps: Array[ApplicationInfo]) { + // MasterWebUI To Master - Utils.checkHost(host, "Required hostname") - assert (port > 0) + case object RequestMasterState - def uri = "spark://" + host + ":" + port -} + // Master to MasterWebUI + + case class MasterStateResponse(host: String, port: Int, workers: Array[WorkerInfo], + activeApps: Array[ApplicationInfo], completedApps: Array[ApplicationInfo]) { + + Utils.checkHost(host, "Required hostname") + assert (port > 0) + + def uri = "spark://" + host + ":" + port + } -// WorkerWebUI to Worker -private[spark] case object RequestWorkerState + // WorkerWebUI to Worker + case object RequestWorkerState -// Worker to WorkerWebUI + // Worker to WorkerWebUI -private[spark] -case class WorkerState(host: String, port: Int, workerId: String, executors: List[ExecutorRunner], - finishedExecutors: List[ExecutorRunner], masterUrl: String, cores: Int, memory: Int, - coresUsed: Int, memoryUsed: Int, masterWebUiUrl: String) { + case class WorkerStateResponse(host: String, port: Int, workerId: String, + executors: List[ExecutorRunner], finishedExecutors: List[ExecutorRunner], masterUrl: String, + cores: Int, memory: Int, coresUsed: Int, memoryUsed: Int, masterWebUiUrl: String) { - Utils.checkHost(host, "Required hostname") - assert (port > 0) + Utils.checkHost(host, "Required hostname") + assert (port > 0) + } } diff --git a/core/src/main/scala/spark/deploy/ExecutorState.scala b/core/src/main/scala/spark/deploy/ExecutorState.scala index 5dc0c54552..08c9a3b725 100644 --- a/core/src/main/scala/spark/deploy/ExecutorState.scala +++ b/core/src/main/scala/spark/deploy/ExecutorState.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy private[spark] object ExecutorState diff --git a/core/src/main/scala/spark/deploy/JsonProtocol.scala b/core/src/main/scala/spark/deploy/JsonProtocol.scala index 335aaeb769..bd1db7c294 100644 --- a/core/src/main/scala/spark/deploy/JsonProtocol.scala +++ b/core/src/main/scala/spark/deploy/JsonProtocol.scala @@ -1,8 +1,28 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy -import master.{ApplicationInfo, WorkerInfo} import net.liftweb.json.JsonDSL._ -import worker.ExecutorRunner + +import spark.deploy.DeployMessages.{MasterStateResponse, WorkerStateResponse} +import spark.deploy.master.{ApplicationInfo, WorkerInfo} +import spark.deploy.worker.ExecutorRunner + private[spark] object JsonProtocol { def writeWorkerInfo(obj: WorkerInfo) = { @@ -40,7 +60,7 @@ private[spark] object JsonProtocol { ("appdesc" -> writeApplicationDescription(obj.appDesc)) } - def writeMasterState(obj: MasterState) = { + def writeMasterState(obj: MasterStateResponse) = { ("url" -> ("spark://" + obj.uri)) ~ ("workers" -> obj.workers.toList.map(writeWorkerInfo)) ~ ("cores" -> obj.workers.map(_.cores).sum) ~ @@ -51,7 +71,7 @@ private[spark] object JsonProtocol { ("completedapps" -> obj.completedApps.toList.map(writeApplicationInfo)) } - def writeWorkerState(obj: WorkerState) = { + def writeWorkerState(obj: WorkerStateResponse) = { ("id" -> obj.workerId) ~ ("masterurl" -> obj.masterUrl) ~ ("masterwebuiurl" -> obj.masterWebUiUrl) ~ @@ -62,4 +82,4 @@ private[spark] object JsonProtocol { ("executors" -> obj.executors.toList.map(writeExecutorRunner)) ~ ("finishedexecutors" -> obj.finishedExecutors.toList.map(writeExecutorRunner)) } -}
\ No newline at end of file +} diff --git a/core/src/main/scala/spark/deploy/LocalSparkCluster.scala b/core/src/main/scala/spark/deploy/LocalSparkCluster.scala index 55bb61b0cc..6b8e9f27af 100644 --- a/core/src/main/scala/spark/deploy/LocalSparkCluster.scala +++ b/core/src/main/scala/spark/deploy/LocalSparkCluster.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy import akka.actor.{ActorRef, Props, Actor, ActorSystem, Terminated} @@ -45,6 +62,7 @@ class LocalSparkCluster(numWorkers: Int, coresPerWorker: Int, memoryPerWorker: I // Stop the workers before the master so they don't get upset that it disconnected workerActorSystems.foreach(_.shutdown()) workerActorSystems.foreach(_.awaitTermination()) + masterActorSystems.foreach(_.shutdown()) masterActorSystems.foreach(_.awaitTermination()) } diff --git a/core/src/main/scala/spark/deploy/WebUI.scala b/core/src/main/scala/spark/deploy/WebUI.scala index 844c4142c7..8ea7792ef4 100644 --- a/core/src/main/scala/spark/deploy/WebUI.scala +++ b/core/src/main/scala/spark/deploy/WebUI.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy import java.text.SimpleDateFormat diff --git a/core/src/main/scala/spark/deploy/client/Client.scala b/core/src/main/scala/spark/deploy/client/Client.scala index 4af44f9c16..9d5ba8a796 100644 --- a/core/src/main/scala/spark/deploy/client/Client.scala +++ b/core/src/main/scala/spark/deploy/client/Client.scala @@ -1,20 +1,39 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.client -import spark.deploy._ +import java.util.concurrent.TimeoutException + import akka.actor._ +import akka.actor.Terminated import akka.pattern.ask import akka.util.Duration -import akka.util.duration._ -import akka.pattern.AskTimeoutException -import spark.{SparkException, Logging} +import akka.remote.RemoteClientDisconnected import akka.remote.RemoteClientLifeCycleEvent import akka.remote.RemoteClientShutdown -import spark.deploy.RegisterApplication -import spark.deploy.master.Master -import akka.remote.RemoteClientDisconnected -import akka.actor.Terminated import akka.dispatch.Await +import spark.Logging +import spark.deploy.{ApplicationDescription, ExecutorState} +import spark.deploy.DeployMessages._ +import spark.deploy.master.Master + + /** * The main class used to talk to a Spark deploy cluster. Takes a master URL, an app description, * and a listener for cluster events, and calls back the listener when various events occur. @@ -117,7 +136,8 @@ private[spark] class Client( val future = actor.ask(StopClient)(timeout) Await.result(future, timeout) } catch { - case e: AskTimeoutException => // Ignore it, maybe master went away + case e: TimeoutException => + logInfo("Stop request to Master timed out; it may already be shut down.") } actor = null } diff --git a/core/src/main/scala/spark/deploy/client/ClientListener.scala b/core/src/main/scala/spark/deploy/client/ClientListener.scala index e8c4083f9d..064024455e 100644 --- a/core/src/main/scala/spark/deploy/client/ClientListener.scala +++ b/core/src/main/scala/spark/deploy/client/ClientListener.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.client /** diff --git a/core/src/main/scala/spark/deploy/client/TestClient.scala b/core/src/main/scala/spark/deploy/client/TestClient.scala index f195082808..4f4daa141a 100644 --- a/core/src/main/scala/spark/deploy/client/TestClient.scala +++ b/core/src/main/scala/spark/deploy/client/TestClient.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.client import spark.util.AkkaUtils diff --git a/core/src/main/scala/spark/deploy/client/TestExecutor.scala b/core/src/main/scala/spark/deploy/client/TestExecutor.scala index 0e46db2272..8a22b6b89f 100644 --- a/core/src/main/scala/spark/deploy/client/TestExecutor.scala +++ b/core/src/main/scala/spark/deploy/client/TestExecutor.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.client private[spark] object TestExecutor { diff --git a/core/src/main/scala/spark/deploy/master/ApplicationInfo.scala b/core/src/main/scala/spark/deploy/master/ApplicationInfo.scala index 785c16e2be..15ff919738 100644 --- a/core/src/main/scala/spark/deploy/master/ApplicationInfo.scala +++ b/core/src/main/scala/spark/deploy/master/ApplicationInfo.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.master import spark.deploy.ApplicationDescription diff --git a/core/src/main/scala/spark/deploy/master/ApplicationState.scala b/core/src/main/scala/spark/deploy/master/ApplicationState.scala index 15016b388d..94f0ad8bae 100644 --- a/core/src/main/scala/spark/deploy/master/ApplicationState.scala +++ b/core/src/main/scala/spark/deploy/master/ApplicationState.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.master private[spark] object ApplicationState diff --git a/core/src/main/scala/spark/deploy/master/ExecutorInfo.scala b/core/src/main/scala/spark/deploy/master/ExecutorInfo.scala index 48e6055fb5..99b60f7d09 100644 --- a/core/src/main/scala/spark/deploy/master/ExecutorInfo.scala +++ b/core/src/main/scala/spark/deploy/master/ExecutorInfo.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.master import spark.deploy.ExecutorState diff --git a/core/src/main/scala/spark/deploy/master/Master.scala b/core/src/main/scala/spark/deploy/master/Master.scala index dbdc8e1057..202d5bcdb7 100644 --- a/core/src/main/scala/spark/deploy/master/Master.scala +++ b/core/src/main/scala/spark/deploy/master/Master.scala @@ -1,19 +1,38 @@ -package spark.deploy.master +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ -import akka.actor._ -import akka.actor.Terminated -import akka.remote.{RemoteClientLifeCycleEvent, RemoteClientDisconnected, RemoteClientShutdown} -import akka.util.duration._ +package spark.deploy.master import java.text.SimpleDateFormat import java.util.Date import scala.collection.mutable.{ArrayBuffer, HashMap, HashSet} -import spark.deploy._ +import akka.actor._ +import akka.actor.Terminated +import akka.remote.{RemoteClientLifeCycleEvent, RemoteClientDisconnected, RemoteClientShutdown} +import akka.util.duration._ + import spark.{Logging, SparkException, Utils} +import spark.deploy.{ApplicationDescription, ExecutorState} +import spark.deploy.DeployMessages._ +import spark.deploy.master.ui.MasterWebUI +import spark.metrics.MetricsSystem import spark.util.AkkaUtils -import ui.MasterWebUI private[spark] class Master(host: String, port: Int, webUiPort: Int) extends Actor with Logging { @@ -36,10 +55,13 @@ private[spark] class Master(host: String, port: Int, webUiPort: Int) extends Act var firstApp: Option[ApplicationInfo] = None - val webUi = new MasterWebUI(self) + val webUi = new MasterWebUI(self, webUiPort) Utils.checkHost(host, "Expected hostname") + val metricsSystem = MetricsSystem.createMetricsSystem("master") + val masterSource = new MasterSource(this) + val masterPublicAddress = { val envVar = System.getenv("SPARK_PUBLIC_DNS") if (envVar != null) envVar else host @@ -56,10 +78,14 @@ private[spark] class Master(host: String, port: Int, webUiPort: Int) extends Act context.system.eventStream.subscribe(self, classOf[RemoteClientLifeCycleEvent]) webUi.start() context.system.scheduler.schedule(0 millis, WORKER_TIMEOUT millis)(timeOutDeadWorkers()) + + metricsSystem.registerSource(masterSource) + metricsSystem.start() } override def postStop() { webUi.stop() + metricsSystem.stop() } override def receive = { @@ -71,7 +97,7 @@ private[spark] class Master(host: String, port: Int, webUiPort: Int) extends Act } else { addWorker(id, host, workerPort, cores, memory, worker_webUiPort, publicAddress) context.watch(sender) // This doesn't work with remote actors but helps for testing - sender ! RegisteredWorker("http://" + masterPublicAddress + ":" + webUi.boundPort) + sender ! RegisteredWorker("http://" + masterPublicAddress + ":" + webUi.boundPort.get) schedule() } } @@ -143,7 +169,7 @@ private[spark] class Master(host: String, port: Int, webUiPort: Int) extends Act } case RequestMasterState => { - sender ! MasterState(host, port, workers.toArray, apps.toArray, completedApps.toArray) + sender ! MasterStateResponse(host, port, workers.toArray, apps.toArray, completedApps.toArray) } } @@ -208,20 +234,27 @@ private[spark] class Master(host: String, port: Int, webUiPort: Int) extends Act def launchExecutor(worker: WorkerInfo, exec: ExecutorInfo, sparkHome: String) { logInfo("Launching executor " + exec.fullId + " on worker " + worker.id) worker.addExecutor(exec) - worker.actor ! LaunchExecutor(exec.application.id, exec.id, exec.application.desc, exec.cores, exec.memory, sparkHome) - exec.application.driver ! ExecutorAdded(exec.id, worker.id, worker.hostPort, exec.cores, exec.memory) + worker.actor ! LaunchExecutor( + exec.application.id, exec.id, exec.application.desc, exec.cores, exec.memory, sparkHome) + exec.application.driver ! ExecutorAdded( + exec.id, worker.id, worker.hostPort, exec.cores, exec.memory) } def addWorker(id: String, host: String, port: Int, cores: Int, memory: Int, webUiPort: Int, publicAddress: String): WorkerInfo = { - // There may be one or more refs to dead workers on this same node (w/ different ID's), remove them. - workers.filter(w => (w.host == host && w.port == port) && (w.state == WorkerState.DEAD)).foreach(workers -= _) + // There may be one or more refs to dead workers on this same node (w/ different ID's), + // remove them. + workers.filter { w => + (w.host == host && w.port == port) && (w.state == WorkerState.DEAD) + }.foreach { w => + workers -= w + } val worker = new WorkerInfo(id, host, port, cores, memory, sender, webUiPort, publicAddress) workers += worker idToWorker(worker.id) = worker actorToWorker(sender) = worker addressToWorker(sender.path.address) = worker - return worker + worker } def removeWorker(worker: WorkerInfo) { @@ -232,7 +265,8 @@ private[spark] class Master(host: String, port: Int, webUiPort: Int) extends Act addressToWorker -= worker.actor.path.address for (exec <- worker.executors.values) { logInfo("Telling app of lost executor: " + exec.id) - exec.application.driver ! ExecutorUpdated(exec.id, ExecutorState.LOST, Some("worker lost"), None) + exec.application.driver ! ExecutorUpdated( + exec.id, ExecutorState.LOST, Some("worker lost"), None) exec.application.removeExecutor(exec) } } @@ -252,7 +286,7 @@ private[spark] class Master(host: String, port: Int, webUiPort: Int) extends Act if (workersAlive.size > 0 && !workersAlive.exists(_.memoryFree >= desc.memoryPerSlave)) { logWarning("Could not find any workers with enough memory for " + firstApp.get.id) } - return app + app } def finishApplication(app: ApplicationInfo) { diff --git a/core/src/main/scala/spark/deploy/master/MasterArguments.scala b/core/src/main/scala/spark/deploy/master/MasterArguments.scala index 3d28ecabb4..0ae0160767 100644 --- a/core/src/main/scala/spark/deploy/master/MasterArguments.scala +++ b/core/src/main/scala/spark/deploy/master/MasterArguments.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.master import spark.util.IntParam @@ -21,7 +38,10 @@ private[spark] class MasterArguments(args: Array[String]) { if (System.getenv("SPARK_MASTER_WEBUI_PORT") != null) { webUiPort = System.getenv("SPARK_MASTER_WEBUI_PORT").toInt } - + if (System.getProperty("master.ui.port") != null) { + webUiPort = System.getProperty("master.ui.port").toInt + } + parse(args.toList) def parse(args: List[String]): Unit = args match { diff --git a/core/src/main/scala/spark/deploy/master/MasterSource.scala b/core/src/main/scala/spark/deploy/master/MasterSource.scala new file mode 100644 index 0000000000..b8cfa6a773 --- /dev/null +++ b/core/src/main/scala/spark/deploy/master/MasterSource.scala @@ -0,0 +1,25 @@ +package spark.deploy.master + +import com.codahale.metrics.{Gauge, MetricRegistry} + +import spark.metrics.source.Source + +private[spark] class MasterSource(val master: Master) extends Source { + val metricRegistry = new MetricRegistry() + val sourceName = "master" + + // Gauge for worker numbers in cluster + metricRegistry.register(MetricRegistry.name("workers","number"), new Gauge[Int] { + override def getValue: Int = master.workers.size + }) + + // Gauge for application numbers in cluster + metricRegistry.register(MetricRegistry.name("apps", "number"), new Gauge[Int] { + override def getValue: Int = master.apps.size + }) + + // Gauge for waiting application numbers in cluster + metricRegistry.register(MetricRegistry.name("waitingApps", "number"), new Gauge[Int] { + override def getValue: Int = master.waitingApps.size + }) +} diff --git a/core/src/main/scala/spark/deploy/master/WorkerInfo.scala b/core/src/main/scala/spark/deploy/master/WorkerInfo.scala index 0c08c5f417..4135cfeb28 100644 --- a/core/src/main/scala/spark/deploy/master/WorkerInfo.scala +++ b/core/src/main/scala/spark/deploy/master/WorkerInfo.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.master import akka.actor.ActorRef diff --git a/core/src/main/scala/spark/deploy/master/WorkerState.scala b/core/src/main/scala/spark/deploy/master/WorkerState.scala index 0bf35014c8..3e50b7748d 100644 --- a/core/src/main/scala/spark/deploy/master/WorkerState.scala +++ b/core/src/main/scala/spark/deploy/master/WorkerState.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.master private[spark] object WorkerState extends Enumeration("ALIVE", "DEAD", "DECOMMISSIONED") { diff --git a/core/src/main/scala/spark/deploy/master/ui/ApplicationPage.scala b/core/src/main/scala/spark/deploy/master/ui/ApplicationPage.scala index 33a16b5d84..b4c62bc224 100644 --- a/core/src/main/scala/spark/deploy/master/ui/ApplicationPage.scala +++ b/core/src/main/scala/spark/deploy/master/ui/ApplicationPage.scala @@ -1,5 +1,24 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.master.ui +import scala.xml.Node + import akka.dispatch.Await import akka.pattern.ask import akka.util.duration._ @@ -8,9 +27,8 @@ import javax.servlet.http.HttpServletRequest import net.liftweb.json.JsonAST.JValue -import scala.xml.Node - -import spark.deploy.{RequestMasterState, JsonProtocol, MasterState} +import spark.deploy.DeployMessages.{MasterStateResponse, RequestMasterState} +import spark.deploy.JsonProtocol import spark.deploy.master.ExecutorInfo import spark.ui.UIUtils @@ -21,7 +39,7 @@ private[spark] class ApplicationPage(parent: MasterWebUI) { /** Executor details for a particular application */ def renderJson(request: HttpServletRequest): JValue = { val appId = request.getParameter("appId") - val stateFuture = (master ? RequestMasterState)(timeout).mapTo[MasterState] + val stateFuture = (master ? RequestMasterState)(timeout).mapTo[MasterStateResponse] val state = Await.result(stateFuture, 30 seconds) val app = state.activeApps.find(_.id == appId).getOrElse({ state.completedApps.find(_.id == appId).getOrElse(null) @@ -32,7 +50,7 @@ private[spark] class ApplicationPage(parent: MasterWebUI) { /** Executor details for a particular application */ def render(request: HttpServletRequest): Seq[Node] = { val appId = request.getParameter("appId") - val stateFuture = (master ? RequestMasterState)(timeout).mapTo[MasterState] + val stateFuture = (master ? RequestMasterState)(timeout).mapTo[MasterStateResponse] val state = Await.result(stateFuture, 30 seconds) val app = state.activeApps.find(_.id == appId).getOrElse({ state.completedApps.find(_.id == appId).getOrElse(null) @@ -90,9 +108,9 @@ private[spark] class ApplicationPage(parent: MasterWebUI) { <td>{executor.memory}</td> <td>{executor.state}</td> <td> - <a href={"%s/log?appId=%s&executorId=%s&logType=stdout" + <a href={"%s/logPage?appId=%s&executorId=%s&logType=stdout" .format(executor.worker.webUiAddress, executor.application.id, executor.id)}>stdout</a> - <a href={"%s/log?appId=%s&executorId=%s&logType=stderr" + <a href={"%s/logPage?appId=%s&executorId=%s&logType=stderr" .format(executor.worker.webUiAddress, executor.application.id, executor.id)}>stderr</a> </td> </tr> diff --git a/core/src/main/scala/spark/deploy/master/ui/IndexPage.scala b/core/src/main/scala/spark/deploy/master/ui/IndexPage.scala index c6de2bafa3..557df89b41 100644 --- a/core/src/main/scala/spark/deploy/master/ui/IndexPage.scala +++ b/core/src/main/scala/spark/deploy/master/ui/IndexPage.scala @@ -1,17 +1,36 @@ -package spark.deploy.master.ui +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ -import akka.dispatch.Await -import akka.pattern.ask -import akka.util.duration._ +package spark.deploy.master.ui import javax.servlet.http.HttpServletRequest import scala.xml.Node -import spark.deploy.{RequestMasterState, DeployWebUI, MasterState} +import akka.dispatch.Await +import akka.pattern.ask +import akka.util.duration._ + import spark.Utils -import spark.ui.UIUtils +import spark.deploy.DeployWebUI +import spark.deploy.DeployMessages.{MasterStateResponse, RequestMasterState} import spark.deploy.master.{ApplicationInfo, WorkerInfo} +import spark.ui.UIUtils + private[spark] class IndexPage(parent: MasterWebUI) { val master = parent.master @@ -19,7 +38,7 @@ private[spark] class IndexPage(parent: MasterWebUI) { /** Index view listing applications and executors */ def render(request: HttpServletRequest): Seq[Node] = { - val stateFuture = (master ? RequestMasterState)(timeout).mapTo[MasterState] + val stateFuture = (master ? RequestMasterState)(timeout).mapTo[MasterStateResponse] val state = Await.result(stateFuture, 30 seconds) val workerHeaders = Seq("Id", "Address", "State", "Cores", "Memory") @@ -38,8 +57,8 @@ private[spark] class IndexPage(parent: MasterWebUI) { <div class="row"> <div class="span12"> <ul class="unstyled"> - <li><strong>URL:</strong>{state.uri}</li> - <li><strong>Workers:</strong>{state.workers.size}</li> + <li><strong>URL:</strong> {state.uri}</li> + <li><strong>Workers:</strong> {state.workers.size}</li> <li><strong>Cores:</strong> {state.workers.map(_.cores).sum} Total, {state.workers.map(_.coresUsed).sum} Used</li> <li><strong>Memory:</strong> @@ -103,7 +122,9 @@ private[spark] class IndexPage(parent: MasterWebUI) { <td> <a href={"app?appId=" + app.id}>{app.id}</a> </td> - <td>{app.desc.name}</td> + <td> + <a href={app.appUiUrl}>{app.desc.name}</a> + </td> <td> {app.coresGranted} </td> diff --git a/core/src/main/scala/spark/deploy/master/ui/MasterWebUI.scala b/core/src/main/scala/spark/deploy/master/ui/MasterWebUI.scala index 82bc6961e2..dabc2d8dc7 100644 --- a/core/src/main/scala/spark/deploy/master/ui/MasterWebUI.scala +++ b/core/src/main/scala/spark/deploy/master/ui/MasterWebUI.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.master.ui import akka.actor.ActorRef @@ -15,12 +32,11 @@ import spark.ui.JettyUtils._ * Web UI server for the standalone master. */ private[spark] -class MasterWebUI(val master: ActorRef, requestedPort: Option[Int] = None) extends Logging { +class MasterWebUI(val master: ActorRef, requestedPort: Int) extends Logging { implicit val timeout = Duration.create( System.getProperty("spark.akka.askTimeout", "10").toLong, "seconds") val host = Utils.localHostName() - val port = requestedPort.getOrElse( - System.getProperty("master.ui.port", MasterWebUI.DEFAULT_PORT).toInt) + val port = requestedPort var server: Option[Server] = None var boundPort: Option[Int] = None @@ -55,5 +71,4 @@ class MasterWebUI(val master: ActorRef, requestedPort: Option[Int] = None) exten private[spark] object MasterWebUI { val STATIC_RESOURCE_DIR = "spark/ui/static" - val DEFAULT_PORT = "8080" -}
\ No newline at end of file +} diff --git a/core/src/main/scala/spark/deploy/worker/ExecutorRunner.scala b/core/src/main/scala/spark/deploy/worker/ExecutorRunner.scala index d7f58b2cb1..345dfe879c 100644 --- a/core/src/main/scala/spark/deploy/worker/ExecutorRunner.scala +++ b/core/src/main/scala/spark/deploy/worker/ExecutorRunner.scala @@ -1,15 +1,30 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.worker import java.io._ import java.lang.System.getenv -import spark.deploy.{ExecutorState, ExecutorStateChanged, ApplicationDescription} + import akka.actor.ActorRef + import spark.{Utils, Logging} -import java.net.{URI, URL} -import org.apache.hadoop.fs.{Path, FileSystem} -import org.apache.hadoop.conf.Configuration -import scala.Some -import spark.deploy.ExecutorStateChanged +import spark.deploy.{ExecutorState, ApplicationDescription} +import spark.deploy.DeployMessages.ExecutorStateChanged /** * Manages the execution of one executor process. @@ -34,6 +49,9 @@ private[spark] class ExecutorRunner( var process: Process = null var shutdownHook: Thread = null + private def getAppEnv(key: String): Option[String] = + appDesc.command.environment.get(key).orElse(Option(getenv(key))) + def start() { workerThread = new Thread("ExecutorRunner for " + fullId) { override def run() { fetchAndRunExecutor() } @@ -78,7 +96,7 @@ private[spark] class ExecutorRunner( def buildCommandSeq(): Seq[String] = { val command = appDesc.command - val runner = Option(getenv("JAVA_HOME")).map(_ + "/bin/java").getOrElse("java") + val runner = getAppEnv("JAVA_HOME").map(_ + "/bin/java").getOrElse("java") // SPARK-698: do not call the run.cmd script, as process.destroy() // fails to kill a process tree on Windows Seq(runner) ++ buildJavaOpts() ++ Seq(command.mainClass) ++ @@ -90,15 +108,17 @@ private[spark] class ExecutorRunner( * the way the JAVA_OPTS are assembled there. */ def buildJavaOpts(): Seq[String] = { - val libraryOpts = Option(getenv("SPARK_LIBRARY_PATH")) + val libraryOpts = getAppEnv("SPARK_LIBRARY_PATH") .map(p => List("-Djava.library.path=" + p)) .getOrElse(Nil) - val userOpts = Option(getenv("SPARK_JAVA_OPTS")).map(Utils.splitCommandString).getOrElse(Nil) + val userOpts = getAppEnv("SPARK_JAVA_OPTS").map(Utils.splitCommandString).getOrElse(Nil) val memoryOpts = Seq("-Xms" + memory + "M", "-Xmx" + memory + "M") // Figure out our classpath with the external compute-classpath script val ext = if (System.getProperty("os.name").startsWith("Windows")) ".cmd" else ".sh" - val classPath = Utils.executeAndGetOutput(Seq(sparkHome + "/bin/compute-classpath" + ext)) + val classPath = Utils.executeAndGetOutput( + Seq(sparkHome + "/bin/compute-classpath" + ext), + extraEnvironment=appDesc.command.environment) Seq("-cp", classPath) ++ libraryOpts ++ userOpts ++ memoryOpts } diff --git a/core/src/main/scala/spark/deploy/worker/Worker.scala b/core/src/main/scala/spark/deploy/worker/Worker.scala index 6ae1cef940..0e46fa281e 100644 --- a/core/src/main/scala/spark/deploy/worker/Worker.scala +++ b/core/src/main/scala/spark/deploy/worker/Worker.scala @@ -1,20 +1,40 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.worker -import scala.collection.mutable.{ArrayBuffer, HashMap} +import java.text.SimpleDateFormat +import java.util.Date +import java.io.File + +import scala.collection.mutable.HashMap + import akka.actor.{ActorRef, Props, Actor, ActorSystem, Terminated} +import akka.remote.{RemoteClientLifeCycleEvent, RemoteClientShutdown, RemoteClientDisconnected} import akka.util.duration._ + import spark.{Logging, Utils} -import spark.util.AkkaUtils -import spark.deploy._ -import akka.remote.{RemoteClientLifeCycleEvent, RemoteClientShutdown, RemoteClientDisconnected} -import java.text.SimpleDateFormat -import java.util.Date -import spark.deploy.RegisterWorker -import spark.deploy.LaunchExecutor -import spark.deploy.RegisterWorkerFailed +import spark.deploy.ExecutorState +import spark.deploy.DeployMessages._ import spark.deploy.master.Master -import java.io.File -import ui.WorkerWebUI +import spark.deploy.worker.ui.WorkerWebUI +import spark.metrics.MetricsSystem +import spark.util.AkkaUtils + private[spark] class Worker( host: String, @@ -50,6 +70,9 @@ private[spark] class Worker( var coresUsed = 0 var memoryUsed = 0 + val metricsSystem = MetricsSystem.createMetricsSystem("worker") + val workerSource = new WorkerSource(this) + def coresFree: Int = cores - coresUsed def memoryFree: Int = memory - memoryUsed @@ -77,9 +100,12 @@ private[spark] class Worker( sparkHome = new File(Option(System.getenv("SPARK_HOME")).getOrElse(".")) logInfo("Spark home: " + sparkHome) createWorkDir() - webUi = new WorkerWebUI(self, workDir, Some(webUiPort)) + webUi = new WorkerWebUI(this, workDir, Some(webUiPort)) webUi.start() connectToMaster() + + metricsSystem.registerSource(workerSource) + metricsSystem.start() } def connectToMaster() { @@ -138,10 +164,10 @@ private[spark] class Worker( case Terminated(_) | RemoteClientDisconnected(_, _) | RemoteClientShutdown(_, _) => masterDisconnected() - + case RequestWorkerState => { - sender ! WorkerState(host, port, workerId, executors.values.toList, - finishedExecutors.values.toList, masterUrl, cores, memory, + sender ! WorkerStateResponse(host, port, workerId, executors.values.toList, + finishedExecutors.values.toList, masterUrl, cores, memory, coresUsed, memoryUsed, masterWebUiUrl) } } @@ -161,6 +187,7 @@ private[spark] class Worker( override def postStop() { executors.values.foreach(_.kill()) webUi.stop() + metricsSystem.stop() } } diff --git a/core/src/main/scala/spark/deploy/worker/WorkerArguments.scala b/core/src/main/scala/spark/deploy/worker/WorkerArguments.scala index 2b96611ee3..9fcd3260ca 100644 --- a/core/src/main/scala/spark/deploy/worker/WorkerArguments.scala +++ b/core/src/main/scala/spark/deploy/worker/WorkerArguments.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.worker import spark.util.IntParam diff --git a/core/src/main/scala/spark/deploy/worker/WorkerSource.scala b/core/src/main/scala/spark/deploy/worker/WorkerSource.scala new file mode 100644 index 0000000000..39cb8e5690 --- /dev/null +++ b/core/src/main/scala/spark/deploy/worker/WorkerSource.scala @@ -0,0 +1,34 @@ +package spark.deploy.worker + +import com.codahale.metrics.{Gauge, MetricRegistry} + +import spark.metrics.source.Source + +private[spark] class WorkerSource(val worker: Worker) extends Source { + val sourceName = "worker" + val metricRegistry = new MetricRegistry() + + metricRegistry.register(MetricRegistry.name("executors", "number"), new Gauge[Int] { + override def getValue: Int = worker.executors.size + }) + + // Gauge for cores used of this worker + metricRegistry.register(MetricRegistry.name("coresUsed", "number"), new Gauge[Int] { + override def getValue: Int = worker.coresUsed + }) + + // Gauge for memory used of this worker + metricRegistry.register(MetricRegistry.name("memUsed", "MBytes"), new Gauge[Int] { + override def getValue: Int = worker.memoryUsed + }) + + // Gauge for cores free of this worker + metricRegistry.register(MetricRegistry.name("coresFree", "number"), new Gauge[Int] { + override def getValue: Int = worker.coresFree + }) + + // Gauge for memory free of this worker + metricRegistry.register(MetricRegistry.name("memFree", "MBytes"), new Gauge[Int] { + override def getValue: Int = worker.memoryFree + }) +} diff --git a/core/src/main/scala/spark/deploy/worker/ui/IndexPage.scala b/core/src/main/scala/spark/deploy/worker/ui/IndexPage.scala index e466129c1a..1619c6a4c2 100644 --- a/core/src/main/scala/spark/deploy/worker/ui/IndexPage.scala +++ b/core/src/main/scala/spark/deploy/worker/ui/IndexPage.scala @@ -1,32 +1,52 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.worker.ui +import javax.servlet.http.HttpServletRequest + +import scala.xml.Node + import akka.dispatch.Await import akka.pattern.ask import akka.util.duration._ -import javax.servlet.http.HttpServletRequest - import net.liftweb.json.JsonAST.JValue -import scala.xml.Node - -import spark.deploy.{RequestWorkerState, JsonProtocol, WorkerState} -import spark.deploy.worker.ExecutorRunner import spark.Utils +import spark.deploy.JsonProtocol +import spark.deploy.DeployMessages.{RequestWorkerState, WorkerStateResponse} +import spark.deploy.worker.ExecutorRunner import spark.ui.UIUtils + private[spark] class IndexPage(parent: WorkerWebUI) { + val workerActor = parent.worker.self val worker = parent.worker val timeout = parent.timeout def renderJson(request: HttpServletRequest): JValue = { - val stateFuture = (worker ? RequestWorkerState)(timeout).mapTo[WorkerState] + val stateFuture = (workerActor ? RequestWorkerState)(timeout).mapTo[WorkerStateResponse] val workerState = Await.result(stateFuture, 30 seconds) JsonProtocol.writeWorkerState(workerState) } def render(request: HttpServletRequest): Seq[Node] = { - val stateFuture = (worker ? RequestWorkerState)(timeout).mapTo[WorkerState] + val stateFuture = (workerActor ? RequestWorkerState)(timeout).mapTo[WorkerStateResponse] val workerState = Await.result(stateFuture, 30 seconds) val executorHeaders = Seq("ExecutorID", "Cores", "Memory", "Job Details", "Logs") @@ -51,7 +71,7 @@ private[spark] class IndexPage(parent: WorkerWebUI) { <p><a href={workerState.masterWebUiUrl}>Back to Master</a></p> </div> </div> - <hr/> + <hr/> <div class="row"> <!-- Running Executors --> <div class="span12"> @@ -70,7 +90,8 @@ private[spark] class IndexPage(parent: WorkerWebUI) { </div> </div>; - UIUtils.basicSparkPage(content, "Spark Worker on %s:%s".format(workerState.host, workerState.port)) + UIUtils.basicSparkPage(content, "Spark Worker on %s:%s".format( + workerState.host, workerState.port)) } def executorRow(executor: ExecutorRunner): Seq[Node] = { @@ -88,11 +109,11 @@ private[spark] class IndexPage(parent: WorkerWebUI) { </ul> </td> <td> - <a href={"log?appId=%s&executorId=%s&logType=stdout" + <a href={"logPage?appId=%s&executorId=%s&logType=stdout" .format(executor.appId, executor.execId)}>stdout</a> - <a href={"log?appId=%s&executorId=%s&logType=stderr" + <a href={"logPage?appId=%s&executorId=%s&logType=stderr" .format(executor.appId, executor.execId)}>stderr</a> - </td> + </td> </tr> } diff --git a/core/src/main/scala/spark/deploy/worker/ui/WorkerWebUI.scala b/core/src/main/scala/spark/deploy/worker/ui/WorkerWebUI.scala index 16564d5619..61d4cd6d99 100644 --- a/core/src/main/scala/spark/deploy/worker/ui/WorkerWebUI.scala +++ b/core/src/main/scala/spark/deploy/worker/ui/WorkerWebUI.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.deploy.worker.ui import akka.actor.ActorRef @@ -9,15 +26,17 @@ import javax.servlet.http.HttpServletRequest import org.eclipse.jetty.server.{Handler, Server} +import spark.deploy.worker.Worker import spark.{Utils, Logging} import spark.ui.JettyUtils import spark.ui.JettyUtils._ +import spark.ui.UIUtils /** * Web UI server for the standalone worker. */ private[spark] -class WorkerWebUI(val worker: ActorRef, val workDir: File, requestedPort: Option[Int] = None) +class WorkerWebUI(val worker: Worker, val workDir: File, requestedPort: Option[Int] = None) extends Logging { implicit val timeout = Timeout( Duration.create(System.getProperty("spark.akka.askTimeout", "10").toLong, "seconds")) @@ -33,6 +52,7 @@ class WorkerWebUI(val worker: ActorRef, val workDir: File, requestedPort: Option val handlers = Array[(String, Handler)]( ("/static", createStaticHandler(WorkerWebUI.STATIC_RESOURCE_DIR)), ("/log", (request: HttpServletRequest) => log(request)), + ("/logPage", (request: HttpServletRequest) => logPage(request)), ("/json", (request: HttpServletRequest) => indexPage.renderJson(request)), ("*", (request: HttpServletRequest) => indexPage.render(request)) ) @@ -51,18 +71,104 @@ class WorkerWebUI(val worker: ActorRef, val workDir: File, requestedPort: Option } def log(request: HttpServletRequest): String = { + val defaultBytes = 100 * 1024 val appId = request.getParameter("appId") val executorId = request.getParameter("executorId") val logType = request.getParameter("logType") + val offset = Option(request.getParameter("offset")).map(_.toLong) + val byteLength = Option(request.getParameter("byteLength")).map(_.toInt).getOrElse(defaultBytes) + val path = "%s/%s/%s/%s".format(workDir.getPath, appId, executorId, logType) - val maxBytes = 1024 * 1024 // Guard against OOM - val defaultBytes = 100 * 1024 - val numBytes = Option(request.getParameter("numBytes")) - .flatMap(s => Some(s.toInt)).getOrElse(defaultBytes) + val (startByte, endByte) = getByteRange(path, offset, byteLength) + val file = new File(path) + val logLength = file.length + + val pre = "==== Bytes %s-%s of %s of %s/%s/%s ====\n" + .format(startByte, endByte, logLength, appId, executorId, logType) + pre + Utils.offsetBytes(path, startByte, endByte) + } + def logPage(request: HttpServletRequest): Seq[scala.xml.Node] = { + val defaultBytes = 100 * 1024 + val appId = request.getParameter("appId") + val executorId = request.getParameter("executorId") + val logType = request.getParameter("logType") + val offset = Option(request.getParameter("offset")).map(_.toLong) + val byteLength = Option(request.getParameter("byteLength")).map(_.toInt).getOrElse(defaultBytes) val path = "%s/%s/%s/%s".format(workDir.getPath, appId, executorId, logType) - val pre = "==== Last %s bytes of %s/%s/%s ====\n".format(numBytes, appId, executorId, logType) - pre + Utils.lastNBytes(path, math.min(numBytes, maxBytes)) + + val (startByte, endByte) = getByteRange(path, offset, byteLength) + val file = new File(path) + val logLength = file.length + + val logText = <node>{Utils.offsetBytes(path, startByte, endByte)}</node> + + val linkToMaster = <p><a href={worker.masterWebUiUrl}>Back to Master</a></p> + + val range = <span>Bytes {startByte.toString} - {endByte.toString} of {logLength}</span> + + val backButton = + if (startByte > 0) { + <a href={"?appId=%s&executorId=%s&logType=%s&offset=%s&byteLength=%s" + .format(appId, executorId, logType, math.max(startByte-byteLength, 0), + byteLength)}> + <button>Previous {Utils.memoryBytesToString(math.min(byteLength, startByte))}</button> + </a> + } + else { + <button disabled="disabled">Previous 0 B</button> + } + + val nextButton = + if (endByte < logLength) { + <a href={"?appId=%s&executorId=%s&logType=%s&offset=%s&byteLength=%s". + format(appId, executorId, logType, endByte, byteLength)}> + <button>Next {Utils.memoryBytesToString(math.min(byteLength, logLength-endByte))}</button> + </a> + } + else { + <button disabled="disabled">Next 0 B</button> + } + + val content = + <html> + <body> + {linkToMaster} + <hr /> + <div> + <div style="float:left;width:40%">{backButton}</div> + <div style="float:left;">{range}</div> + <div style="float:right;">{nextButton}</div> + </div> + <br /> + <div style="height:500px;overflow:auto;padding:5px;"> + <pre>{logText}</pre> + </div> + </body> + </html> + UIUtils.basicSparkPage(content, logType + " log page for " + appId) + } + + /** Determine the byte range for a log or log page. */ + def getByteRange(path: String, offset: Option[Long], byteLength: Int) + : (Long, Long) = { + val defaultBytes = 100 * 1024 + val maxBytes = 1024 * 1024 + + val file = new File(path) + val logLength = file.length() + val getOffset = offset.getOrElse(logLength-defaultBytes) + + val startByte = + if (getOffset < 0) 0L + else if (getOffset > logLength) logLength + else getOffset + + val logPageLength = math.min(byteLength, maxBytes) + + val endByte = math.min(startByte+logPageLength, logLength) + + (startByte, endByte) } def stop() { diff --git a/core/src/main/scala/spark/executor/Executor.scala b/core/src/main/scala/spark/executor/Executor.scala index 8360547a74..8a74a8d853 100644 --- a/core/src/main/scala/spark/executor/Executor.scala +++ b/core/src/main/scala/spark/executor/Executor.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.executor import java.io.{File, FileOutputStream} @@ -52,7 +69,7 @@ private[spark] class Executor(executorId: String, slaveHostname: String, propert override def uncaughtException(thread: Thread, exception: Throwable) { try { logError("Uncaught exception in thread " + thread, exception) - + // We may have been called from a shutdown hook. If so, we must not call System.exit(). // (If we do, we will deadlock.) if (!Utils.inShutdown()) { @@ -70,9 +87,13 @@ private[spark] class Executor(executorId: String, slaveHostname: String, propert } ) + val executorSource = new ExecutorSource(this) + // Initialize Spark environment (using system properties read above) val env = SparkEnv.createFromSystemProperties(executorId, slaveHostname, 0, false, false) SparkEnv.set(env) + env.metricsSystem.registerSource(executorSource) + private val akkaFrameSize = env.actorSystem.settings.config.getBytes("akka.remote.netty.message-frame-size") // Start worker thread pool diff --git a/core/src/main/scala/spark/executor/ExecutorBackend.scala b/core/src/main/scala/spark/executor/ExecutorBackend.scala index e97e509700..33a6f8a824 100644 --- a/core/src/main/scala/spark/executor/ExecutorBackend.scala +++ b/core/src/main/scala/spark/executor/ExecutorBackend.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.executor import java.nio.ByteBuffer diff --git a/core/src/main/scala/spark/executor/ExecutorExitCode.scala b/core/src/main/scala/spark/executor/ExecutorExitCode.scala index fd76029cb3..64b9fb88f8 100644 --- a/core/src/main/scala/spark/executor/ExecutorExitCode.scala +++ b/core/src/main/scala/spark/executor/ExecutorExitCode.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.executor /** diff --git a/core/src/main/scala/spark/executor/ExecutorSource.scala b/core/src/main/scala/spark/executor/ExecutorSource.scala new file mode 100644 index 0000000000..94116edfcf --- /dev/null +++ b/core/src/main/scala/spark/executor/ExecutorSource.scala @@ -0,0 +1,30 @@ +package spark.executor + +import com.codahale.metrics.{Gauge, MetricRegistry} + +import spark.metrics.source.Source + +class ExecutorSource(val executor: Executor) extends Source { + val metricRegistry = new MetricRegistry() + val sourceName = "executor" + + // Gauge for executor thread pool's actively executing task counts + metricRegistry.register(MetricRegistry.name("threadpool", "activeTask", "count"), new Gauge[Int] { + override def getValue: Int = executor.threadPool.getActiveCount() + }) + + // Gauge for executor thread pool's approximate total number of tasks that have been completed + metricRegistry.register(MetricRegistry.name("threadpool", "completeTask", "count"), new Gauge[Long] { + override def getValue: Long = executor.threadPool.getCompletedTaskCount() + }) + + // Gauge for executor thread pool's current number of threads + metricRegistry.register(MetricRegistry.name("threadpool", "currentPool", "size"), new Gauge[Int] { + override def getValue: Int = executor.threadPool.getPoolSize() + }) + + // Gauge got executor thread pool's largest number of threads that have ever simultaneously been in th pool + metricRegistry.register(MetricRegistry.name("threadpool", "maxPool", "size"), new Gauge[Int] { + override def getValue: Int = executor.threadPool.getMaximumPoolSize() + }) +} diff --git a/core/src/main/scala/spark/executor/ExecutorURLClassLoader.scala b/core/src/main/scala/spark/executor/ExecutorURLClassLoader.scala index 5beb4d049e..09d12fb65b 100644 --- a/core/src/main/scala/spark/executor/ExecutorURLClassLoader.scala +++ b/core/src/main/scala/spark/executor/ExecutorURLClassLoader.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.executor import java.net.{URLClassLoader, URL} diff --git a/core/src/main/scala/spark/executor/MesosExecutorBackend.scala b/core/src/main/scala/spark/executor/MesosExecutorBackend.scala index 10f3531df0..4961c42fad 100644 --- a/core/src/main/scala/spark/executor/MesosExecutorBackend.scala +++ b/core/src/main/scala/spark/executor/MesosExecutorBackend.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.executor import java.nio.ByteBuffer diff --git a/core/src/main/scala/spark/executor/StandaloneExecutorBackend.scala b/core/src/main/scala/spark/executor/StandaloneExecutorBackend.scala index ebe2ac68d8..e47fe50021 100644 --- a/core/src/main/scala/spark/executor/StandaloneExecutorBackend.scala +++ b/core/src/main/scala/spark/executor/StandaloneExecutorBackend.scala @@ -1,19 +1,33 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.executor import java.nio.ByteBuffer -import spark.Logging -import spark.TaskState.TaskState -import spark.util.AkkaUtils + import akka.actor.{ActorRef, Actor, Props, Terminated} import akka.remote.{RemoteClientLifeCycleEvent, RemoteClientShutdown, RemoteClientDisconnected} -import java.util.concurrent.{TimeUnit, ThreadPoolExecutor, SynchronousQueue} -import spark.scheduler.cluster._ -import spark.scheduler.cluster.RegisteredExecutor -import spark.scheduler.cluster.LaunchTask -import spark.scheduler.cluster.RegisterExecutorFailed -import spark.scheduler.cluster.RegisterExecutor -import spark.Utils + +import spark.{Logging, Utils} +import spark.TaskState.TaskState import spark.deploy.SparkHadoopUtil +import spark.scheduler.cluster.StandaloneClusterMessages._ +import spark.util.AkkaUtils + private[spark] class StandaloneExecutorBackend( driverUrl: String, diff --git a/core/src/main/scala/spark/executor/TaskMetrics.scala b/core/src/main/scala/spark/executor/TaskMetrics.scala index 1dc13754f9..3151627839 100644 --- a/core/src/main/scala/spark/executor/TaskMetrics.scala +++ b/core/src/main/scala/spark/executor/TaskMetrics.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.executor class TaskMetrics extends Serializable { diff --git a/core/src/main/scala/spark/metrics/MetricsConfig.scala b/core/src/main/scala/spark/metrics/MetricsConfig.scala new file mode 100644 index 0000000000..ed505b0aa7 --- /dev/null +++ b/core/src/main/scala/spark/metrics/MetricsConfig.scala @@ -0,0 +1,79 @@ +package spark.metrics + +import java.util.Properties +import java.io.{File, FileInputStream, InputStream, IOException} + +import scala.collection.mutable +import scala.util.matching.Regex + +import spark.Logging + +private[spark] class MetricsConfig(val configFile: Option[String]) extends Logging { + initLogging() + + val DEFAULT_PREFIX = "*" + val INSTANCE_REGEX = "^(\\*|[a-zA-Z]+)\\.(.+)".r + val METRICS_CONF = "metrics.properties" + + val properties = new Properties() + var propertyCategories: mutable.HashMap[String, Properties] = null + + private def setDefaultProperties(prop: Properties) { + // empty function, any default property can be set here + } + + def initialize() { + //Add default properties in case there's no properties file + setDefaultProperties(properties) + + // If spark.metrics.conf is not set, try to get file in class path + var is: InputStream = null + try { + is = configFile match { + case Some(f) => new FileInputStream(f) + case None => getClass.getClassLoader.getResourceAsStream(METRICS_CONF) + } + + if (is != null) { + properties.load(is) + } + } catch { + case e: Exception => logError("Error loading configure file", e) + } finally { + if (is != null) is.close() + } + + propertyCategories = subProperties(properties, INSTANCE_REGEX) + if (propertyCategories.contains(DEFAULT_PREFIX)) { + import scala.collection.JavaConversions._ + + val defaultProperty = propertyCategories(DEFAULT_PREFIX) + for { (inst, prop) <- propertyCategories + if (inst != DEFAULT_PREFIX) + (k, v) <- defaultProperty + if (prop.getProperty(k) == null) } { + prop.setProperty(k, v) + } + } + } + + def subProperties(prop: Properties, regex: Regex): mutable.HashMap[String, Properties] = { + val subProperties = new mutable.HashMap[String, Properties] + import scala.collection.JavaConversions._ + prop.foreach { kv => + if (regex.findPrefixOf(kv._1) != None) { + val regex(prefix, suffix) = kv._1 + subProperties.getOrElseUpdate(prefix, new Properties).setProperty(suffix, kv._2) + } + } + subProperties + } + + def getInstance(inst: String): Properties = { + propertyCategories.get(inst) match { + case Some(s) => s + case None => propertyCategories.getOrElse(DEFAULT_PREFIX, new Properties) + } + } +} + diff --git a/core/src/main/scala/spark/metrics/MetricsSystem.scala b/core/src/main/scala/spark/metrics/MetricsSystem.scala new file mode 100644 index 0000000000..2f87577ff3 --- /dev/null +++ b/core/src/main/scala/spark/metrics/MetricsSystem.scala @@ -0,0 +1,129 @@ +package spark.metrics + +import com.codahale.metrics.{JmxReporter, MetricSet, MetricRegistry} + +import java.util.Properties +import java.util.concurrent.TimeUnit + +import scala.collection.mutable + +import spark.Logging +import spark.metrics.sink.Sink +import spark.metrics.source.Source + +/** + * Spark Metrics System, created by specific "instance", combined by source, + * sink, periodically poll source metrics data to sink destinations. + * + * "instance" specify "who" (the role) use metrics system. In spark there are several roles + * like master, worker, executor, client driver, these roles will create metrics system + * for monitoring. So instance represents these roles. Currently in Spark, several instances + * have already implemented: master, worker, executor, driver. + * + * "source" specify "where" (source) to collect metrics data. In metrics system, there exists + * two kinds of source: + * 1. Spark internal source, like MasterSource, WorkerSource, etc, which will collect + * Spark component's internal state, these sources are related to instance and will be + * added after specific metrics system is created. + * 2. Common source, like JvmSource, which will collect low level state, is configured by + * configuration and loaded through reflection. + * + * "sink" specify "where" (destination) to output metrics data to. Several sinks can be + * coexisted and flush metrics to all these sinks. + * + * Metrics configuration format is like below: + * [instance].[sink|source].[name].[options] = xxxx + * + * [instance] can be "master", "worker", "executor", "driver", which means only the specified + * instance has this property. + * wild card "*" can be used to replace instance name, which means all the instances will have + * this property. + * + * [sink|source] means this property belongs to source or sink. This field can only be source or sink. + * + * [name] specify the name of sink or source, it is custom defined. + * + * [options] is the specific property of this source or sink. + */ +private[spark] class MetricsSystem private (val instance: String) extends Logging { + initLogging() + + val confFile = System.getProperty("spark.metrics.conf") + val metricsConfig = new MetricsConfig(Option(confFile)) + + val sinks = new mutable.ArrayBuffer[Sink] + val sources = new mutable.ArrayBuffer[Source] + val registry = new MetricRegistry() + + metricsConfig.initialize() + registerSources() + registerSinks() + + def start() { + sinks.foreach(_.start) + } + + def stop() { + sinks.foreach(_.stop) + } + + def registerSource(source: Source) { + sources += source + try { + registry.register(source.sourceName, source.metricRegistry) + } catch { + case e: IllegalArgumentException => logInfo("Metrics already registered", e) + } + } + + def registerSources() { + val instConfig = metricsConfig.getInstance(instance) + val sourceConfigs = metricsConfig.subProperties(instConfig, MetricsSystem.SOURCE_REGEX) + + // Register all the sources related to instance + sourceConfigs.foreach { kv => + val classPath = kv._2.getProperty("class") + try { + val source = Class.forName(classPath).newInstance() + registerSource(source.asInstanceOf[Source]) + } catch { + case e: Exception => logError("Source class " + classPath + " cannot be instantialized", e) + } + } + } + + def registerSinks() { + val instConfig = metricsConfig.getInstance(instance) + val sinkConfigs = metricsConfig.subProperties(instConfig, MetricsSystem.SINK_REGEX) + + sinkConfigs.foreach { kv => + val classPath = kv._2.getProperty("class") + try { + val sink = Class.forName(classPath) + .getConstructor(classOf[Properties], classOf[MetricRegistry]) + .newInstance(kv._2, registry) + sinks += sink.asInstanceOf[Sink] + } catch { + case e: Exception => logError("Sink class " + classPath + " cannot be instantialized", e) + } + } + } +} + +private[spark] object MetricsSystem { + val SINK_REGEX = "^sink\\.(.+)\\.(.+)".r + val SOURCE_REGEX = "^source\\.(.+)\\.(.+)".r + + val MINIMAL_POLL_UNIT = TimeUnit.SECONDS + val MINIMAL_POLL_PERIOD = 1 + + def checkMinimalPollingPeriod(pollUnit: TimeUnit, pollPeriod: Int) { + val period = MINIMAL_POLL_UNIT.convert(pollPeriod, pollUnit) + if (period < MINIMAL_POLL_PERIOD) { + throw new IllegalArgumentException("Polling period " + pollPeriod + " " + pollUnit + + " below than minimal polling period ") + } + } + + def createMetricsSystem(instance: String): MetricsSystem = new MetricsSystem(instance) +} diff --git a/core/src/main/scala/spark/metrics/sink/ConsoleSink.scala b/core/src/main/scala/spark/metrics/sink/ConsoleSink.scala new file mode 100644 index 0000000000..eaaac5d153 --- /dev/null +++ b/core/src/main/scala/spark/metrics/sink/ConsoleSink.scala @@ -0,0 +1,42 @@ +package spark.metrics.sink + +import com.codahale.metrics.{ConsoleReporter, MetricRegistry} + +import java.util.Properties +import java.util.concurrent.TimeUnit + +import spark.metrics.MetricsSystem + +class ConsoleSink(val property: Properties, val registry: MetricRegistry) extends Sink { + val CONSOLE_DEFAULT_PERIOD = 10 + val CONSOLE_DEFAULT_UNIT = "SECONDS" + + val CONSOLE_KEY_PERIOD = "period" + val CONSOLE_KEY_UNIT = "unit" + + val pollPeriod = Option(property.getProperty(CONSOLE_KEY_PERIOD)) match { + case Some(s) => s.toInt + case None => CONSOLE_DEFAULT_PERIOD + } + + val pollUnit = Option(property.getProperty(CONSOLE_KEY_UNIT)) match { + case Some(s) => TimeUnit.valueOf(s.toUpperCase()) + case None => TimeUnit.valueOf(CONSOLE_DEFAULT_UNIT) + } + + MetricsSystem.checkMinimalPollingPeriod(pollUnit, pollPeriod) + + val reporter: ConsoleReporter = ConsoleReporter.forRegistry(registry) + .convertDurationsTo(TimeUnit.MILLISECONDS) + .convertRatesTo(TimeUnit.SECONDS) + .build() + + override def start() { + reporter.start(pollPeriod, pollUnit) + } + + override def stop() { + reporter.stop() + } +} + diff --git a/core/src/main/scala/spark/metrics/sink/CsvSink.scala b/core/src/main/scala/spark/metrics/sink/CsvSink.scala new file mode 100644 index 0000000000..aa5bff0d34 --- /dev/null +++ b/core/src/main/scala/spark/metrics/sink/CsvSink.scala @@ -0,0 +1,51 @@ +package spark.metrics.sink + +import com.codahale.metrics.{CsvReporter, MetricRegistry} + +import java.io.File +import java.util.{Locale, Properties} +import java.util.concurrent.TimeUnit + +import spark.metrics.MetricsSystem + +class CsvSink(val property: Properties, val registry: MetricRegistry) extends Sink { + val CSV_KEY_PERIOD = "period" + val CSV_KEY_UNIT = "unit" + val CSV_KEY_DIR = "directory" + + val CSV_DEFAULT_PERIOD = 10 + val CSV_DEFAULT_UNIT = "SECONDS" + val CSV_DEFAULT_DIR = "/tmp/" + + val pollPeriod = Option(property.getProperty(CSV_KEY_PERIOD)) match { + case Some(s) => s.toInt + case None => CSV_DEFAULT_PERIOD + } + + val pollUnit = Option(property.getProperty(CSV_KEY_UNIT)) match { + case Some(s) => TimeUnit.valueOf(s.toUpperCase()) + case None => TimeUnit.valueOf(CSV_DEFAULT_UNIT) + } + + MetricsSystem.checkMinimalPollingPeriod(pollUnit, pollPeriod) + + val pollDir = Option(property.getProperty(CSV_KEY_DIR)) match { + case Some(s) => s + case None => CSV_DEFAULT_DIR + } + + val reporter: CsvReporter = CsvReporter.forRegistry(registry) + .formatFor(Locale.US) + .convertDurationsTo(TimeUnit.MILLISECONDS) + .convertRatesTo(TimeUnit.SECONDS) + .build(new File(pollDir)) + + override def start() { + reporter.start(pollPeriod, pollUnit) + } + + override def stop() { + reporter.stop() + } +} + diff --git a/core/src/main/scala/spark/metrics/sink/JmxSink.scala b/core/src/main/scala/spark/metrics/sink/JmxSink.scala new file mode 100644 index 0000000000..6a40885b78 --- /dev/null +++ b/core/src/main/scala/spark/metrics/sink/JmxSink.scala @@ -0,0 +1,18 @@ +package spark.metrics.sink + +import com.codahale.metrics.{JmxReporter, MetricRegistry} + +import java.util.Properties + +class JmxSink(val property: Properties, val registry: MetricRegistry) extends Sink { + val reporter: JmxReporter = JmxReporter.forRegistry(registry).build() + + override def start() { + reporter.start() + } + + override def stop() { + reporter.stop() + } + +} diff --git a/core/src/main/scala/spark/metrics/sink/Sink.scala b/core/src/main/scala/spark/metrics/sink/Sink.scala new file mode 100644 index 0000000000..3ffdcbdaba --- /dev/null +++ b/core/src/main/scala/spark/metrics/sink/Sink.scala @@ -0,0 +1,6 @@ +package spark.metrics.sink + +trait Sink { + def start: Unit + def stop: Unit +}
\ No newline at end of file diff --git a/core/src/main/scala/spark/metrics/source/JvmSource.scala b/core/src/main/scala/spark/metrics/source/JvmSource.scala new file mode 100644 index 0000000000..79f505079c --- /dev/null +++ b/core/src/main/scala/spark/metrics/source/JvmSource.scala @@ -0,0 +1,15 @@ +package spark.metrics.source + +import com.codahale.metrics.MetricRegistry +import com.codahale.metrics.jvm.{GarbageCollectorMetricSet, MemoryUsageGaugeSet} + +class JvmSource extends Source { + val sourceName = "jvm" + val metricRegistry = new MetricRegistry() + + val gcMetricSet = new GarbageCollectorMetricSet + val memGaugeSet = new MemoryUsageGaugeSet + + metricRegistry.registerAll(gcMetricSet) + metricRegistry.registerAll(memGaugeSet) +} diff --git a/core/src/main/scala/spark/metrics/source/Source.scala b/core/src/main/scala/spark/metrics/source/Source.scala new file mode 100644 index 0000000000..5607e2c40a --- /dev/null +++ b/core/src/main/scala/spark/metrics/source/Source.scala @@ -0,0 +1,8 @@ +package spark.metrics.source + +import com.codahale.metrics.MetricRegistry + +trait Source { + def sourceName: String + def metricRegistry: MetricRegistry +} diff --git a/core/src/main/scala/spark/network/BufferMessage.scala b/core/src/main/scala/spark/network/BufferMessage.scala index 7b0e489a6c..e566aeac13 100644 --- a/core/src/main/scala/spark/network/BufferMessage.scala +++ b/core/src/main/scala/spark/network/BufferMessage.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.network import java.nio.ByteBuffer @@ -91,4 +108,4 @@ class BufferMessage(id_ : Int, val buffers: ArrayBuffer[ByteBuffer], var ackId: "BufferMessage(id = " + id + ", size = " + size + ")" } } -}
\ No newline at end of file +} diff --git a/core/src/main/scala/spark/network/Connection.scala b/core/src/main/scala/spark/network/Connection.scala index 6e28f677a3..b66c00b58c 100644 --- a/core/src/main/scala/spark/network/Connection.scala +++ b/core/src/main/scala/spark/network/Connection.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.network import spark._ diff --git a/core/src/main/scala/spark/network/ConnectionManager.scala b/core/src/main/scala/spark/network/ConnectionManager.scala index 624a094856..6c4e7dc03e 100644 --- a/core/src/main/scala/spark/network/ConnectionManager.scala +++ b/core/src/main/scala/spark/network/ConnectionManager.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.network import spark._ diff --git a/core/src/main/scala/spark/network/ConnectionManagerId.scala b/core/src/main/scala/spark/network/ConnectionManagerId.scala index b554e84251..9d5c518293 100644 --- a/core/src/main/scala/spark/network/ConnectionManagerId.scala +++ b/core/src/main/scala/spark/network/ConnectionManagerId.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.network import java.net.InetSocketAddress diff --git a/core/src/main/scala/spark/network/ConnectionManagerTest.scala b/core/src/main/scala/spark/network/ConnectionManagerTest.scala index 533e4610f3..9e3827aaf5 100644 --- a/core/src/main/scala/spark/network/ConnectionManagerTest.scala +++ b/core/src/main/scala/spark/network/ConnectionManagerTest.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.network import spark._ diff --git a/core/src/main/scala/spark/network/Message.scala b/core/src/main/scala/spark/network/Message.scala index d4f03610eb..a25457ea35 100644 --- a/core/src/main/scala/spark/network/Message.scala +++ b/core/src/main/scala/spark/network/Message.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.network import java.nio.ByteBuffer diff --git a/core/src/main/scala/spark/network/MessageChunk.scala b/core/src/main/scala/spark/network/MessageChunk.scala index aaf9204d0e..784db5ab62 100644 --- a/core/src/main/scala/spark/network/MessageChunk.scala +++ b/core/src/main/scala/spark/network/MessageChunk.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.network import java.nio.ByteBuffer diff --git a/core/src/main/scala/spark/network/MessageChunkHeader.scala b/core/src/main/scala/spark/network/MessageChunkHeader.scala index 3693d509d6..18d0cbcc14 100644 --- a/core/src/main/scala/spark/network/MessageChunkHeader.scala +++ b/core/src/main/scala/spark/network/MessageChunkHeader.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.network import java.net.InetAddress diff --git a/core/src/main/scala/spark/network/ReceiverTest.scala b/core/src/main/scala/spark/network/ReceiverTest.scala index a174d5f403..2bbc736f40 100644 --- a/core/src/main/scala/spark/network/ReceiverTest.scala +++ b/core/src/main/scala/spark/network/ReceiverTest.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.network import java.nio.ByteBuffer diff --git a/core/src/main/scala/spark/network/SenderTest.scala b/core/src/main/scala/spark/network/SenderTest.scala index a4ff69e4d2..542c54c36b 100644 --- a/core/src/main/scala/spark/network/SenderTest.scala +++ b/core/src/main/scala/spark/network/SenderTest.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.network import java.nio.ByteBuffer diff --git a/core/src/main/scala/spark/network/netty/FileHeader.scala b/core/src/main/scala/spark/network/netty/FileHeader.scala index aed4254234..bf46d32aa3 100644 --- a/core/src/main/scala/spark/network/netty/FileHeader.scala +++ b/core/src/main/scala/spark/network/netty/FileHeader.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.network.netty import io.netty.buffer._ diff --git a/core/src/main/scala/spark/network/netty/ShuffleCopier.scala b/core/src/main/scala/spark/network/netty/ShuffleCopier.scala index 8d5194a737..b01f6369f6 100644 --- a/core/src/main/scala/spark/network/netty/ShuffleCopier.scala +++ b/core/src/main/scala/spark/network/netty/ShuffleCopier.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.network.netty import java.util.concurrent.Executors diff --git a/core/src/main/scala/spark/network/netty/ShuffleSender.scala b/core/src/main/scala/spark/network/netty/ShuffleSender.scala index d6fa4b1e80..cdf88b03a0 100644 --- a/core/src/main/scala/spark/network/netty/ShuffleSender.scala +++ b/core/src/main/scala/spark/network/netty/ShuffleSender.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.network.netty import java.io.File diff --git a/core/src/main/scala/spark/package.scala b/core/src/main/scala/spark/package.scala index 389ec4da3e..b244bfbf06 100644 --- a/core/src/main/scala/spark/package.scala +++ b/core/src/main/scala/spark/package.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + /** * Core Spark functionality. [[spark.SparkContext]] serves as the main entry point to Spark, while * [[spark.RDD]] is the data type representing a distributed collection, and provides most diff --git a/core/src/main/scala/spark/partial/ApproximateActionListener.scala b/core/src/main/scala/spark/partial/ApproximateActionListener.scala index de2dce161a..691d939150 100644 --- a/core/src/main/scala/spark/partial/ApproximateActionListener.scala +++ b/core/src/main/scala/spark/partial/ApproximateActionListener.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.partial import spark._ diff --git a/core/src/main/scala/spark/partial/ApproximateEvaluator.scala b/core/src/main/scala/spark/partial/ApproximateEvaluator.scala index 75713b2eaa..5eae144dfb 100644 --- a/core/src/main/scala/spark/partial/ApproximateEvaluator.scala +++ b/core/src/main/scala/spark/partial/ApproximateEvaluator.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.partial /** diff --git a/core/src/main/scala/spark/partial/BoundedDouble.scala b/core/src/main/scala/spark/partial/BoundedDouble.scala index 463c33d6e2..8bdbe6c012 100644 --- a/core/src/main/scala/spark/partial/BoundedDouble.scala +++ b/core/src/main/scala/spark/partial/BoundedDouble.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.partial /** diff --git a/core/src/main/scala/spark/partial/CountEvaluator.scala b/core/src/main/scala/spark/partial/CountEvaluator.scala index daf2c5170c..6aa92094eb 100644 --- a/core/src/main/scala/spark/partial/CountEvaluator.scala +++ b/core/src/main/scala/spark/partial/CountEvaluator.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.partial import cern.jet.stat.Probability diff --git a/core/src/main/scala/spark/partial/GroupedCountEvaluator.scala b/core/src/main/scala/spark/partial/GroupedCountEvaluator.scala index 01fbb8a11b..ebe2e5a1e3 100644 --- a/core/src/main/scala/spark/partial/GroupedCountEvaluator.scala +++ b/core/src/main/scala/spark/partial/GroupedCountEvaluator.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.partial import java.util.{HashMap => JHashMap} diff --git a/core/src/main/scala/spark/partial/GroupedMeanEvaluator.scala b/core/src/main/scala/spark/partial/GroupedMeanEvaluator.scala index c622df5220..2dadbbd5fb 100644 --- a/core/src/main/scala/spark/partial/GroupedMeanEvaluator.scala +++ b/core/src/main/scala/spark/partial/GroupedMeanEvaluator.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.partial import java.util.{HashMap => JHashMap} diff --git a/core/src/main/scala/spark/partial/GroupedSumEvaluator.scala b/core/src/main/scala/spark/partial/GroupedSumEvaluator.scala index 20fa55cff2..ae2b63f7cb 100644 --- a/core/src/main/scala/spark/partial/GroupedSumEvaluator.scala +++ b/core/src/main/scala/spark/partial/GroupedSumEvaluator.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.partial import java.util.{HashMap => JHashMap} diff --git a/core/src/main/scala/spark/partial/MeanEvaluator.scala b/core/src/main/scala/spark/partial/MeanEvaluator.scala index 762c85400d..5ddcad7075 100644 --- a/core/src/main/scala/spark/partial/MeanEvaluator.scala +++ b/core/src/main/scala/spark/partial/MeanEvaluator.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.partial import cern.jet.stat.Probability diff --git a/core/src/main/scala/spark/partial/PartialResult.scala b/core/src/main/scala/spark/partial/PartialResult.scala index 200ed4ea1e..922a9f9bc6 100644 --- a/core/src/main/scala/spark/partial/PartialResult.scala +++ b/core/src/main/scala/spark/partial/PartialResult.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.partial class PartialResult[R](initialVal: R, isFinal: Boolean) { diff --git a/core/src/main/scala/spark/partial/StudentTCacher.scala b/core/src/main/scala/spark/partial/StudentTCacher.scala index 443abba5cd..f3bb987d46 100644 --- a/core/src/main/scala/spark/partial/StudentTCacher.scala +++ b/core/src/main/scala/spark/partial/StudentTCacher.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.partial import cern.jet.stat.Probability diff --git a/core/src/main/scala/spark/partial/SumEvaluator.scala b/core/src/main/scala/spark/partial/SumEvaluator.scala index 58fb60f441..4083abef03 100644 --- a/core/src/main/scala/spark/partial/SumEvaluator.scala +++ b/core/src/main/scala/spark/partial/SumEvaluator.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.partial import cern.jet.stat.Probability diff --git a/core/src/main/scala/spark/rdd/BlockRDD.scala b/core/src/main/scala/spark/rdd/BlockRDD.scala index 719d4bf03e..0ebb722d73 100644 --- a/core/src/main/scala/spark/rdd/BlockRDD.scala +++ b/core/src/main/scala/spark/rdd/BlockRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import spark.{RDD, SparkContext, SparkEnv, Partition, TaskContext} diff --git a/core/src/main/scala/spark/rdd/CartesianRDD.scala b/core/src/main/scala/spark/rdd/CartesianRDD.scala index 38600b8be4..150e5bca29 100644 --- a/core/src/main/scala/spark/rdd/CartesianRDD.scala +++ b/core/src/main/scala/spark/rdd/CartesianRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import java.io.{ObjectOutputStream, IOException} diff --git a/core/src/main/scala/spark/rdd/CheckpointRDD.scala b/core/src/main/scala/spark/rdd/CheckpointRDD.scala index 377b1bdbe0..6794e0e201 100644 --- a/core/src/main/scala/spark/rdd/CheckpointRDD.scala +++ b/core/src/main/scala/spark/rdd/CheckpointRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import spark._ diff --git a/core/src/main/scala/spark/rdd/CoGroupedRDD.scala b/core/src/main/scala/spark/rdd/CoGroupedRDD.scala index 8966f9f86e..c540cd36eb 100644 --- a/core/src/main/scala/spark/rdd/CoGroupedRDD.scala +++ b/core/src/main/scala/spark/rdd/CoGroupedRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import java.io.{ObjectOutputStream, IOException} diff --git a/core/src/main/scala/spark/rdd/CoalescedRDD.scala b/core/src/main/scala/spark/rdd/CoalescedRDD.scala index 6d862c0c28..2b5bf18541 100644 --- a/core/src/main/scala/spark/rdd/CoalescedRDD.scala +++ b/core/src/main/scala/spark/rdd/CoalescedRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import spark.{Dependency, OneToOneDependency, NarrowDependency, RDD, Partition, TaskContext} diff --git a/core/src/main/scala/spark/rdd/EmptyRDD.scala b/core/src/main/scala/spark/rdd/EmptyRDD.scala index e4dd3a7fa7..d7d4db5d30 100644 --- a/core/src/main/scala/spark/rdd/EmptyRDD.scala +++ b/core/src/main/scala/spark/rdd/EmptyRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import spark.{RDD, SparkContext, SparkEnv, Partition, TaskContext} diff --git a/core/src/main/scala/spark/rdd/FilteredRDD.scala b/core/src/main/scala/spark/rdd/FilteredRDD.scala index c84ec39d21..783508cfd1 100644 --- a/core/src/main/scala/spark/rdd/FilteredRDD.scala +++ b/core/src/main/scala/spark/rdd/FilteredRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import spark.{OneToOneDependency, RDD, Partition, TaskContext} diff --git a/core/src/main/scala/spark/rdd/FlatMappedRDD.scala b/core/src/main/scala/spark/rdd/FlatMappedRDD.scala index 8ebc778925..ed75eac3ff 100644 --- a/core/src/main/scala/spark/rdd/FlatMappedRDD.scala +++ b/core/src/main/scala/spark/rdd/FlatMappedRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import spark.{RDD, Partition, TaskContext} diff --git a/core/src/main/scala/spark/rdd/GlommedRDD.scala b/core/src/main/scala/spark/rdd/GlommedRDD.scala index e16c7ba881..1573f8a289 100644 --- a/core/src/main/scala/spark/rdd/GlommedRDD.scala +++ b/core/src/main/scala/spark/rdd/GlommedRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import spark.{RDD, Partition, TaskContext} diff --git a/core/src/main/scala/spark/rdd/HadoopRDD.scala b/core/src/main/scala/spark/rdd/HadoopRDD.scala index 07c103503c..d0fdeb741e 100644 --- a/core/src/main/scala/spark/rdd/HadoopRDD.scala +++ b/core/src/main/scala/spark/rdd/HadoopRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import java.io.EOFException diff --git a/core/src/main/scala/spark/rdd/JdbcRDD.scala b/core/src/main/scala/spark/rdd/JdbcRDD.scala index a50f407737..59132437d2 100644 --- a/core/src/main/scala/spark/rdd/JdbcRDD.scala +++ b/core/src/main/scala/spark/rdd/JdbcRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import java.sql.{Connection, ResultSet} diff --git a/core/src/main/scala/spark/rdd/MapPartitionsRDD.scala b/core/src/main/scala/spark/rdd/MapPartitionsRDD.scala index d283c5b2bb..af8f0a112f 100644 --- a/core/src/main/scala/spark/rdd/MapPartitionsRDD.scala +++ b/core/src/main/scala/spark/rdd/MapPartitionsRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import spark.{RDD, Partition, TaskContext} diff --git a/core/src/main/scala/spark/rdd/MapPartitionsWithIndexRDD.scala b/core/src/main/scala/spark/rdd/MapPartitionsWithIndexRDD.scala index afb7504ba1..3b4e9518fd 100644 --- a/core/src/main/scala/spark/rdd/MapPartitionsWithIndexRDD.scala +++ b/core/src/main/scala/spark/rdd/MapPartitionsWithIndexRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import spark.{RDD, Partition, TaskContext} diff --git a/core/src/main/scala/spark/rdd/MappedRDD.scala b/core/src/main/scala/spark/rdd/MappedRDD.scala index af07311b6d..8b411dd85d 100644 --- a/core/src/main/scala/spark/rdd/MappedRDD.scala +++ b/core/src/main/scala/spark/rdd/MappedRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import spark.{RDD, Partition, TaskContext} diff --git a/core/src/main/scala/spark/rdd/NewHadoopRDD.scala b/core/src/main/scala/spark/rdd/NewHadoopRDD.scala index 901d01ef30..17fe805fd4 100644 --- a/core/src/main/scala/spark/rdd/NewHadoopRDD.scala +++ b/core/src/main/scala/spark/rdd/NewHadoopRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import java.text.SimpleDateFormat diff --git a/core/src/main/scala/spark/rdd/ParallelCollectionRDD.scala b/core/src/main/scala/spark/rdd/ParallelCollectionRDD.scala index 07585a88ce..16ba0c26f8 100644 --- a/core/src/main/scala/spark/rdd/ParallelCollectionRDD.scala +++ b/core/src/main/scala/spark/rdd/ParallelCollectionRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import scala.collection.immutable.NumericRange diff --git a/core/src/main/scala/spark/rdd/PartitionPruningRDD.scala b/core/src/main/scala/spark/rdd/PartitionPruningRDD.scala index 41ff62dd22..d8700becb0 100644 --- a/core/src/main/scala/spark/rdd/PartitionPruningRDD.scala +++ b/core/src/main/scala/spark/rdd/PartitionPruningRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import spark.{NarrowDependency, RDD, SparkEnv, Partition, TaskContext} @@ -16,8 +33,9 @@ class PruneDependency[T](rdd: RDD[T], @transient partitionFilterFunc: Int => Boo extends NarrowDependency[T](rdd) { @transient - val partitions: Array[Partition] = rdd.partitions.filter(s => partitionFilterFunc(s.index)) - .zipWithIndex.map { case(split, idx) => new PartitionPruningRDDPartition(idx, split) : Partition } + val partitions: Array[Partition] = rdd.partitions.zipWithIndex + .filter(s => partitionFilterFunc(s._2)) + .map { case(split, idx) => new PartitionPruningRDDPartition(idx, split) : Partition } override def getParents(partitionId: Int) = List(partitions(partitionId).index) } diff --git a/core/src/main/scala/spark/rdd/PipedRDD.scala b/core/src/main/scala/spark/rdd/PipedRDD.scala index c0baf43d43..2cefdc78b0 100644 --- a/core/src/main/scala/spark/rdd/PipedRDD.scala +++ b/core/src/main/scala/spark/rdd/PipedRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import java.io.PrintWriter diff --git a/core/src/main/scala/spark/rdd/SampledRDD.scala b/core/src/main/scala/spark/rdd/SampledRDD.scala index 243673f151..574c9b141d 100644 --- a/core/src/main/scala/spark/rdd/SampledRDD.scala +++ b/core/src/main/scala/spark/rdd/SampledRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import java.util.Random diff --git a/core/src/main/scala/spark/rdd/ShuffledRDD.scala b/core/src/main/scala/spark/rdd/ShuffledRDD.scala index c7d1926b83..0137f80953 100644 --- a/core/src/main/scala/spark/rdd/ShuffledRDD.scala +++ b/core/src/main/scala/spark/rdd/ShuffledRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import spark.{Partitioner, RDD, SparkEnv, ShuffleDependency, Partition, TaskContext} diff --git a/core/src/main/scala/spark/rdd/SubtractedRDD.scala b/core/src/main/scala/spark/rdd/SubtractedRDD.scala index 8a9efc5da2..0402b9f250 100644 --- a/core/src/main/scala/spark/rdd/SubtractedRDD.scala +++ b/core/src/main/scala/spark/rdd/SubtractedRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import java.util.{HashMap => JHashMap} diff --git a/core/src/main/scala/spark/rdd/UnionRDD.scala b/core/src/main/scala/spark/rdd/UnionRDD.scala index 2c52a67e22..2776826f18 100644 --- a/core/src/main/scala/spark/rdd/UnionRDD.scala +++ b/core/src/main/scala/spark/rdd/UnionRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import scala.collection.mutable.ArrayBuffer diff --git a/core/src/main/scala/spark/rdd/ZippedPartitionsRDD.scala b/core/src/main/scala/spark/rdd/ZippedPartitionsRDD.scala index b234428ab2..6a4fa13ad6 100644 --- a/core/src/main/scala/spark/rdd/ZippedPartitionsRDD.scala +++ b/core/src/main/scala/spark/rdd/ZippedPartitionsRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import spark.{Utils, OneToOneDependency, RDD, SparkContext, Partition, TaskContext} diff --git a/core/src/main/scala/spark/rdd/ZippedRDD.scala b/core/src/main/scala/spark/rdd/ZippedRDD.scala index f728e93d24..b1c43b3195 100644 --- a/core/src/main/scala/spark/rdd/ZippedRDD.scala +++ b/core/src/main/scala/spark/rdd/ZippedRDD.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import spark.{Utils, OneToOneDependency, RDD, SparkContext, Partition, TaskContext} diff --git a/core/src/main/scala/spark/scheduler/ActiveJob.scala b/core/src/main/scala/spark/scheduler/ActiveJob.scala index 105eaecb22..71cc94edb6 100644 --- a/core/src/main/scala/spark/scheduler/ActiveJob.scala +++ b/core/src/main/scala/spark/scheduler/ActiveJob.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler import spark.TaskContext diff --git a/core/src/main/scala/spark/scheduler/DAGScheduler.scala b/core/src/main/scala/spark/scheduler/DAGScheduler.scala index 5fda78e152..88c1b7baff 100644 --- a/core/src/main/scala/spark/scheduler/DAGScheduler.scala +++ b/core/src/main/scala/spark/scheduler/DAGScheduler.scala @@ -1,18 +1,33 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler -import cluster.TaskInfo -import java.util.concurrent.atomic.AtomicInteger -import java.util.concurrent.LinkedBlockingQueue -import java.util.concurrent.TimeUnit +import java.io.NotSerializableException import java.util.Properties +import java.util.concurrent.{LinkedBlockingQueue, TimeUnit} +import java.util.concurrent.atomic.AtomicInteger import scala.collection.mutable.{ArrayBuffer, HashMap, HashSet, Map} import spark._ import spark.executor.TaskMetrics -import spark.partial.ApproximateActionListener -import spark.partial.ApproximateEvaluator -import spark.partial.PartialResult +import spark.partial.{ApproximateActionListener, ApproximateEvaluator, PartialResult} +import spark.scheduler.cluster.TaskInfo import spark.storage.{BlockManager, BlockManagerMaster} import spark.util.{MetadataCleaner, TimeStampedHashMap} @@ -35,6 +50,11 @@ class DAGScheduler( } taskSched.setListener(this) + // Called by TaskScheduler to report task's starting. + override def taskStarted(task: Task[_], taskInfo: TaskInfo) { + eventQueue.put(BeginEvent(task, taskInfo)) + } + // Called by TaskScheduler to report task completions or failures. override def taskEnded( task: Task[_], @@ -150,7 +170,13 @@ class DAGScheduler( * as a result stage for the final RDD used directly in an action. The stage will also be given * the provided priority. */ - private def newStage(rdd: RDD[_], shuffleDep: Option[ShuffleDependency[_,_]], priority: Int): Stage = { + private def newStage( + rdd: RDD[_], + shuffleDep: Option[ShuffleDependency[_,_]], + priority: Int, + callSite: Option[String] = None) + : Stage = + { if (shuffleDep != None) { // Kind of ugly: need to register RDDs with the cache and map output tracker here // since we can't do it in the RDD constructor because # of partitions is unknown @@ -158,7 +184,7 @@ class DAGScheduler( mapOutputTracker.registerShuffle(shuffleDep.get.shuffleId, rdd.partitions.size) } val id = nextStageId.getAndIncrement() - val stage = new Stage(id, rdd, shuffleDep, getParentStages(rdd, priority), priority) + val stage = new Stage(id, rdd, shuffleDep, getParentStages(rdd, priority), priority, callSite) idToStage(id) = stage stageToInfos(stage) = StageInfo(stage) stage @@ -235,7 +261,8 @@ class DAGScheduler( assert(partitions.size > 0) val waiter = new JobWaiter(partitions.size, resultHandler) val func2 = func.asInstanceOf[(TaskContext, Iterator[_]) => _] - val toSubmit = JobSubmitted(finalRdd, func2, partitions.toArray, allowLocal, callSite, waiter, properties) + val toSubmit = JobSubmitted(finalRdd, func2, partitions.toArray, allowLocal, callSite, waiter, + properties) return (toSubmit, waiter) } @@ -251,7 +278,16 @@ class DAGScheduler( if (partitions.size == 0) { return } - val (toSubmit, waiter) = prepareJob( + + // Check to make sure we are not launching a task on a partition that does not exist. + val maxPartitions = finalRdd.partitions.length + partitions.find(p => p >= maxPartitions).foreach { p => + throw new IllegalArgumentException( + "Attempting to access a non-existent partition: " + p + ". " + + "Total number of partitions: " + maxPartitions) + } + + val (toSubmit: JobSubmitted, waiter: JobWaiter[_]) = prepareJob( finalRdd, func, partitions, callSite, allowLocal, resultHandler, properties) eventQueue.put(toSubmit) waiter.awaitResult() match { @@ -286,12 +322,12 @@ class DAGScheduler( event match { case JobSubmitted(finalRDD, func, partitions, allowLocal, callSite, listener, properties) => val runId = nextRunId.getAndIncrement() - val finalStage = newStage(finalRDD, None, runId) + val finalStage = newStage(finalRDD, None, runId, Some(callSite)) val job = new ActiveJob(runId, finalStage, func, partitions, callSite, listener, properties) clearCacheLocs() logInfo("Got job " + job.runId + " (" + callSite + ") with " + partitions.length + " output partitions (allowLocal=" + allowLocal + ")") - logInfo("Final stage: " + finalStage + " (" + finalStage.origin + ")") + logInfo("Final stage: " + finalStage + " (" + finalStage.name + ")") logInfo("Parents of final stage: " + finalStage.parents) logInfo("Missing parents: " + getMissingParentStages(finalStage)) if (allowLocal && finalStage.parents.size == 0 && partitions.length == 1) { @@ -311,6 +347,9 @@ class DAGScheduler( case ExecutorLost(execId) => handleExecutorLost(execId) + case begin: BeginEvent => + sparkListeners.foreach(_.onTaskStart(SparkListenerTaskStart(begin.task, begin.taskInfo))) + case completion: CompletionEvent => sparkListeners.foreach(_.onTaskEnd(SparkListenerTaskEnd(completion.task, completion.reason, completion.taskInfo, completion.taskMetrics))) @@ -472,6 +511,18 @@ class DAGScheduler( } } if (tasks.size > 0) { + // Preemptively serialize a task to make sure it can be serialized. We are catching this + // exception here because it would be fairly hard to catch the non-serializable exception + // down the road, where we have several different implementations for local scheduler and + // cluster schedulers. + try { + SparkEnv.get.closureSerializer.newInstance().serialize(tasks.head) + } catch { + case e: NotSerializableException => + abortStage(stage, e.toString) + running -= stage + return + } val properties = idToActiveJob(stage.priority).properties sparkListeners.foreach(_.onStageSubmitted( SparkListenerStageSubmitted(stage, tasks.size, properties))) @@ -503,7 +554,7 @@ class DAGScheduler( case Some(t) => "%.03f".format((System.currentTimeMillis() - t) / 1000.0) case _ => "Unkown" } - logInfo("%s (%s) finished in %s s".format(stage, stage.origin, serviceTime)) + logInfo("%s (%s) finished in %s s".format(stage, stage.name, serviceTime)) stage.completionTime = Some(System.currentTimeMillis) val stageComp = StageCompleted(stageToInfos(stage)) sparkListeners.foreach{_.onStageCompleted(stageComp)} @@ -569,7 +620,7 @@ class DAGScheduler( if (stage.outputLocs.count(_ == Nil) != 0) { // Some tasks had failed; let's resubmit this stage // TODO: Lower-level scheduler should also deal with this - logInfo("Resubmitting " + stage + " (" + stage.origin + + logInfo("Resubmitting " + stage + " (" + stage.name + ") because some of its tasks had failed: " + stage.outputLocs.zipWithIndex.filter(_._1 == Nil).map(_._2).mkString(", ")) submitStage(stage) @@ -601,7 +652,7 @@ class DAGScheduler( running -= failedStage failed += failedStage // TODO: Cancel running tasks in the stage - logInfo("Marking " + failedStage + " (" + failedStage.origin + + logInfo("Marking " + failedStage + " (" + failedStage.name + ") for resubmision due to a fetch failure") // Mark the map whose fetch failed as broken in the map stage val mapStage = shuffleToMapStage(shuffleId) @@ -609,7 +660,7 @@ class DAGScheduler( mapStage.removeOutputLoc(mapId, bmAddress) mapOutputTracker.unregisterMapOutput(shuffleId, mapId, bmAddress) } - logInfo("The failed fetch was from " + mapStage + " (" + mapStage.origin + + logInfo("The failed fetch was from " + mapStage + " (" + mapStage.name + "); marking it for resubmission") failed += mapStage // Remember that a fetch failed now; this is used to resubmit the broken diff --git a/core/src/main/scala/spark/scheduler/DAGSchedulerEvent.scala b/core/src/main/scala/spark/scheduler/DAGSchedulerEvent.scala index acad915f13..3b4ee6287a 100644 --- a/core/src/main/scala/spark/scheduler/DAGSchedulerEvent.scala +++ b/core/src/main/scala/spark/scheduler/DAGSchedulerEvent.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler import java.util.Properties @@ -26,6 +43,8 @@ private[spark] case class JobSubmitted( properties: Properties = null) extends DAGSchedulerEvent +private[spark] case class BeginEvent(task: Task[_], taskInfo: TaskInfo) extends DAGSchedulerEvent + private[spark] case class CompletionEvent( task: Task[_], reason: TaskEndReason, diff --git a/core/src/main/scala/spark/scheduler/DAGSchedulerSource.scala b/core/src/main/scala/spark/scheduler/DAGSchedulerSource.scala new file mode 100644 index 0000000000..87d27cc70d --- /dev/null +++ b/core/src/main/scala/spark/scheduler/DAGSchedulerSource.scala @@ -0,0 +1,30 @@ +package spark.scheduler + +import com.codahale.metrics.{Gauge,MetricRegistry} + +import spark.metrics.source.Source + +private[spark] class DAGSchedulerSource(val dagScheduler: DAGScheduler) extends Source { + val metricRegistry = new MetricRegistry() + val sourceName = "DAGScheduler" + + metricRegistry.register(MetricRegistry.name("stage", "failedStages", "number"), new Gauge[Int] { + override def getValue: Int = dagScheduler.failed.size + }) + + metricRegistry.register(MetricRegistry.name("stage", "runningStages", "number"), new Gauge[Int] { + override def getValue: Int = dagScheduler.running.size + }) + + metricRegistry.register(MetricRegistry.name("stage", "waitingStages", "number"), new Gauge[Int] { + override def getValue: Int = dagScheduler.waiting.size + }) + + metricRegistry.register(MetricRegistry.name("job", "allJobs", "number"), new Gauge[Int] { + override def getValue: Int = dagScheduler.nextRunId.get() + }) + + metricRegistry.register(MetricRegistry.name("job", "activeJobs", "number"), new Gauge[Int] { + override def getValue: Int = dagScheduler.activeJobs.size + }) +} diff --git a/core/src/main/scala/spark/scheduler/InputFormatInfo.scala b/core/src/main/scala/spark/scheduler/InputFormatInfo.scala index 17d0ea4f80..65f8c3200e 100644 --- a/core/src/main/scala/spark/scheduler/InputFormatInfo.scala +++ b/core/src/main/scala/spark/scheduler/InputFormatInfo.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler import spark.Logging diff --git a/core/src/main/scala/spark/scheduler/JobListener.scala b/core/src/main/scala/spark/scheduler/JobListener.scala index f46b9d551d..af108b8fec 100644 --- a/core/src/main/scala/spark/scheduler/JobListener.scala +++ b/core/src/main/scala/spark/scheduler/JobListener.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler /** diff --git a/core/src/main/scala/spark/scheduler/JobLogger.scala b/core/src/main/scala/spark/scheduler/JobLogger.scala index 8e5540873f..c7e8f8a9a1 100644 --- a/core/src/main/scala/spark/scheduler/JobLogger.scala +++ b/core/src/main/scala/spark/scheduler/JobLogger.scala @@ -1,3 +1,20 @@ +/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
package spark.scheduler
import java.io.PrintWriter
@@ -51,6 +68,8 @@ class JobLogger(val logDirName: String) extends SparkListener with Logging { processStageCompletedEvent(stageInfo)
case SparkListenerJobEnd(job, result) =>
processJobEndEvent(job, result)
+ case SparkListenerTaskStart(task, taskInfo) =>
+ processTaskStartEvent(task, taskInfo)
case SparkListenerTaskEnd(task, reason, taskInfo, taskMetrics) =>
processTaskEndEvent(task, reason, taskInfo, taskMetrics)
case _ =>
@@ -235,7 +254,19 @@ class JobLogger(val logDirName: String) extends SparkListener with Logging { stageInfo.stage.id + " STATUS=COMPLETED")
}
-
+
+ override def onTaskStart(taskStart: SparkListenerTaskStart) {
+ eventQueue.put(taskStart)
+ }
+
+ protected def processTaskStartEvent(task: Task[_], taskInfo: TaskInfo) {
+ var taskStatus = ""
+ task match {
+ case resultTask: ResultTask[_, _] => taskStatus = "TASK_TYPE=RESULT_TASK"
+ case shuffleMapTask: ShuffleMapTask => taskStatus = "TASK_TYPE=SHUFFLE_MAP_TASK"
+ }
+ }
+
override def onTaskEnd(taskEnd: SparkListenerTaskEnd) {
eventQueue.put(taskEnd)
}
diff --git a/core/src/main/scala/spark/scheduler/JobResult.scala b/core/src/main/scala/spark/scheduler/JobResult.scala index a0fdf391e6..a61b335152 100644 --- a/core/src/main/scala/spark/scheduler/JobResult.scala +++ b/core/src/main/scala/spark/scheduler/JobResult.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler /** diff --git a/core/src/main/scala/spark/scheduler/JobWaiter.scala b/core/src/main/scala/spark/scheduler/JobWaiter.scala index 6ff2e29434..69cd161c1f 100644 --- a/core/src/main/scala/spark/scheduler/JobWaiter.scala +++ b/core/src/main/scala/spark/scheduler/JobWaiter.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler import scala.collection.mutable.ArrayBuffer diff --git a/core/src/main/scala/spark/scheduler/MapStatus.scala b/core/src/main/scala/spark/scheduler/MapStatus.scala index 203abb917b..2f6a68ee85 100644 --- a/core/src/main/scala/spark/scheduler/MapStatus.scala +++ b/core/src/main/scala/spark/scheduler/MapStatus.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler import spark.storage.BlockManagerId diff --git a/core/src/main/scala/spark/scheduler/ResultTask.scala b/core/src/main/scala/spark/scheduler/ResultTask.scala index 83166bce22..361b1e6b91 100644 --- a/core/src/main/scala/spark/scheduler/ResultTask.scala +++ b/core/src/main/scala/spark/scheduler/ResultTask.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler import spark._ diff --git a/core/src/main/scala/spark/scheduler/ShuffleMapTask.scala b/core/src/main/scala/spark/scheduler/ShuffleMapTask.scala index 95647389c3..1c25605f75 100644 --- a/core/src/main/scala/spark/scheduler/ShuffleMapTask.scala +++ b/core/src/main/scala/spark/scheduler/ShuffleMapTask.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler import java.io._ diff --git a/core/src/main/scala/spark/scheduler/SparkListener.scala b/core/src/main/scala/spark/scheduler/SparkListener.scala index 49f7c85c29..2a09a956ad 100644 --- a/core/src/main/scala/spark/scheduler/SparkListener.scala +++ b/core/src/main/scala/spark/scheduler/SparkListener.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler import java.util.Properties @@ -13,6 +30,8 @@ case class SparkListenerStageSubmitted(stage: Stage, taskSize: Int, properties: case class StageCompleted(val stageInfo: StageInfo) extends SparkListenerEvents +case class SparkListenerTaskStart(task: Task[_], taskInfo: TaskInfo) extends SparkListenerEvents + case class SparkListenerTaskEnd(task: Task[_], reason: TaskEndReason, taskInfo: TaskInfo, taskMetrics: TaskMetrics) extends SparkListenerEvents @@ -34,6 +53,11 @@ trait SparkListener { def onStageSubmitted(stageSubmitted: SparkListenerStageSubmitted) { } /** + * Called when a task starts + */ + def onTaskStart(taskEnd: SparkListenerTaskStart) { } + + /** * Called when a task ends */ def onTaskEnd(taskEnd: SparkListenerTaskEnd) { } diff --git a/core/src/main/scala/spark/scheduler/SplitInfo.scala b/core/src/main/scala/spark/scheduler/SplitInfo.scala index 6abfb7a1f7..4e3661ec5d 100644 --- a/core/src/main/scala/spark/scheduler/SplitInfo.scala +++ b/core/src/main/scala/spark/scheduler/SplitInfo.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler import collection.mutable.ArrayBuffer diff --git a/core/src/main/scala/spark/scheduler/Stage.scala b/core/src/main/scala/spark/scheduler/Stage.scala index 539cf8233b..5428daeb94 100644 --- a/core/src/main/scala/spark/scheduler/Stage.scala +++ b/core/src/main/scala/spark/scheduler/Stage.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler import java.net.URI @@ -24,7 +41,8 @@ private[spark] class Stage( val rdd: RDD[_], val shuffleDep: Option[ShuffleDependency[_,_]], // Output shuffle if stage is a map stage val parents: List[Stage], - val priority: Int) + val priority: Int, + callSite: Option[String]) extends Logging { val isShuffleMap = shuffleDep != None @@ -85,7 +103,7 @@ private[spark] class Stage( return id } - def origin: String = rdd.origin + val name = callSite.getOrElse(rdd.origin) override def toString = "Stage " + id diff --git a/core/src/main/scala/spark/scheduler/StageInfo.scala b/core/src/main/scala/spark/scheduler/StageInfo.scala index 8d83ff10c4..c4026f995a 100644 --- a/core/src/main/scala/spark/scheduler/StageInfo.scala +++ b/core/src/main/scala/spark/scheduler/StageInfo.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler import spark.scheduler.cluster.TaskInfo @@ -9,4 +26,4 @@ case class StageInfo( val taskInfos: mutable.Buffer[(TaskInfo, TaskMetrics)] = mutable.Buffer[(TaskInfo, TaskMetrics)]() ) { override def toString = stage.rdd.toString -}
\ No newline at end of file +} diff --git a/core/src/main/scala/spark/scheduler/Task.scala b/core/src/main/scala/spark/scheduler/Task.scala index a6462c6968..50768d43e0 100644 --- a/core/src/main/scala/spark/scheduler/Task.scala +++ b/core/src/main/scala/spark/scheduler/Task.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler import spark.serializer.SerializerInstance diff --git a/core/src/main/scala/spark/scheduler/TaskResult.scala b/core/src/main/scala/spark/scheduler/TaskResult.scala index 6de0aa7adf..dc0621ea7b 100644 --- a/core/src/main/scala/spark/scheduler/TaskResult.scala +++ b/core/src/main/scala/spark/scheduler/TaskResult.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler import java.io._ diff --git a/core/src/main/scala/spark/scheduler/TaskScheduler.scala b/core/src/main/scala/spark/scheduler/TaskScheduler.scala index 5cdf846032..4943d58e25 100644 --- a/core/src/main/scala/spark/scheduler/TaskScheduler.scala +++ b/core/src/main/scala/spark/scheduler/TaskScheduler.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler import spark.scheduler.cluster.Pool diff --git a/core/src/main/scala/spark/scheduler/TaskSchedulerListener.scala b/core/src/main/scala/spark/scheduler/TaskSchedulerListener.scala index b75d3736cf..2cdeb1c8c0 100644 --- a/core/src/main/scala/spark/scheduler/TaskSchedulerListener.scala +++ b/core/src/main/scala/spark/scheduler/TaskSchedulerListener.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler import spark.scheduler.cluster.TaskInfo @@ -10,6 +27,9 @@ import spark.executor.TaskMetrics * Interface for getting events back from the TaskScheduler. */ private[spark] trait TaskSchedulerListener { + // A task has started. + def taskStarted(task: Task[_], taskInfo: TaskInfo) + // A task has finished or failed. def taskEnded(task: Task[_], reason: TaskEndReason, result: Any, accumUpdates: Map[Long, Any], taskInfo: TaskInfo, taskMetrics: TaskMetrics): Unit diff --git a/core/src/main/scala/spark/scheduler/TaskSet.scala b/core/src/main/scala/spark/scheduler/TaskSet.scala index e4b5fcaedb..dc3550dd0b 100644 --- a/core/src/main/scala/spark/scheduler/TaskSet.scala +++ b/core/src/main/scala/spark/scheduler/TaskSet.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler import java.util.Properties diff --git a/core/src/main/scala/spark/scheduler/cluster/ClusterScheduler.scala b/core/src/main/scala/spark/scheduler/cluster/ClusterScheduler.scala index 20680bbf87..d92dd3de8b 100644 --- a/core/src/main/scala/spark/scheduler/cluster/ClusterScheduler.scala +++ b/core/src/main/scala/spark/scheduler/cluster/ClusterScheduler.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler.cluster import java.lang.{Boolean => JBoolean} @@ -257,7 +274,7 @@ private[spark] class ClusterScheduler(val sc: SparkContext) val sortedTaskSetQueue = rootPool.getSortedTaskSetQueue() for (manager <- sortedTaskSetQueue) { - logInfo("parentName:%s, name:%s, runningTasks:%s".format( + logDebug("parentName:%s, name:%s, runningTasks:%s".format( manager.parent.name, manager.name, manager.runningTasks)) } diff --git a/core/src/main/scala/spark/scheduler/cluster/ClusterTaskSetManager.scala b/core/src/main/scala/spark/scheduler/cluster/ClusterTaskSetManager.scala index 5e2351bafd..40a11b25b7 100644 --- a/core/src/main/scala/spark/scheduler/cluster/ClusterTaskSetManager.scala +++ b/core/src/main/scala/spark/scheduler/cluster/ClusterTaskSetManager.scala @@ -1,6 +1,24 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler.cluster -import java.util.{HashMap => JHashMap, NoSuchElementException, Arrays} +import java.nio.ByteBuffer +import java.util.{Arrays, NoSuchElementException} import scala.collection.mutable.ArrayBuffer import scala.collection.mutable.HashMap @@ -8,12 +26,14 @@ import scala.collection.mutable.HashSet import scala.math.max import scala.math.min -import spark._ -import spark.scheduler._ +import spark.{FetchFailed, Logging, Resubmitted, SparkEnv, Success, TaskEndReason, TaskState, Utils} +import spark.{ExceptionFailure, SparkException, TaskResultTooBigFailure} import spark.TaskState.TaskState -import java.nio.ByteBuffer +import spark.scheduler.{ShuffleMapTask, Task, TaskResult, TaskSet} + -private[spark] object TaskLocality extends Enumeration("PROCESS_LOCAL", "NODE_LOCAL", "RACK_LOCAL", "ANY") with Logging { +private[spark] object TaskLocality + extends Enumeration("PROCESS_LOCAL", "NODE_LOCAL", "RACK_LOCAL", "ANY") with Logging { // process local is expected to be used ONLY within tasksetmanager for now. val PROCESS_LOCAL, NODE_LOCAL, RACK_LOCAL, ANY = Value @@ -26,8 +46,10 @@ private[spark] object TaskLocality extends Enumeration("PROCESS_LOCAL", "NODE_LO assert (constraint != TaskLocality.PROCESS_LOCAL) constraint match { - case TaskLocality.NODE_LOCAL => condition == TaskLocality.NODE_LOCAL - case TaskLocality.RACK_LOCAL => condition == TaskLocality.NODE_LOCAL || condition == TaskLocality.RACK_LOCAL + case TaskLocality.NODE_LOCAL => + condition == TaskLocality.NODE_LOCAL + case TaskLocality.RACK_LOCAL => + condition == TaskLocality.NODE_LOCAL || condition == TaskLocality.RACK_LOCAL // For anything else, allow case _ => true } @@ -39,11 +61,10 @@ private[spark] object TaskLocality extends Enumeration("PROCESS_LOCAL", "NODE_LO val retval = TaskLocality.withName(str) // Must not specify PROCESS_LOCAL ! assert (retval != TaskLocality.PROCESS_LOCAL) - retval } catch { case nEx: NoSuchElementException => { - logWarning("Invalid task locality specified '" + str + "', defaulting to NODE_LOCAL"); + logWarning("Invalid task locality specified '" + str + "', defaulting to NODE_LOCAL") // default to preserve earlier behavior NODE_LOCAL } @@ -54,11 +75,8 @@ private[spark] object TaskLocality extends Enumeration("PROCESS_LOCAL", "NODE_LO /** * Schedules the tasks within a single TaskSet in the ClusterScheduler. */ -private[spark] class ClusterTaskSetManager( - sched: ClusterScheduler, - val taskSet: TaskSet) - extends TaskSetManager - with Logging { +private[spark] class ClusterTaskSetManager(sched: ClusterScheduler, val taskSet: TaskSet) + extends TaskSetManager with Logging { // Maximum time to wait to run a task in a preferred location (in ms) val LOCALITY_WAIT = System.getProperty("spark.locality.wait", "3000").toLong @@ -94,7 +112,8 @@ private[spark] class ClusterTaskSetManager( // Last time when we launched a preferred task (for delay scheduling) var lastPreferredLaunchTime = System.currentTimeMillis - // List of pending tasks for each node (process local to container). These collections are actually + // List of pending tasks for each node (process local to container). + // These collections are actually // treated as stacks, in which new tasks are added to the end of the // ArrayBuffer and removed from the end. This makes it faster to detect // tasks that repeatedly fail because whenever a task failed, it is put @@ -154,9 +173,11 @@ private[spark] class ClusterTaskSetManager( // Note that it follows the hierarchy. // if we search for NODE_LOCAL, the output will include PROCESS_LOCAL and // if we search for RACK_LOCAL, it will include PROCESS_LOCAL & NODE_LOCAL - private def findPreferredLocations(_taskPreferredLocations: Seq[String], scheduler: ClusterScheduler, - taskLocality: TaskLocality.TaskLocality): HashSet[String] = { - + private def findPreferredLocations( + _taskPreferredLocations: Seq[String], + scheduler: ClusterScheduler, + taskLocality: TaskLocality.TaskLocality): HashSet[String] = + { if (TaskLocality.PROCESS_LOCAL == taskLocality) { // straight forward comparison ! Special case it. val retval = new HashSet[String]() @@ -171,13 +192,14 @@ private[spark] class ClusterTaskSetManager( return retval } - val taskPreferredLocations = + val taskPreferredLocations = { if (TaskLocality.NODE_LOCAL == taskLocality) { _taskPreferredLocations } else { assert (TaskLocality.RACK_LOCAL == taskLocality) // Expand set to include all 'seen' rack local hosts. - // This works since container allocation/management happens within master - so any rack locality information is updated in msater. + // This works since container allocation/management happens within master - + // so any rack locality information is updated in msater. // Best case effort, and maybe sort of kludge for now ... rework it later ? val hosts = new HashSet[String] _taskPreferredLocations.foreach(h => { @@ -195,6 +217,7 @@ private[spark] class ClusterTaskSetManager( hosts } + } val retval = new HashSet[String] scheduler.synchronized { @@ -211,11 +234,13 @@ private[spark] class ClusterTaskSetManager( // Add a task to all the pending-task lists that it should be on. private def addPendingTask(index: Int) { - // We can infer hostLocalLocations from rackLocalLocations by joining it against tasks(index).preferredLocations (with appropriate - // hostPort <-> host conversion). But not doing it for simplicity sake. If this becomes a performance issue, modify it. - val processLocalLocations = findPreferredLocations(tasks(index).preferredLocations, sched, TaskLocality.PROCESS_LOCAL) - val hostLocalLocations = findPreferredLocations(tasks(index).preferredLocations, sched, TaskLocality.NODE_LOCAL) - val rackLocalLocations = findPreferredLocations(tasks(index).preferredLocations, sched, TaskLocality.RACK_LOCAL) + // We can infer hostLocalLocations from rackLocalLocations by joining it against + // tasks(index).preferredLocations (with appropriate hostPort <-> host conversion). + // But not doing it for simplicity sake. If this becomes a performance issue, modify it. + val locs = tasks(index).preferredLocations + val processLocalLocations = findPreferredLocations(locs, sched, TaskLocality.PROCESS_LOCAL) + val hostLocalLocations = findPreferredLocations(locs, sched, TaskLocality.NODE_LOCAL) + val rackLocalLocations = findPreferredLocations(locs, sched, TaskLocality.RACK_LOCAL) if (rackLocalLocations.size == 0) { // Current impl ensures this. @@ -280,18 +305,24 @@ private[spark] class ClusterTaskSetManager( } // Number of pending tasks for a given host Port (which would be process local) - def numPendingTasksForHostPort(hostPort: String): Int = { - getPendingTasksForHostPort(hostPort).count( index => copiesRunning(index) == 0 && !finished(index) ) + override def numPendingTasksForHostPort(hostPort: String): Int = { + getPendingTasksForHostPort(hostPort).count { index => + copiesRunning(index) == 0 && !finished(index) + } } // Number of pending tasks for a given host (which would be data local) - def numPendingTasksForHost(hostPort: String): Int = { - getPendingTasksForHost(hostPort).count( index => copiesRunning(index) == 0 && !finished(index) ) + override def numPendingTasksForHost(hostPort: String): Int = { + getPendingTasksForHost(hostPort).count { index => + copiesRunning(index) == 0 && !finished(index) + } } // Number of pending rack local tasks for a given host - def numRackLocalPendingTasksForHost(hostPort: String): Int = { - getRackLocalPendingTasksForHost(hostPort).count( index => copiesRunning(index) == 0 && !finished(index) ) + override def numRackLocalPendingTasksForHost(hostPort: String): Int = { + getRackLocalPendingTasksForHost(hostPort).count { index => + copiesRunning(index) == 0 && !finished(index) + } } @@ -319,12 +350,12 @@ private[spark] class ClusterTaskSetManager( speculatableTasks.retain(index => !finished(index)) // Remove finished tasks from set if (speculatableTasks.size > 0) { - val localTask = speculatableTasks.find { - index => - val locations = findPreferredLocations(tasks(index).preferredLocations, sched, TaskLocality.NODE_LOCAL) - val attemptLocs = taskAttempts(index).map(_.hostPort) - (locations.size == 0 || locations.contains(hostPort)) && !attemptLocs.contains(hostPort) - } + val localTask = speculatableTasks.find { index => + val locations = findPreferredLocations(tasks(index).preferredLocations, sched, + TaskLocality.NODE_LOCAL) + val attemptLocs = taskAttempts(index).map(_.hostPort) + (locations.size == 0 || locations.contains(hostPort)) && !attemptLocs.contains(hostPort) + } if (localTask != None) { speculatableTasks -= localTask.get @@ -333,11 +364,11 @@ private[spark] class ClusterTaskSetManager( // check for rack locality if (TaskLocality.isAllowed(locality, TaskLocality.RACK_LOCAL)) { - val rackTask = speculatableTasks.find { - index => - val locations = findPreferredLocations(tasks(index).preferredLocations, sched, TaskLocality.RACK_LOCAL) - val attemptLocs = taskAttempts(index).map(_.hostPort) - locations.contains(hostPort) && !attemptLocs.contains(hostPort) + val rackTask = speculatableTasks.find { index => + val locations = findPreferredLocations(tasks(index).preferredLocations, sched, + TaskLocality.RACK_LOCAL) + val attemptLocs = taskAttempts(index).map(_.hostPort) + locations.contains(hostPort) && !attemptLocs.contains(hostPort) } if (rackTask != None) { @@ -349,7 +380,9 @@ private[spark] class ClusterTaskSetManager( // Any task ... if (TaskLocality.isAllowed(locality, TaskLocality.ANY)) { // Check for attemptLocs also ? - val nonLocalTask = speculatableTasks.find(i => !taskAttempts(i).map(_.hostPort).contains(hostPort)) + val nonLocalTask = speculatableTasks.find { i => + !taskAttempts(i).map(_.hostPort).contains(hostPort) + } if (nonLocalTask != None) { speculatableTasks -= nonLocalTask.get return nonLocalTask @@ -379,7 +412,8 @@ private[spark] class ClusterTaskSetManager( } } - // Look for no pref tasks AFTER rack local tasks - this has side effect that we will get to failed tasks later rather than sooner. + // Look for no pref tasks AFTER rack local tasks - this has side effect that we will get to + // failed tasks later rather than sooner. // TODO: That code path needs to be revisited (adding to no prefs list when host:port goes down). val noPrefTask = findTaskFromList(pendingTasksWithNoPrefs) if (noPrefTask != None) { @@ -415,7 +449,8 @@ private[spark] class ClusterTaskSetManager( locs.find(h => Utils.parseHostPort(h)._1 == host).isDefined } - // Does a host count as a rack local preferred location for a task? (assumes host is NOT preferred location). + // Does a host count as a rack local preferred location for a task? + // (assumes host is NOT preferred location). // This is true if either the task has preferred locations and this host is one, or it has // no preferred locations (in which we still count the launch as preferred). private def isRackLocalLocation(task: Task[_], hostPort: String): Boolean = { @@ -436,14 +471,22 @@ private[spark] class ClusterTaskSetManager( } // Respond to an offer of a single slave from the scheduler by finding a task - def slaveOffer(execId: String, hostPort: String, availableCpus: Double, overrideLocality: TaskLocality.TaskLocality = null): Option[TaskDescription] = { - + override def slaveOffer( + execId: String, + hostPort: String, + availableCpus: Double, + overrideLocality: TaskLocality.TaskLocality = null): Option[TaskDescription] = + { if (tasksFinished < numTasks && availableCpus >= CPUS_PER_TASK) { // If explicitly specified, use that val locality = if (overrideLocality != null) overrideLocality else { // expand only if we have waited for more than LOCALITY_WAIT for a host local task ... val time = System.currentTimeMillis - if (time - lastPreferredLaunchTime < LOCALITY_WAIT) TaskLocality.NODE_LOCAL else TaskLocality.ANY + if (time - lastPreferredLaunchTime < LOCALITY_WAIT) { + TaskLocality.NODE_LOCAL + } else { + TaskLocality.ANY + } } findTask(hostPort, locality) match { @@ -453,10 +496,10 @@ private[spark] class ClusterTaskSetManager( val taskId = sched.newTaskId() // Figure out whether this should count as a preferred launch val taskLocality = - if (isProcessLocalLocation(task, hostPort)) TaskLocality.PROCESS_LOCAL else - if (isHostLocalLocation(task, hostPort)) TaskLocality.NODE_LOCAL else - if (isRackLocalLocation(task, hostPort)) TaskLocality.RACK_LOCAL else - TaskLocality.ANY + if (isProcessLocalLocation(task, hostPort)) TaskLocality.PROCESS_LOCAL + else if (isHostLocalLocation(task, hostPort)) TaskLocality.NODE_LOCAL + else if (isRackLocalLocation(task, hostPort)) TaskLocality.RACK_LOCAL + else TaskLocality.ANY val prefStr = taskLocality.toString logInfo("Starting task %s:%d as TID %s on slave %s: %s (%s)".format( taskSet.id, index, taskId, execId, hostPort, prefStr)) @@ -466,11 +509,13 @@ private[spark] class ClusterTaskSetManager( val info = new TaskInfo(taskId, index, time, execId, hostPort, taskLocality) taskInfos(taskId) = info taskAttempts(index) = info :: taskAttempts(index) - if (TaskLocality.NODE_LOCAL == taskLocality) { + if (taskLocality == TaskLocality.PROCESS_LOCAL || taskLocality == TaskLocality.NODE_LOCAL) { lastPreferredLaunchTime = time } // Serialize and return the task val startTime = System.currentTimeMillis + // We rely on the DAGScheduler to catch non-serializable closures and RDDs, so in here + // we assume the task can be serialized without exceptions. val serializedTask = Task.serializeWithDependencies( task, sched.sc.addedFiles, sched.sc.addedJars, ser) val timeTaken = System.currentTimeMillis - startTime @@ -478,6 +523,8 @@ private[spark] class ClusterTaskSetManager( logInfo("Serialized task %s:%d as %d bytes in %d ms".format( taskSet.id, index, serializedTask.limit, timeTaken)) val taskName = "task %s:%d".format(taskSet.id, index) + if (taskAttempts(index).size == 1) + taskStarted(task,info) return Some(new TaskDescription(taskId, execId, taskName, serializedTask)) } case _ => @@ -486,7 +533,7 @@ private[spark] class ClusterTaskSetManager( return None } - def statusUpdate(tid: Long, state: TaskState, serializedData: ByteBuffer) { + override def statusUpdate(tid: Long, state: TaskState, serializedData: ByteBuffer) { state match { case TaskState.FINISHED => taskFinished(tid, state, serializedData) @@ -500,6 +547,10 @@ private[spark] class ClusterTaskSetManager( } } + def taskStarted(task: Task[_], info: TaskInfo) { + sched.listener.taskStarted(task, info) + } + def taskFinished(tid: Long, state: TaskState, serializedData: ByteBuffer) { val info = taskInfos(tid) if (info.failed) { @@ -512,13 +563,14 @@ private[spark] class ClusterTaskSetManager( decreaseRunningTasks(1) if (!finished(index)) { tasksFinished += 1 - logInfo("Finished TID %s in %d ms (progress: %d/%d)".format( - tid, info.duration, tasksFinished, numTasks)) + logInfo("Finished TID %s in %d ms on %s (progress: %d/%d)".format( + tid, info.duration, info.hostPort, tasksFinished, numTasks)) // Deserialize task result and pass it to the scheduler try { val result = ser.deserialize[TaskResult[_]](serializedData) result.metrics.resultSize = serializedData.limit() - sched.listener.taskEnded(tasks(index), Success, result.value, result.accumUpdates, info, result.metrics) + sched.listener.taskEnded( + tasks(index), Success, result.value, result.accumUpdates, info, result.metrics) } catch { case cnf: ClassNotFoundException => val loader = Thread.currentThread().getContextClassLoader @@ -564,8 +616,8 @@ private[spark] class ClusterTaskSetManager( return case taskResultTooBig: TaskResultTooBigFailure => - logInfo("Loss was due to task %s result exceeding Akka frame size; " + - "aborting job".format(tid)) + logInfo("Loss was due to task %s result exceeding Akka frame size; aborting job".format( + tid)) abort("Task %s result exceeded Akka frame size".format(tid)) return @@ -616,7 +668,7 @@ private[spark] class ClusterTaskSetManager( } } - def error(message: String) { + override def error(message: String) { // Save the error message abort("Error: " + message) } @@ -644,7 +696,7 @@ private[spark] class ClusterTaskSetManager( } } - // TODO: for now we just find Pool not TaskSetManager + // TODO(xiajunluan): for now we just find Pool not TaskSetManager // we can extend this function in future if needed override def getSchedulableByName(name: String): Schedulable = { return null @@ -670,13 +722,15 @@ private[spark] class ClusterTaskSetManager( // If some task has preferred locations only on hostname, and there are no more executors there, // put it in the no-prefs list to avoid the wait from delay scheduling - // host local tasks - should we push this to rack local or no pref list ? For now, preserving behavior and moving to - // no prefs list. Note, this was done due to impliations related to 'waiting' for data local tasks, etc. - // Note: NOT checking process local list - since host local list is super set of that. We need to ad to no prefs only if - // there is no host local node for the task (not if there is no process local node for the task) + // host local tasks - should we push this to rack local or no pref list ? For now, preserving + // behavior and moving to no prefs list. Note, this was done due to impliations related to + // 'waiting' for data local tasks, etc. + // Note: NOT checking process local list - since host local list is super set of that. We need + // to ad to no prefs only if there is no host local node for the task (not if there is no + // process local node for the task) for (index <- getPendingTasksForHost(Utils.parseHostPort(hostPort)._1)) { - // val newLocs = findPreferredLocations(tasks(index).preferredLocations, sched, TaskLocality.RACK_LOCAL) - val newLocs = findPreferredLocations(tasks(index).preferredLocations, sched, TaskLocality.NODE_LOCAL) + val newLocs = findPreferredLocations( + tasks(index).preferredLocations, sched, TaskLocality.NODE_LOCAL) if (newLocs.isEmpty) { pendingTasksWithNoPrefs += index } diff --git a/core/src/main/scala/spark/scheduler/cluster/ExecutorLossReason.scala b/core/src/main/scala/spark/scheduler/cluster/ExecutorLossReason.scala index 8bf838209f..8825f2dd24 100644 --- a/core/src/main/scala/spark/scheduler/cluster/ExecutorLossReason.scala +++ b/core/src/main/scala/spark/scheduler/cluster/ExecutorLossReason.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler.cluster import spark.executor.ExecutorExitCode diff --git a/core/src/main/scala/spark/scheduler/cluster/Pool.scala b/core/src/main/scala/spark/scheduler/cluster/Pool.scala index 941ba7a3f1..83708f07e1 100644 --- a/core/src/main/scala/spark/scheduler/cluster/Pool.scala +++ b/core/src/main/scala/spark/scheduler/cluster/Pool.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler.cluster import scala.collection.mutable.ArrayBuffer diff --git a/core/src/main/scala/spark/scheduler/cluster/Schedulable.scala b/core/src/main/scala/spark/scheduler/cluster/Schedulable.scala index c410af8af4..e77e8e4162 100644 --- a/core/src/main/scala/spark/scheduler/cluster/Schedulable.scala +++ b/core/src/main/scala/spark/scheduler/cluster/Schedulable.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler.cluster import spark.scheduler.cluster.SchedulingMode.SchedulingMode diff --git a/core/src/main/scala/spark/scheduler/cluster/SchedulableBuilder.scala b/core/src/main/scala/spark/scheduler/cluster/SchedulableBuilder.scala index db51b48494..b2d089f31d 100644 --- a/core/src/main/scala/spark/scheduler/cluster/SchedulableBuilder.scala +++ b/core/src/main/scala/spark/scheduler/cluster/SchedulableBuilder.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler.cluster import java.io.{File, FileInputStream, FileOutputStream} diff --git a/core/src/main/scala/spark/scheduler/cluster/SchedulerBackend.scala b/core/src/main/scala/spark/scheduler/cluster/SchedulerBackend.scala index 8844057a5c..4431744ec3 100644 --- a/core/src/main/scala/spark/scheduler/cluster/SchedulerBackend.scala +++ b/core/src/main/scala/spark/scheduler/cluster/SchedulerBackend.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler.cluster import spark.{SparkContext, Utils} diff --git a/core/src/main/scala/spark/scheduler/cluster/SchedulingAlgorithm.scala b/core/src/main/scala/spark/scheduler/cluster/SchedulingAlgorithm.scala index f33310a34a..69e0ac2a6b 100644 --- a/core/src/main/scala/spark/scheduler/cluster/SchedulingAlgorithm.scala +++ b/core/src/main/scala/spark/scheduler/cluster/SchedulingAlgorithm.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler.cluster /** diff --git a/core/src/main/scala/spark/scheduler/cluster/SchedulingMode.scala b/core/src/main/scala/spark/scheduler/cluster/SchedulingMode.scala index cd0642772d..55cdf4791f 100644 --- a/core/src/main/scala/spark/scheduler/cluster/SchedulingMode.scala +++ b/core/src/main/scala/spark/scheduler/cluster/SchedulingMode.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler.cluster /** diff --git a/core/src/main/scala/spark/scheduler/cluster/SparkDeploySchedulerBackend.scala b/core/src/main/scala/spark/scheduler/cluster/SparkDeploySchedulerBackend.scala index 170ede0f44..55d6c0a47e 100644 --- a/core/src/main/scala/spark/scheduler/cluster/SparkDeploySchedulerBackend.scala +++ b/core/src/main/scala/spark/scheduler/cluster/SparkDeploySchedulerBackend.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler.cluster import spark.{Utils, Logging, SparkContext} diff --git a/core/src/main/scala/spark/scheduler/cluster/StandaloneClusterMessage.scala b/core/src/main/scala/spark/scheduler/cluster/StandaloneClusterMessage.scala index 3335294844..05c29eb72f 100644 --- a/core/src/main/scala/spark/scheduler/cluster/StandaloneClusterMessage.scala +++ b/core/src/main/scala/spark/scheduler/cluster/StandaloneClusterMessage.scala @@ -1,45 +1,63 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler.cluster -import spark.TaskState.TaskState import java.nio.ByteBuffer -import spark.util.SerializableBuffer + +import spark.TaskState.TaskState import spark.Utils +import spark.util.SerializableBuffer + private[spark] sealed trait StandaloneClusterMessage extends Serializable -// Driver to executors -private[spark] -case class LaunchTask(task: TaskDescription) extends StandaloneClusterMessage +private[spark] object StandaloneClusterMessages { -private[spark] -case class RegisteredExecutor(sparkProperties: Seq[(String, String)]) - extends StandaloneClusterMessage + // Driver to executors + case class LaunchTask(task: TaskDescription) extends StandaloneClusterMessage -private[spark] -case class RegisterExecutorFailed(message: String) extends StandaloneClusterMessage + case class RegisteredExecutor(sparkProperties: Seq[(String, String)]) + extends StandaloneClusterMessage -// Executors to driver -private[spark] -case class RegisterExecutor(executorId: String, hostPort: String, cores: Int) - extends StandaloneClusterMessage { - Utils.checkHostPort(hostPort, "Expected host port") -} + case class RegisterExecutorFailed(message: String) extends StandaloneClusterMessage + + // Executors to driver + case class RegisterExecutor(executorId: String, hostPort: String, cores: Int) + extends StandaloneClusterMessage { + Utils.checkHostPort(hostPort, "Expected host port") + } -private[spark] -case class StatusUpdate(executorId: String, taskId: Long, state: TaskState, data: SerializableBuffer) - extends StandaloneClusterMessage + case class StatusUpdate(executorId: String, taskId: Long, state: TaskState, + data: SerializableBuffer) extends StandaloneClusterMessage -private[spark] -object StatusUpdate { - /** Alternate factory method that takes a ByteBuffer directly for the data field */ - def apply(executorId: String, taskId: Long, state: TaskState, data: ByteBuffer): StatusUpdate = { - StatusUpdate(executorId, taskId, state, new SerializableBuffer(data)) + object StatusUpdate { + /** Alternate factory method that takes a ByteBuffer directly for the data field */ + def apply(executorId: String, taskId: Long, state: TaskState, data: ByteBuffer) + : StatusUpdate = { + StatusUpdate(executorId, taskId, state, new SerializableBuffer(data)) + } } -} -// Internal messages in driver -private[spark] case object ReviveOffers extends StandaloneClusterMessage -private[spark] case object StopDriver extends StandaloneClusterMessage + // Internal messages in driver + case object ReviveOffers extends StandaloneClusterMessage -private[spark] case class RemoveExecutor(executorId: String, reason: String) - extends StandaloneClusterMessage + case object StopDriver extends StandaloneClusterMessage + + case class RemoveExecutor(executorId: String, reason: String) extends StandaloneClusterMessage + +} diff --git a/core/src/main/scala/spark/scheduler/cluster/StandaloneSchedulerBackend.scala b/core/src/main/scala/spark/scheduler/cluster/StandaloneSchedulerBackend.scala index 004592a540..075a7cbf7e 100644 --- a/core/src/main/scala/spark/scheduler/cluster/StandaloneSchedulerBackend.scala +++ b/core/src/main/scala/spark/scheduler/cluster/StandaloneSchedulerBackend.scala @@ -1,16 +1,34 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler.cluster +import java.util.concurrent.atomic.AtomicInteger + import scala.collection.mutable.{ArrayBuffer, HashMap, HashSet} import akka.actor._ -import akka.util.duration._ +import akka.dispatch.Await import akka.pattern.ask +import akka.remote.{RemoteClientShutdown, RemoteClientDisconnected, RemoteClientLifeCycleEvent} import akka.util.Duration import spark.{Utils, SparkException, Logging, TaskState} -import akka.dispatch.Await -import java.util.concurrent.atomic.AtomicInteger -import akka.remote.{RemoteClientShutdown, RemoteClientDisconnected, RemoteClientLifeCycleEvent} +import spark.scheduler.cluster.StandaloneClusterMessages._ /** * A standalone scheduler backend, which waits for standalone executors to connect to it through diff --git a/core/src/main/scala/spark/scheduler/cluster/TaskDescription.scala b/core/src/main/scala/spark/scheduler/cluster/TaskDescription.scala index b41e951be9..761fdf6919 100644 --- a/core/src/main/scala/spark/scheduler/cluster/TaskDescription.scala +++ b/core/src/main/scala/spark/scheduler/cluster/TaskDescription.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler.cluster import java.nio.ByteBuffer diff --git a/core/src/main/scala/spark/scheduler/cluster/TaskInfo.scala b/core/src/main/scala/spark/scheduler/cluster/TaskInfo.scala index 718f26bfbd..c693b722ac 100644 --- a/core/src/main/scala/spark/scheduler/cluster/TaskInfo.scala +++ b/core/src/main/scala/spark/scheduler/cluster/TaskInfo.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler.cluster import spark.Utils @@ -34,6 +51,17 @@ class TaskInfo( def running: Boolean = !finished + def status: String = { + if (running) + "RUNNING" + else if (failed) + "FAILED" + else if (successful) + "SUCCESS" + else + "UNKNOWN" + } + def duration: Long = { if (!finished) { throw new UnsupportedOperationException("duration() called on unfinished tasks") diff --git a/core/src/main/scala/spark/scheduler/cluster/TaskSetManager.scala b/core/src/main/scala/spark/scheduler/cluster/TaskSetManager.scala index 4e6bc51278..1a92a5ed6f 100644 --- a/core/src/main/scala/spark/scheduler/cluster/TaskSetManager.scala +++ b/core/src/main/scala/spark/scheduler/cluster/TaskSetManager.scala @@ -1,19 +1,47 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler.cluster -import scala.collection.mutable.ArrayBuffer -import spark.scheduler._ -import spark.TaskState.TaskState import java.nio.ByteBuffer +import spark.TaskState.TaskState +import spark.scheduler.TaskSet + private[spark] trait TaskSetManager extends Schedulable { def schedulableQueue = null + def schedulingMode = SchedulingMode.NONE + def taskSet: TaskSet - def slaveOffer(execId: String, hostPort: String, availableCpus: Double, - overrideLocality: TaskLocality.TaskLocality = null): Option[TaskDescription] + + def slaveOffer( + execId: String, + hostPort: String, + availableCpus: Double, + overrideLocality: TaskLocality.TaskLocality = null): Option[TaskDescription] + def numPendingTasksForHostPort(hostPort: String): Int - def numRackLocalPendingTasksForHost(hostPort :String): Int + + def numRackLocalPendingTasksForHost(hostPort: String): Int + def numPendingTasksForHost(hostPort: String): Int + def statusUpdate(tid: Long, state: TaskState, serializedData: ByteBuffer) + def error(message: String) } diff --git a/core/src/main/scala/spark/scheduler/cluster/WorkerOffer.scala b/core/src/main/scala/spark/scheduler/cluster/WorkerOffer.scala index c47824315c..06d1203f70 100644 --- a/core/src/main/scala/spark/scheduler/cluster/WorkerOffer.scala +++ b/core/src/main/scala/spark/scheduler/cluster/WorkerOffer.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler.cluster /** diff --git a/core/src/main/scala/spark/scheduler/local/LocalScheduler.scala b/core/src/main/scala/spark/scheduler/local/LocalScheduler.scala index f4411582f1..bb0c836e86 100644 --- a/core/src/main/scala/spark/scheduler/local/LocalScheduler.scala +++ b/core/src/main/scala/spark/scheduler/local/LocalScheduler.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler.local import java.io.File @@ -21,10 +38,15 @@ import akka.actor._ * testing fault recovery. */ -private[spark] case class LocalReviveOffers() -private[spark] case class LocalStatusUpdate(taskId: Long, state: TaskState, serializedData: ByteBuffer) +private[spark] +case class LocalReviveOffers() + +private[spark] +case class LocalStatusUpdate(taskId: Long, state: TaskState, serializedData: ByteBuffer) + +private[spark] +class LocalActor(localScheduler: LocalScheduler, var freeCores: Int) extends Actor with Logging { -private[spark] class LocalActor(localScheduler: LocalScheduler, var freeCores: Int) extends Actor with Logging { def receive = { case LocalReviveOffers => launchTask(localScheduler.resourceOffer(freeCores)) @@ -39,7 +61,7 @@ private[spark] class LocalActor(localScheduler: LocalScheduler, var freeCores: I freeCores -= 1 localScheduler.threadPool.submit(new Runnable { def run() { - localScheduler.runTask(task.taskId,task.serializedTask) + localScheduler.runTask(task.taskId, task.serializedTask) } }) } @@ -94,7 +116,7 @@ private[spark] class LocalScheduler(threads: Int, val maxFailures: Int, val sc: override def submitTasks(taskSet: TaskSet) { synchronized { - var manager = new LocalTaskSetManager(this, taskSet) + val manager = new LocalTaskSetManager(this, taskSet) schedulableBuilder.addTaskSetManager(manager, manager.taskSet.properties) activeTaskSets(taskSet.id) = manager taskSetTaskIds(taskSet.id) = new HashSet[Long]() @@ -108,14 +130,15 @@ private[spark] class LocalScheduler(threads: Int, val maxFailures: Int, val sc: val tasks = new ArrayBuffer[TaskDescription](freeCores) val sortedTaskSetQueue = rootPool.getSortedTaskSetQueue() for (manager <- sortedTaskSetQueue) { - logDebug("parentName:%s,name:%s,runningTasks:%s".format(manager.parent.name, manager.name, manager.runningTasks)) + logDebug("parentName:%s,name:%s,runningTasks:%s".format( + manager.parent.name, manager.name, manager.runningTasks)) } var launchTask = false for (manager <- sortedTaskSetQueue) { do { launchTask = false - manager.slaveOffer(null,null,freeCpuCores) match { + manager.slaveOffer(null, null, freeCpuCores) match { case Some(task) => tasks += task taskIdToTaskSetId(task.taskId) = manager.taskSet.id @@ -123,7 +146,7 @@ private[spark] class LocalScheduler(threads: Int, val maxFailures: Int, val sc: freeCpuCores -= 1 launchTask = true case None => {} - } + } } while(launchTask) } return tasks diff --git a/core/src/main/scala/spark/scheduler/local/LocalTaskSetManager.scala b/core/src/main/scala/spark/scheduler/local/LocalTaskSetManager.scala index cc27f1ecca..4ab15532cf 100644 --- a/core/src/main/scala/spark/scheduler/local/LocalTaskSetManager.scala +++ b/core/src/main/scala/spark/scheduler/local/LocalTaskSetManager.scala @@ -1,26 +1,42 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler.local -import java.io.File -import java.util.concurrent.atomic.AtomicInteger import java.nio.ByteBuffer import scala.collection.mutable.ArrayBuffer import scala.collection.mutable.HashMap -import scala.collection.mutable.HashSet -import spark._ +import spark.{ExceptionFailure, Logging, SparkEnv, Success, TaskState} import spark.TaskState.TaskState -import spark.scheduler._ -import spark.scheduler.cluster._ +import spark.scheduler.{Task, TaskResult, TaskSet} +import spark.scheduler.cluster.{Schedulable, TaskDescription, TaskInfo, TaskLocality, TaskSetManager} + + +private[spark] class LocalTaskSetManager(sched: LocalScheduler, val taskSet: TaskSet) + extends TaskSetManager with Logging { -private[spark] class LocalTaskSetManager(sched: LocalScheduler, val taskSet: TaskSet) extends TaskSetManager with Logging { var parent: Schedulable = null var weight: Int = 1 var minShare: Int = 0 var runningTasks: Int = 0 var priority: Int = taskSet.priority var stageId: Int = taskSet.stageId - var name: String = "TaskSet_"+taskSet.stageId.toString - + var name: String = "TaskSet_" + taskSet.stageId.toString var failCount = new Array[Int](taskSet.tasks.size) val taskInfos = new HashMap[Long, TaskInfo] @@ -32,49 +48,45 @@ private[spark] class LocalTaskSetManager(sched: LocalScheduler, val taskSet: Tas val numFailures = new Array[Int](numTasks) val MAX_TASK_FAILURES = sched.maxFailures - def increaseRunningTasks(taskNum: Int): Unit = { - runningTasks += taskNum - if (parent != null) { - parent.increaseRunningTasks(taskNum) - } + override def increaseRunningTasks(taskNum: Int): Unit = { + runningTasks += taskNum + if (parent != null) { + parent.increaseRunningTasks(taskNum) + } } - def decreaseRunningTasks(taskNum: Int): Unit = { + override def decreaseRunningTasks(taskNum: Int): Unit = { runningTasks -= taskNum if (parent != null) { parent.decreaseRunningTasks(taskNum) } } - def addSchedulable(schedulable: Schedulable): Unit = { + override def addSchedulable(schedulable: Schedulable): Unit = { // nothing } - def removeSchedulable(schedulable: Schedulable): Unit = { + override def removeSchedulable(schedulable: Schedulable): Unit = { // nothing } - def getSchedulableByName(name: String): Schedulable = { + override def getSchedulableByName(name: String): Schedulable = { return null } - def executorLost(executorId: String, host: String): Unit = { + override def executorLost(executorId: String, host: String): Unit = { // nothing } - def checkSpeculatableTasks(): Boolean = { - return true - } + override def checkSpeculatableTasks() = true - def getSortedTaskSetQueue(): ArrayBuffer[TaskSetManager] = { + override def getSortedTaskSetQueue(): ArrayBuffer[TaskSetManager] = { var sortedTaskSetQueue = new ArrayBuffer[TaskSetManager] sortedTaskSetQueue += this return sortedTaskSetQueue } - def hasPendingTasks(): Boolean = { - return true - } + override def hasPendingTasks() = true def findTask(): Option[Int] = { for (i <- 0 to numTasks-1) { @@ -85,21 +97,32 @@ private[spark] class LocalTaskSetManager(sched: LocalScheduler, val taskSet: Tas return None } - def slaveOffer(execId: String, hostPort: String, availableCpus: Double, overrideLocality: TaskLocality.TaskLocality = null): Option[TaskDescription] = { + override def slaveOffer( + execId: String, + hostPort: String, + availableCpus: Double, + overrideLocality: TaskLocality.TaskLocality = null): Option[TaskDescription] = + { SparkEnv.set(sched.env) - logDebug("availableCpus:%d,numFinished:%d,numTasks:%d".format(availableCpus.toInt, numFinished, numTasks)) + logDebug("availableCpus:%d,numFinished:%d,numTasks:%d".format( + availableCpus.toInt, numFinished, numTasks)) if (availableCpus > 0 && numFinished < numTasks) { findTask() match { case Some(index) => val taskId = sched.attemptId.getAndIncrement() val task = taskSet.tasks(index) - val info = new TaskInfo(taskId, index, System.currentTimeMillis(), "local", "local:1", TaskLocality.NODE_LOCAL) + val info = new TaskInfo(taskId, index, System.currentTimeMillis(), "local", "local:1", + TaskLocality.NODE_LOCAL) taskInfos(taskId) = info - val bytes = Task.serializeWithDependencies(task, sched.sc.addedFiles, sched.sc.addedJars, ser) + // We rely on the DAGScheduler to catch non-serializable closures and RDDs, so in here + // we assume the task can be serialized without exceptions. + val bytes = Task.serializeWithDependencies( + task, sched.sc.addedFiles, sched.sc.addedJars, ser) logInfo("Size of task " + taskId + " is " + bytes.limit + " bytes") val taskName = "task %s:%d".format(taskSet.id, index) copiesRunning(index) += 1 increaseRunningTasks(1) + taskStarted(task, info) return Some(new TaskDescription(taskId, null, taskName, bytes)) case None => {} } @@ -107,19 +130,19 @@ private[spark] class LocalTaskSetManager(sched: LocalScheduler, val taskSet: Tas return None } - def numPendingTasksForHostPort(hostPort: String): Int = { + override def numPendingTasksForHostPort(hostPort: String): Int = { return 0 } - def numRackLocalPendingTasksForHost(hostPort :String): Int = { + override def numRackLocalPendingTasksForHost(hostPort :String): Int = { return 0 } - def numPendingTasksForHost(hostPort: String): Int = { + override def numPendingTasksForHost(hostPort: String): Int = { return 0 } - def statusUpdate(tid: Long, state: TaskState, serializedData: ByteBuffer) { + override def statusUpdate(tid: Long, state: TaskState, serializedData: ByteBuffer) { state match { case TaskState.FINISHED => taskEnded(tid, state, serializedData) @@ -129,6 +152,10 @@ private[spark] class LocalTaskSetManager(sched: LocalScheduler, val taskSet: Tas } } + def taskStarted(task: Task[_], info: TaskInfo) { + sched.listener.taskStarted(task, info) + } + def taskEnded(tid: Long, state: TaskState, serializedData: ByteBuffer) { val info = taskInfos(tid) val index = info.index @@ -151,15 +178,18 @@ private[spark] class LocalTaskSetManager(sched: LocalScheduler, val taskSet: Tas val task = taskSet.tasks(index) info.markFailed() decreaseRunningTasks(1) - val reason: ExceptionFailure = ser.deserialize[ExceptionFailure](serializedData, getClass.getClassLoader) + val reason: ExceptionFailure = ser.deserialize[ExceptionFailure]( + serializedData, getClass.getClassLoader) sched.listener.taskEnded(task, reason, null, null, info, reason.metrics.getOrElse(null)) if (!finished(index)) { copiesRunning(index) -= 1 numFailures(index) += 1 val locs = reason.stackTrace.map(loc => "\tat %s".format(loc.toString)) - logInfo("Loss was due to %s\n%s\n%s".format(reason.className, reason.description, locs.mkString("\n"))) + logInfo("Loss was due to %s\n%s\n%s".format( + reason.className, reason.description, locs.mkString("\n"))) if (numFailures(index) > MAX_TASK_FAILURES) { - val errorMessage = "Task %s:%d failed more than %d times; aborting job %s".format(taskSet.id, index, 4, reason.description) + val errorMessage = "Task %s:%d failed more than %d times; aborting job %s".format( + taskSet.id, index, 4, reason.description) decreaseRunningTasks(runningTasks) sched.listener.taskSetFailed(taskSet, errorMessage) // need to delete failed Taskset from schedule queue @@ -168,6 +198,6 @@ private[spark] class LocalTaskSetManager(sched: LocalScheduler, val taskSet: Tas } } - def error(message: String) { + override def error(message: String) { } } diff --git a/core/src/main/scala/spark/scheduler/mesos/CoarseMesosSchedulerBackend.scala b/core/src/main/scala/spark/scheduler/mesos/CoarseMesosSchedulerBackend.scala index f4a2994b6d..7bc6040544 100644 --- a/core/src/main/scala/spark/scheduler/mesos/CoarseMesosSchedulerBackend.scala +++ b/core/src/main/scala/spark/scheduler/mesos/CoarseMesosSchedulerBackend.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler.mesos import com.google.protobuf.ByteString diff --git a/core/src/main/scala/spark/scheduler/mesos/MesosSchedulerBackend.scala b/core/src/main/scala/spark/scheduler/mesos/MesosSchedulerBackend.scala index ca7fab4cc5..75b8268b55 100644 --- a/core/src/main/scala/spark/scheduler/mesos/MesosSchedulerBackend.scala +++ b/core/src/main/scala/spark/scheduler/mesos/MesosSchedulerBackend.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler.mesos import com.google.protobuf.ByteString @@ -43,8 +60,12 @@ private[spark] class MesosSchedulerBackend( // An ExecutorInfo for our tasks var execArgs: Array[Byte] = null + var classLoader: ClassLoader = null + override def start() { synchronized { + classLoader = Thread.currentThread.getContextClassLoader + new Thread("MesosSchedulerBackend driver") { setDaemon(true) override def run() { @@ -114,13 +135,28 @@ private[spark] class MesosSchedulerBackend( return execArgs } + private def setClassLoader(): ClassLoader = { + val oldClassLoader = Thread.currentThread.getContextClassLoader + Thread.currentThread.setContextClassLoader(classLoader) + return oldClassLoader + } + + private def restoreClassLoader(oldClassLoader: ClassLoader) { + Thread.currentThread.setContextClassLoader(oldClassLoader) + } + override def offerRescinded(d: SchedulerDriver, o: OfferID) {} override def registered(d: SchedulerDriver, frameworkId: FrameworkID, masterInfo: MasterInfo) { - logInfo("Registered as framework ID " + frameworkId.getValue) - registeredLock.synchronized { - isRegistered = true - registeredLock.notifyAll() + val oldClassLoader = setClassLoader() + try { + logInfo("Registered as framework ID " + frameworkId.getValue) + registeredLock.synchronized { + isRegistered = true + registeredLock.notifyAll() + } + } finally { + restoreClassLoader(oldClassLoader) } } @@ -142,49 +178,54 @@ private[spark] class MesosSchedulerBackend( * tasks are balanced across the cluster. */ override def resourceOffers(d: SchedulerDriver, offers: JList[Offer]) { - synchronized { - // Build a big list of the offerable workers, and remember their indices so that we can - // figure out which Offer to reply to for each worker - val offerableIndices = new ArrayBuffer[Int] - val offerableWorkers = new ArrayBuffer[WorkerOffer] - - def enoughMemory(o: Offer) = { - val mem = getResource(o.getResourcesList, "mem") - val slaveId = o.getSlaveId.getValue - mem >= executorMemory || slaveIdsWithExecutors.contains(slaveId) - } + val oldClassLoader = setClassLoader() + try { + synchronized { + // Build a big list of the offerable workers, and remember their indices so that we can + // figure out which Offer to reply to for each worker + val offerableIndices = new ArrayBuffer[Int] + val offerableWorkers = new ArrayBuffer[WorkerOffer] + + def enoughMemory(o: Offer) = { + val mem = getResource(o.getResourcesList, "mem") + val slaveId = o.getSlaveId.getValue + mem >= executorMemory || slaveIdsWithExecutors.contains(slaveId) + } - for ((offer, index) <- offers.zipWithIndex if enoughMemory(offer)) { - offerableIndices += index - offerableWorkers += new WorkerOffer( - offer.getSlaveId.getValue, - offer.getHostname, - getResource(offer.getResourcesList, "cpus").toInt) - } + for ((offer, index) <- offers.zipWithIndex if enoughMemory(offer)) { + offerableIndices += index + offerableWorkers += new WorkerOffer( + offer.getSlaveId.getValue, + offer.getHostname, + getResource(offer.getResourcesList, "cpus").toInt) + } - // Call into the ClusterScheduler - val taskLists = scheduler.resourceOffers(offerableWorkers) - - // Build a list of Mesos tasks for each slave - val mesosTasks = offers.map(o => Collections.emptyList[MesosTaskInfo]()) - for ((taskList, index) <- taskLists.zipWithIndex) { - if (!taskList.isEmpty) { - val offerNum = offerableIndices(index) - val slaveId = offers(offerNum).getSlaveId.getValue - slaveIdsWithExecutors += slaveId - mesosTasks(offerNum) = new JArrayList[MesosTaskInfo](taskList.size) - for (taskDesc <- taskList) { - taskIdToSlaveId(taskDesc.taskId) = slaveId - mesosTasks(offerNum).add(createMesosTask(taskDesc, slaveId)) + // Call into the ClusterScheduler + val taskLists = scheduler.resourceOffers(offerableWorkers) + + // Build a list of Mesos tasks for each slave + val mesosTasks = offers.map(o => Collections.emptyList[MesosTaskInfo]()) + for ((taskList, index) <- taskLists.zipWithIndex) { + if (!taskList.isEmpty) { + val offerNum = offerableIndices(index) + val slaveId = offers(offerNum).getSlaveId.getValue + slaveIdsWithExecutors += slaveId + mesosTasks(offerNum) = new JArrayList[MesosTaskInfo](taskList.size) + for (taskDesc <- taskList) { + taskIdToSlaveId(taskDesc.taskId) = slaveId + mesosTasks(offerNum).add(createMesosTask(taskDesc, slaveId)) + } } } - } - // Reply to the offers - val filters = Filters.newBuilder().setRefuseSeconds(1).build() // TODO: lower timeout? - for (i <- 0 until offers.size) { - d.launchTasks(offers(i).getId, mesosTasks(i), filters) + // Reply to the offers + val filters = Filters.newBuilder().setRefuseSeconds(1).build() // TODO: lower timeout? + for (i <- 0 until offers.size) { + d.launchTasks(offers(i).getId, mesosTasks(i), filters) + } } + } finally { + restoreClassLoader(oldClassLoader) } } @@ -224,23 +265,33 @@ private[spark] class MesosSchedulerBackend( } override def statusUpdate(d: SchedulerDriver, status: TaskStatus) { - val tid = status.getTaskId.getValue.toLong - val state = TaskState.fromMesos(status.getState) - synchronized { - if (status.getState == MesosTaskState.TASK_LOST && taskIdToSlaveId.contains(tid)) { - // We lost the executor on this slave, so remember that it's gone - slaveIdsWithExecutors -= taskIdToSlaveId(tid) - } - if (isFinished(status.getState)) { - taskIdToSlaveId.remove(tid) + val oldClassLoader = setClassLoader() + try { + val tid = status.getTaskId.getValue.toLong + val state = TaskState.fromMesos(status.getState) + synchronized { + if (status.getState == MesosTaskState.TASK_LOST && taskIdToSlaveId.contains(tid)) { + // We lost the executor on this slave, so remember that it's gone + slaveIdsWithExecutors -= taskIdToSlaveId(tid) + } + if (isFinished(status.getState)) { + taskIdToSlaveId.remove(tid) + } } + scheduler.statusUpdate(tid, state, status.getData.asReadOnlyByteBuffer) + } finally { + restoreClassLoader(oldClassLoader) } - scheduler.statusUpdate(tid, state, status.getData.asReadOnlyByteBuffer) } override def error(d: SchedulerDriver, message: String) { - logError("Mesos error: " + message) - scheduler.error(message) + val oldClassLoader = setClassLoader() + try { + logError("Mesos error: " + message) + scheduler.error(message) + } finally { + restoreClassLoader(oldClassLoader) + } } override def stop() { @@ -256,11 +307,16 @@ private[spark] class MesosSchedulerBackend( override def frameworkMessage(d: SchedulerDriver, e: ExecutorID, s: SlaveID, b: Array[Byte]) {} private def recordSlaveLost(d: SchedulerDriver, slaveId: SlaveID, reason: ExecutorLossReason) { - logInfo("Mesos slave lost: " + slaveId.getValue) - synchronized { - slaveIdsWithExecutors -= slaveId.getValue + val oldClassLoader = setClassLoader() + try { + logInfo("Mesos slave lost: " + slaveId.getValue) + synchronized { + slaveIdsWithExecutors -= slaveId.getValue + } + scheduler.executorLost(slaveId.getValue, reason) + } finally { + restoreClassLoader(oldClassLoader) } - scheduler.executorLost(slaveId.getValue, reason) } override def slaveLost(d: SchedulerDriver, slaveId: SlaveID) { diff --git a/core/src/main/scala/spark/serializer/Serializer.scala b/core/src/main/scala/spark/serializer/Serializer.scala index 2ad73b711d..dc94d42bb6 100644 --- a/core/src/main/scala/spark/serializer/Serializer.scala +++ b/core/src/main/scala/spark/serializer/Serializer.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.serializer import java.io.{EOFException, InputStream, OutputStream} diff --git a/core/src/main/scala/spark/serializer/SerializerManager.scala b/core/src/main/scala/spark/serializer/SerializerManager.scala index 60b2aac797..b7b24705a2 100644 --- a/core/src/main/scala/spark/serializer/SerializerManager.scala +++ b/core/src/main/scala/spark/serializer/SerializerManager.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.serializer import java.util.concurrent.ConcurrentHashMap diff --git a/core/src/main/scala/spark/storage/BlockException.scala b/core/src/main/scala/spark/storage/BlockException.scala index f275d476df..8ebfaf3cbf 100644 --- a/core/src/main/scala/spark/storage/BlockException.scala +++ b/core/src/main/scala/spark/storage/BlockException.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.storage private[spark] diff --git a/core/src/main/scala/spark/storage/BlockFetchTracker.scala b/core/src/main/scala/spark/storage/BlockFetchTracker.scala index 0718156b1b..265e554ad8 100644 --- a/core/src/main/scala/spark/storage/BlockFetchTracker.scala +++ b/core/src/main/scala/spark/storage/BlockFetchTracker.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.storage private[spark] trait BlockFetchTracker { diff --git a/core/src/main/scala/spark/storage/BlockFetcherIterator.scala b/core/src/main/scala/spark/storage/BlockFetcherIterator.scala index bec876213e..1965c5bc19 100644 --- a/core/src/main/scala/spark/storage/BlockFetcherIterator.scala +++ b/core/src/main/scala/spark/storage/BlockFetcherIterator.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.storage import java.nio.ByteBuffer diff --git a/core/src/main/scala/spark/storage/BlockManager.scala b/core/src/main/scala/spark/storage/BlockManager.scala index 9b39d3aadf..e4ffa57ad2 100644 --- a/core/src/main/scala/spark/storage/BlockManager.scala +++ b/core/src/main/scala/spark/storage/BlockManager.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.storage import java.io.{InputStream, OutputStream} diff --git a/core/src/main/scala/spark/storage/BlockManagerId.scala b/core/src/main/scala/spark/storage/BlockManagerId.scala index 1e557d6148..b36a6176c0 100644 --- a/core/src/main/scala/spark/storage/BlockManagerId.scala +++ b/core/src/main/scala/spark/storage/BlockManagerId.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.storage import java.io.{Externalizable, IOException, ObjectInput, ObjectOutput} diff --git a/core/src/main/scala/spark/storage/BlockManagerMaster.scala b/core/src/main/scala/spark/storage/BlockManagerMaster.scala index 58888b1ebb..76128e8cff 100644 --- a/core/src/main/scala/spark/storage/BlockManagerMaster.scala +++ b/core/src/main/scala/spark/storage/BlockManagerMaster.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.storage import akka.actor.ActorRef @@ -6,6 +23,7 @@ import akka.pattern.ask import akka.util.Duration import spark.{Logging, SparkException} +import spark.storage.BlockManagerMessages._ private[spark] class BlockManagerMaster(var driverActor: ActorRef) extends Logging { diff --git a/core/src/main/scala/spark/storage/BlockManagerMasterActor.scala b/core/src/main/scala/spark/storage/BlockManagerMasterActor.scala index 0d4384ba1f..011bb6b83d 100644 --- a/core/src/main/scala/spark/storage/BlockManagerMasterActor.scala +++ b/core/src/main/scala/spark/storage/BlockManagerMasterActor.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.storage import java.util.{HashMap => JHashMap} @@ -12,6 +29,8 @@ import akka.util.Duration import akka.util.duration._ import spark.{Logging, Utils, SparkException} +import spark.storage.BlockManagerMessages._ + /** * BlockManagerMasterActor is an actor on the master node to track statuses of diff --git a/core/src/main/scala/spark/storage/BlockManagerMessages.scala b/core/src/main/scala/spark/storage/BlockManagerMessages.scala index 0010726c8d..9375a9ca54 100644 --- a/core/src/main/scala/spark/storage/BlockManagerMessages.scala +++ b/core/src/main/scala/spark/storage/BlockManagerMessages.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.storage import java.io.{Externalizable, ObjectInput, ObjectOutput} @@ -5,102 +22,89 @@ import java.io.{Externalizable, ObjectInput, ObjectOutput} import akka.actor.ActorRef -////////////////////////////////////////////////////////////////////////////////// -// Messages from the master to slaves. -////////////////////////////////////////////////////////////////////////////////// -private[spark] -sealed trait ToBlockManagerSlave - -// Remove a block from the slaves that have it. This can only be used to remove -// blocks that the master knows about. -private[spark] -case class RemoveBlock(blockId: String) extends ToBlockManagerSlave - -// Remove all blocks belonging to a specific RDD. -private[spark] case class RemoveRdd(rddId: Int) extends ToBlockManagerSlave - - -////////////////////////////////////////////////////////////////////////////////// -// Messages from slaves to the master. -////////////////////////////////////////////////////////////////////////////////// -private[spark] -sealed trait ToBlockManagerMaster - -private[spark] -case class RegisterBlockManager( - blockManagerId: BlockManagerId, - maxMemSize: Long, - sender: ActorRef) - extends ToBlockManagerMaster - -private[spark] -case class HeartBeat(blockManagerId: BlockManagerId) extends ToBlockManagerMaster - -private[spark] -class UpdateBlockInfo( - var blockManagerId: BlockManagerId, - var blockId: String, - var storageLevel: StorageLevel, - var memSize: Long, - var diskSize: Long) - extends ToBlockManagerMaster - with Externalizable { - - def this() = this(null, null, null, 0, 0) // For deserialization only - - override def writeExternal(out: ObjectOutput) { - blockManagerId.writeExternal(out) - out.writeUTF(blockId) - storageLevel.writeExternal(out) - out.writeLong(memSize) - out.writeLong(diskSize) - } - - override def readExternal(in: ObjectInput) { - blockManagerId = BlockManagerId(in) - blockId = in.readUTF() - storageLevel = StorageLevel(in) - memSize = in.readLong() - diskSize = in.readLong() - } -} - -private[spark] -object UpdateBlockInfo { - def apply(blockManagerId: BlockManagerId, - blockId: String, - storageLevel: StorageLevel, - memSize: Long, - diskSize: Long): UpdateBlockInfo = { - new UpdateBlockInfo(blockManagerId, blockId, storageLevel, memSize, diskSize) +private[storage] object BlockManagerMessages { + ////////////////////////////////////////////////////////////////////////////////// + // Messages from the master to slaves. + ////////////////////////////////////////////////////////////////////////////////// + sealed trait ToBlockManagerSlave + + // Remove a block from the slaves that have it. This can only be used to remove + // blocks that the master knows about. + case class RemoveBlock(blockId: String) extends ToBlockManagerSlave + + // Remove all blocks belonging to a specific RDD. + case class RemoveRdd(rddId: Int) extends ToBlockManagerSlave + + + ////////////////////////////////////////////////////////////////////////////////// + // Messages from slaves to the master. + ////////////////////////////////////////////////////////////////////////////////// + sealed trait ToBlockManagerMaster + + case class RegisterBlockManager( + blockManagerId: BlockManagerId, + maxMemSize: Long, + sender: ActorRef) + extends ToBlockManagerMaster + + case class HeartBeat(blockManagerId: BlockManagerId) extends ToBlockManagerMaster + + class UpdateBlockInfo( + var blockManagerId: BlockManagerId, + var blockId: String, + var storageLevel: StorageLevel, + var memSize: Long, + var diskSize: Long) + extends ToBlockManagerMaster + with Externalizable { + + def this() = this(null, null, null, 0, 0) // For deserialization only + + override def writeExternal(out: ObjectOutput) { + blockManagerId.writeExternal(out) + out.writeUTF(blockId) + storageLevel.writeExternal(out) + out.writeLong(memSize) + out.writeLong(diskSize) + } + + override def readExternal(in: ObjectInput) { + blockManagerId = BlockManagerId(in) + blockId = in.readUTF() + storageLevel = StorageLevel(in) + memSize = in.readLong() + diskSize = in.readLong() + } } - // For pattern-matching - def unapply(h: UpdateBlockInfo): Option[(BlockManagerId, String, StorageLevel, Long, Long)] = { - Some((h.blockManagerId, h.blockId, h.storageLevel, h.memSize, h.diskSize)) + object UpdateBlockInfo { + def apply(blockManagerId: BlockManagerId, + blockId: String, + storageLevel: StorageLevel, + memSize: Long, + diskSize: Long): UpdateBlockInfo = { + new UpdateBlockInfo(blockManagerId, blockId, storageLevel, memSize, diskSize) + } + + // For pattern-matching + def unapply(h: UpdateBlockInfo): Option[(BlockManagerId, String, StorageLevel, Long, Long)] = { + Some((h.blockManagerId, h.blockId, h.storageLevel, h.memSize, h.diskSize)) + } } -} -private[spark] -case class GetLocations(blockId: String) extends ToBlockManagerMaster + case class GetLocations(blockId: String) extends ToBlockManagerMaster -private[spark] -case class GetLocationsMultipleBlockIds(blockIds: Array[String]) extends ToBlockManagerMaster + case class GetLocationsMultipleBlockIds(blockIds: Array[String]) extends ToBlockManagerMaster -private[spark] -case class GetPeers(blockManagerId: BlockManagerId, size: Int) extends ToBlockManagerMaster + case class GetPeers(blockManagerId: BlockManagerId, size: Int) extends ToBlockManagerMaster -private[spark] -case class RemoveExecutor(execId: String) extends ToBlockManagerMaster + case class RemoveExecutor(execId: String) extends ToBlockManagerMaster -private[spark] -case object StopBlockManagerMaster extends ToBlockManagerMaster + case object StopBlockManagerMaster extends ToBlockManagerMaster -private[spark] -case object GetMemoryStatus extends ToBlockManagerMaster + case object GetMemoryStatus extends ToBlockManagerMaster -private[spark] -case object ExpireDeadHosts extends ToBlockManagerMaster + case object ExpireDeadHosts extends ToBlockManagerMaster -private[spark] -case object GetStorageStatus extends ToBlockManagerMaster + case object GetStorageStatus extends ToBlockManagerMaster +} diff --git a/core/src/main/scala/spark/storage/BlockManagerSlaveActor.scala b/core/src/main/scala/spark/storage/BlockManagerSlaveActor.scala index b264d1deb5..6e5fb43732 100644 --- a/core/src/main/scala/spark/storage/BlockManagerSlaveActor.scala +++ b/core/src/main/scala/spark/storage/BlockManagerSlaveActor.scala @@ -1,8 +1,25 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.storage import akka.actor.Actor -import spark.{Logging, SparkException, Utils} +import spark.storage.BlockManagerMessages._ /** diff --git a/core/src/main/scala/spark/storage/BlockManagerSource.scala b/core/src/main/scala/spark/storage/BlockManagerSource.scala new file mode 100644 index 0000000000..2aecd1ea71 --- /dev/null +++ b/core/src/main/scala/spark/storage/BlockManagerSource.scala @@ -0,0 +1,48 @@ +package spark.storage + +import com.codahale.metrics.{Gauge,MetricRegistry} + +import spark.metrics.source.Source + + +private[spark] class BlockManagerSource(val blockManager: BlockManager) extends Source { + val metricRegistry = new MetricRegistry() + val sourceName = "BlockManager" + + metricRegistry.register(MetricRegistry.name("memory", "maxMem", "MBytes"), new Gauge[Long] { + override def getValue: Long = { + val storageStatusList = blockManager.master.getStorageStatus + val maxMem = storageStatusList.map(_.maxMem).reduce(_ + _) + maxMem / 1024 / 1024 + } + }) + + metricRegistry.register(MetricRegistry.name("memory", "remainingMem", "MBytes"), new Gauge[Long] { + override def getValue: Long = { + val storageStatusList = blockManager.master.getStorageStatus + val remainingMem = storageStatusList.map(_.memRemaining).reduce(_ + _) + remainingMem / 1024 / 1024 + } + }) + + metricRegistry.register(MetricRegistry.name("memory", "memUsed", "MBytes"), new Gauge[Long] { + override def getValue: Long = { + val storageStatusList = blockManager.master.getStorageStatus + val maxMem = storageStatusList.map(_.maxMem).reduce(_ + _) + val remainingMem = storageStatusList.map(_.memRemaining).reduce(_ + _) + (maxMem - remainingMem) / 1024 / 1024 + } + }) + + metricRegistry.register(MetricRegistry.name("disk", "diskSpaceUsed", "MBytes"), new Gauge[Long] { + override def getValue: Long = { + val storageStatusList = blockManager.master.getStorageStatus + val diskSpaceUsed = storageStatusList + .flatMap(_.blocks.values.map(_.diskSize)) + .reduceOption(_ + _) + .getOrElse(0L) + + diskSpaceUsed / 1024 / 1024 + } + }) +} diff --git a/core/src/main/scala/spark/storage/BlockManagerWorker.scala b/core/src/main/scala/spark/storage/BlockManagerWorker.scala index 3057ade233..39064bce92 100644 --- a/core/src/main/scala/spark/storage/BlockManagerWorker.scala +++ b/core/src/main/scala/spark/storage/BlockManagerWorker.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.storage import java.nio.ByteBuffer diff --git a/core/src/main/scala/spark/storage/BlockMessage.scala b/core/src/main/scala/spark/storage/BlockMessage.scala index 30d7500e01..bcce26b7c1 100644 --- a/core/src/main/scala/spark/storage/BlockMessage.scala +++ b/core/src/main/scala/spark/storage/BlockMessage.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.storage import java.nio.ByteBuffer @@ -5,7 +22,6 @@ import java.nio.ByteBuffer import scala.collection.mutable.StringBuilder import scala.collection.mutable.ArrayBuffer -import spark._ import spark.network._ private[spark] case class GetBlock(id: String) diff --git a/core/src/main/scala/spark/storage/BlockMessageArray.scala b/core/src/main/scala/spark/storage/BlockMessageArray.scala index ee0c5ff9a2..ee2fc167d5 100644 --- a/core/src/main/scala/spark/storage/BlockMessageArray.scala +++ b/core/src/main/scala/spark/storage/BlockMessageArray.scala @@ -1,8 +1,24 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.storage import java.nio.ByteBuffer -import scala.collection.mutable.StringBuilder import scala.collection.mutable.ArrayBuffer import spark._ @@ -96,7 +112,7 @@ private[spark] object BlockMessageArray { def main(args: Array[String]) { val blockMessages = - (0 until 10).map(i => { + (0 until 10).map { i => if (i % 2 == 0) { val buffer = ByteBuffer.allocate(100) buffer.clear @@ -104,7 +120,7 @@ private[spark] object BlockMessageArray { } else { BlockMessage.fromGetBlock(GetBlock(i.toString)) } - }) + } val blockMessageArray = new BlockMessageArray(blockMessages) println("Block message array created") diff --git a/core/src/main/scala/spark/storage/BlockObjectWriter.scala b/core/src/main/scala/spark/storage/BlockObjectWriter.scala index 42e2b07d5c..3812009ca1 100644 --- a/core/src/main/scala/spark/storage/BlockObjectWriter.scala +++ b/core/src/main/scala/spark/storage/BlockObjectWriter.scala @@ -1,6 +1,21 @@ -package spark.storage +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ -import java.nio.ByteBuffer +package spark.storage /** diff --git a/core/src/main/scala/spark/storage/BlockStore.scala b/core/src/main/scala/spark/storage/BlockStore.scala index 8188d3595e..c8db0022b0 100644 --- a/core/src/main/scala/spark/storage/BlockStore.scala +++ b/core/src/main/scala/spark/storage/BlockStore.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.storage import java.nio.ByteBuffer diff --git a/core/src/main/scala/spark/storage/DiskStore.scala b/core/src/main/scala/spark/storage/DiskStore.scala index da859eebcb..3495d653bd 100644 --- a/core/src/main/scala/spark/storage/DiskStore.scala +++ b/core/src/main/scala/spark/storage/DiskStore.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.storage import java.io.{File, FileOutputStream, OutputStream, RandomAccessFile} diff --git a/core/src/main/scala/spark/storage/MemoryStore.scala b/core/src/main/scala/spark/storage/MemoryStore.scala index eba5ee507f..b5a86b85a7 100644 --- a/core/src/main/scala/spark/storage/MemoryStore.scala +++ b/core/src/main/scala/spark/storage/MemoryStore.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.storage import java.util.LinkedHashMap diff --git a/core/src/main/scala/spark/storage/PutResult.scala b/core/src/main/scala/spark/storage/PutResult.scala index 76f236057b..3a0974fe15 100644 --- a/core/src/main/scala/spark/storage/PutResult.scala +++ b/core/src/main/scala/spark/storage/PutResult.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.storage import java.nio.ByteBuffer diff --git a/core/src/main/scala/spark/storage/ShuffleBlockManager.scala b/core/src/main/scala/spark/storage/ShuffleBlockManager.scala index 44638e0c2d..8a7a6f9ed3 100644 --- a/core/src/main/scala/spark/storage/ShuffleBlockManager.scala +++ b/core/src/main/scala/spark/storage/ShuffleBlockManager.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.storage import spark.serializer.Serializer diff --git a/core/src/main/scala/spark/storage/StorageLevel.scala b/core/src/main/scala/spark/storage/StorageLevel.scala index cc0c354e7e..f52650988c 100644 --- a/core/src/main/scala/spark/storage/StorageLevel.scala +++ b/core/src/main/scala/spark/storage/StorageLevel.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.storage import java.io.{Externalizable, IOException, ObjectInput, ObjectOutput} diff --git a/core/src/main/scala/spark/storage/StorageUtils.scala b/core/src/main/scala/spark/storage/StorageUtils.scala index 950c0cdf35..2aeed4ea3c 100644 --- a/core/src/main/scala/spark/storage/StorageUtils.scala +++ b/core/src/main/scala/spark/storage/StorageUtils.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.storage import spark.{Utils, SparkContext} @@ -39,12 +56,19 @@ case class RDDInfo(id: Int, name: String, storageLevel: StorageLevel, private[spark] object StorageUtils { - /* Given the current storage status of the BlockManager, returns information for each RDD */ - def rddInfoFromStorageStatus(storageStatusList: Array[StorageStatus], + /* Returns RDD-level information, compiled from a list of StorageStatus objects */ + def rddInfoFromStorageStatus(storageStatusList: Seq[StorageStatus], sc: SparkContext) : Array[RDDInfo] = { rddInfoFromBlockStatusList(storageStatusList.flatMap(_.blocks).toMap, sc) } + /* Returns a map of blocks to their locations, compiled from a list of StorageStatus objects */ + def blockLocationsFromStorageStatus(storageStatusList: Seq[StorageStatus]) = { + val blockLocationPairs = storageStatusList + .flatMap(s => s.blocks.map(b => (b._1, s.blockManagerId.hostPort))) + blockLocationPairs.groupBy(_._1).map{case (k, v) => (k, v.unzip._2)}.toMap + } + /* Given a list of BlockStatus objets, returns information for each RDD */ def rddInfoFromBlockStatusList(infos: Map[String, BlockStatus], sc: SparkContext) : Array[RDDInfo] = { diff --git a/core/src/main/scala/spark/storage/ThreadingTest.scala b/core/src/main/scala/spark/storage/ThreadingTest.scala index 5c406e68cb..b3ab1ff4b4 100644 --- a/core/src/main/scala/spark/storage/ThreadingTest.scala +++ b/core/src/main/scala/spark/storage/ThreadingTest.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.storage import akka.actor._ diff --git a/core/src/main/scala/spark/ui/JettyUtils.scala b/core/src/main/scala/spark/ui/JettyUtils.scala index bc6f9c10d5..ca6088ad93 100644 --- a/core/src/main/scala/spark/ui/JettyUtils.scala +++ b/core/src/main/scala/spark/ui/JettyUtils.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.ui import annotation.tailrec diff --git a/core/src/main/scala/spark/ui/Page.scala b/core/src/main/scala/spark/ui/Page.scala index c853b44b76..03034a4520 100644 --- a/core/src/main/scala/spark/ui/Page.scala +++ b/core/src/main/scala/spark/ui/Page.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.ui -private[spark] object Page extends Enumeration { val Storage, Jobs = Value }
\ No newline at end of file +private[spark] object Page extends Enumeration { val Storage, Jobs, Environment, Executors = Value } diff --git a/core/src/main/scala/spark/ui/SparkUI.scala b/core/src/main/scala/spark/ui/SparkUI.scala index b3bdc2c490..7599f82a94 100644 --- a/core/src/main/scala/spark/ui/SparkUI.scala +++ b/core/src/main/scala/spark/ui/SparkUI.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.ui import javax.servlet.http.HttpServletRequest @@ -5,15 +22,14 @@ import javax.servlet.http.HttpServletRequest import org.eclipse.jetty.server.{Handler, Server} import spark.{Logging, SparkContext, Utils} +import spark.ui.env.EnvironmentUI +import spark.ui.exec.ExecutorsUI import spark.ui.storage.BlockManagerUI import spark.ui.jobs.JobProgressUI -import spark.ui.UIUtils._ import spark.ui.JettyUtils._ /** Top level user interface for Spark */ private[spark] class SparkUI(sc: SparkContext) extends Logging { - // TODO(pwendell): It would be nice to add a view that prints out environment information - val host = Utils.localHostName() val port = Option(System.getProperty("spark.ui.port")).getOrElse(SparkUI.DEFAULT_PORT).toInt var boundPort: Option[Int] = None @@ -25,7 +41,10 @@ private[spark] class SparkUI(sc: SparkContext) extends Logging { ) val storage = new BlockManagerUI(sc) val jobs = new JobProgressUI(sc) - val allHandlers = storage.getHandlers ++ jobs.getHandlers ++ handlers + val env = new EnvironmentUI(sc) + val exec = new ExecutorsUI(sc) + val allHandlers = storage.getHandlers ++ jobs.getHandlers ++ env.getHandlers ++ + exec.getHandlers ++ handlers /** Bind the HTTP server which backs this web interface */ def bind() { @@ -48,6 +67,7 @@ private[spark] class SparkUI(sc: SparkContext) extends Logging { // This server must register all handlers, including JobProgressUI, before binding // JobProgressUI registers a listener with SparkContext, which requires sc to initialize jobs.start() + exec.start() } def stop() { diff --git a/core/src/main/scala/spark/ui/UIUtils.scala b/core/src/main/scala/spark/ui/UIUtils.scala index 7b79290d1b..e33c80282a 100644 --- a/core/src/main/scala/spark/ui/UIUtils.scala +++ b/core/src/main/scala/spark/ui/UIUtils.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.ui import scala.xml.Node @@ -19,6 +36,14 @@ private[spark] object UIUtils { case Jobs => <li class="active"><a href="/stages">Jobs</a></li> case _ => <li><a href="/stages">Jobs</a></li> } + val environment = page match { + case Environment => <li class="active"><a href="/environment">Environment</a></li> + case _ => <li><a href="/environment">Environment</a></li> + } + val executors = page match { + case Executors => <li class="active"><a href="/executors">Executors</a></li> + case _ => <li><a href="/executors">Executors</a></li> + } <html> <head> @@ -27,7 +52,7 @@ private[spark] object UIUtils { <link rel="stylesheet" href="/static/webui.css" type="text/css" /> <link rel="stylesheet" href="/static/bootstrap-responsive.min.css" type="text/css" /> <script src="/static/sorttable.js"></script> - <title>{title}</title> + <title>{sc.appName} - {title}</title> <style type="text/css"> table.sortable thead {{ cursor: pointer; }} </style> @@ -44,6 +69,8 @@ private[spark] object UIUtils { <ul class="nav"> {storage} {jobs} + {environment} + {executors} </ul> <ul id="infolist"> <li>Application: <strong>{sc.appName}</strong></li> @@ -110,4 +137,4 @@ private[spark] object UIUtils { </tbody> </table> } -}
\ No newline at end of file +} diff --git a/core/src/main/scala/spark/ui/UIWorkloadGenerator.scala b/core/src/main/scala/spark/ui/UIWorkloadGenerator.scala index 840ac9773e..c738cc4065 100644 --- a/core/src/main/scala/spark/ui/UIWorkloadGenerator.scala +++ b/core/src/main/scala/spark/ui/UIWorkloadGenerator.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.ui import scala.util.Random @@ -32,6 +49,9 @@ private[spark] object UIWorkloadGenerator { // NOTE: Right now there is no easy way for us to show spark.job.annotation for a given phase, // but we pass it here anyways since it will be useful once we do. def setName(s: String) = { + if(schedulingMode == SchedulingMode.FAIR) { + sc.addLocalProperties("spark.scheduler.cluster.fair.pool",desc) + } sc.addLocalProperties("spark.job.annotation", s) } val baseData = sc.makeRDD(1 to NUM_PARTITIONS * 10, NUM_PARTITIONS) @@ -67,10 +87,7 @@ private[spark] object UIWorkloadGenerator { while (true) { for ((desc, job) <- jobs) { new Thread { - override def run() { - if(schedulingMode == SchedulingMode.FAIR) { - sc.addLocalProperties("spark.scheduler.cluster.fair.pool",desc) - } + override def run() { try { setName(desc) job() diff --git a/core/src/main/scala/spark/ui/env/EnvironmentUI.scala b/core/src/main/scala/spark/ui/env/EnvironmentUI.scala new file mode 100644 index 0000000000..5ae7935ed4 --- /dev/null +++ b/core/src/main/scala/spark/ui/env/EnvironmentUI.scala @@ -0,0 +1,86 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +package spark.ui.env + +import javax.servlet.http.HttpServletRequest + +import org.eclipse.jetty.server.Handler + +import scala.collection.JavaConversions._ +import scala.util.Properties + +import spark.ui.JettyUtils._ +import spark.ui.UIUtils.headerSparkPage +import spark.ui.Page.Environment +import spark.SparkContext +import spark.ui.UIUtils + +import scala.xml.Node + +private[spark] class EnvironmentUI(sc: SparkContext) { + + def getHandlers = Seq[(String, Handler)]( + ("/environment", (request: HttpServletRequest) => envDetails(request)) + ) + + def envDetails(request: HttpServletRequest): Seq[Node] = { + val jvmInformation = Seq( + ("Java Version", "%s (%s)".format(Properties.javaVersion, Properties.javaVendor)), + ("Java Home", Properties.javaHome), + ("Scala Version", Properties.versionString), + ("Scala Home", Properties.scalaHome) + ) + def jvmRow(kv: (String, String)) = <tr><td>{kv._1}</td><td>{kv._2}</td></tr> + def jvmTable = UIUtils.listingTable(Seq("Name", "Value"), jvmRow, jvmInformation) + + val properties = System.getProperties.iterator.toSeq + val classPathProperty = properties + .filter{case (k, v) => k.contains("java.class.path")} + .headOption + .getOrElse("", "") + val sparkProperties = properties.filter(_._1.startsWith("spark")) + val otherProperties = properties.diff(sparkProperties :+ classPathProperty) + + val propertyHeaders = Seq("Name", "Value") + def propertyRow(kv: (String, String)) = <tr><td>{kv._1}</td><td>{kv._2}</td></tr> + val sparkPropertyTable = UIUtils.listingTable(propertyHeaders, propertyRow, sparkProperties) + val otherPropertyTable = UIUtils.listingTable(propertyHeaders, propertyRow, otherProperties) + + val classPathEntries = classPathProperty._2 + .split(System.getProperty("path.separator", ":")) + .filterNot(e => e.isEmpty) + .map(e => (e, "System Classpath")) + val addedJars = sc.addedJars.iterator.toSeq.map{case (path, time) => (path, "Added By User")} + val addedFiles = sc.addedFiles.iterator.toSeq.map{case (path, time) => (path, "Added By User")} + val classPath = addedJars ++ addedFiles ++ classPathEntries + + val classPathHeaders = Seq("Resource", "Source") + def classPathRow(data: (String, String)) = <tr><td>{data._1}</td><td>{data._2}</td></tr> + val classPathTable = UIUtils.listingTable(classPathHeaders, classPathRow, classPath) + + val content = + <span> + <h2>Runtime Information</h2> {jvmTable} + <h2>Spark Properties</h2> {sparkPropertyTable} + <h2>System Properties</h2> {otherPropertyTable} + <h2>Classpath Entries</h2> {classPathTable} + </span> + + headerSparkPage(content, sc, "Environment", Environment) + } +} diff --git a/core/src/main/scala/spark/ui/exec/ExecutorsUI.scala b/core/src/main/scala/spark/ui/exec/ExecutorsUI.scala new file mode 100644 index 0000000000..4be2bfa413 --- /dev/null +++ b/core/src/main/scala/spark/ui/exec/ExecutorsUI.scala @@ -0,0 +1,153 @@ +package spark.ui.exec + + +import javax.servlet.http.HttpServletRequest + +import org.eclipse.jetty.server.Handler + +import scala.collection.mutable.{ArrayBuffer, HashMap, HashSet} +import scala.util.Properties + +import spark.{ExceptionFailure, Logging, SparkContext, Success, Utils} +import spark.executor.TaskMetrics +import spark.scheduler.cluster.TaskInfo +import spark.scheduler._ +import spark.SparkContext +import spark.storage.{StorageStatus, StorageUtils} +import spark.ui.JettyUtils._ +import spark.ui.Page.Executors +import spark.ui.UIUtils.headerSparkPage +import spark.ui.UIUtils + +import scala.xml.{Node, XML} + +private[spark] class ExecutorsUI(val sc: SparkContext) { + + private var _listener: Option[ExecutorsListener] = None + def listener = _listener.get + + def start() { + _listener = Some(new ExecutorsListener) + sc.addSparkListener(listener) + } + + def getHandlers = Seq[(String, Handler)]( + ("/executors", (request: HttpServletRequest) => render(request)) + ) + + def render(request: HttpServletRequest): Seq[Node] = { + val storageStatusList = sc.getExecutorStorageStatus + + val maxMem = storageStatusList.map(_.maxMem).reduce(_+_) + val memUsed = storageStatusList.map(_.memUsed()).reduce(_+_) + val diskSpaceUsed = storageStatusList.flatMap(_.blocks.values.map(_.diskSize)) + .reduceOption(_+_).getOrElse(0L) + + val execHead = Seq("Executor ID", "Address", "RDD blocks", "Memory used", "Disk used", + "Active tasks", "Failed tasks", "Complete tasks", "Total tasks") + def execRow(kv: Seq[String]) = + <tr> + <td>{kv(0)}</td> + <td>{kv(1)}</td> + <td>{kv(2)}</td> + <td sorttable_customkey={kv(3)}> + {Utils.memoryBytesToString(kv(3).toLong)} / {Utils.memoryBytesToString(kv(4).toLong)} + </td> + <td sorttable_customkey={kv(5)}> + {Utils.memoryBytesToString(kv(5).toLong)} + </td> + <td>{kv(6)}</td> + <td>{kv(7)}</td> + <td>{kv(8)}</td> + <td>{kv(9)}</td> + </tr> + val execInfo = + for (b <- 0 until storageStatusList.size) + yield getExecInfo(b) + val execTable = UIUtils.listingTable(execHead, execRow, execInfo) + + val content = + <div class="row"> + <div class="span12"> + <ul class="unstyled"> + <li><strong>Memory:</strong> + {Utils.memoryBytesToString(memUsed)} Used + ({Utils.memoryBytesToString(maxMem)} Total) </li> + <li><strong>Disk:</strong> {Utils.memoryBytesToString(diskSpaceUsed)} Used </li> + </ul> + </div> + </div> + <div class = "row"> + <div class="span12"> + {execTable} + </div> + </div>; + + headerSparkPage(content, sc, "Executors", Executors) + } + + def getExecInfo(a: Int): Seq[String] = { + val execId = sc.getExecutorStorageStatus(a).blockManagerId.executorId + val hostPort = sc.getExecutorStorageStatus(a).blockManagerId.hostPort + val rddBlocks = sc.getExecutorStorageStatus(a).blocks.size.toString + val memUsed = sc.getExecutorStorageStatus(a).memUsed().toString + val maxMem = sc.getExecutorStorageStatus(a).maxMem.toString + val diskUsed = sc.getExecutorStorageStatus(a).diskUsed().toString + val activeTasks = listener.executorToTasksActive.get(a.toString).map(l => l.size) + .getOrElse(0).toString + val failedTasks = listener.executorToTasksFailed.getOrElse(a.toString, 0).toString + val completedTasks = listener.executorToTasksComplete.getOrElse(a.toString, 0).toString + val totalTasks = listener.executorToTaskInfos(a.toString).size.toString + + Seq( + execId, + hostPort, + rddBlocks, + memUsed, + maxMem, + diskUsed, + activeTasks, + failedTasks, + completedTasks, + totalTasks + ) + } + + private[spark] class ExecutorsListener extends SparkListener with Logging { + val executorToTasksActive = HashMap[String, HashSet[TaskInfo]]() + val executorToTasksComplete = HashMap[String, Int]() + val executorToTasksFailed = HashMap[String, Int]() + val executorToTaskInfos = + HashMap[String, ArrayBuffer[(TaskInfo, Option[TaskMetrics], Option[ExceptionFailure])]]() + + override def onTaskStart(taskStart: SparkListenerTaskStart) { + val eid = taskStart.taskInfo.executorId + val activeTasks = executorToTasksActive.getOrElseUpdate(eid, new HashSet[TaskInfo]()) + activeTasks += taskStart.taskInfo + val taskList = executorToTaskInfos.getOrElse( + eid, ArrayBuffer[(TaskInfo, Option[TaskMetrics], Option[ExceptionFailure])]()) + taskList += ((taskStart.taskInfo, None, None)) + executorToTaskInfos(eid) = taskList + } + + override def onTaskEnd(taskEnd: SparkListenerTaskEnd) { + val eid = taskEnd.taskInfo.executorId + val activeTasks = executorToTasksActive.getOrElseUpdate(eid, new HashSet[TaskInfo]()) + activeTasks -= taskEnd.taskInfo + val (failureInfo, metrics): (Option[ExceptionFailure], Option[TaskMetrics]) = + taskEnd.reason match { + case e: ExceptionFailure => + executorToTasksFailed(eid) = executorToTasksFailed.getOrElse(eid, 0) + 1 + (Some(e), e.metrics) + case _ => + executorToTasksComplete(eid) = executorToTasksComplete.getOrElse(eid, 0) + 1 + (None, Option(taskEnd.taskMetrics)) + } + val taskList = executorToTaskInfos.getOrElse( + eid, ArrayBuffer[(TaskInfo, Option[TaskMetrics], Option[ExceptionFailure])]()) + taskList -= ((taskEnd.taskInfo, None, None)) + taskList += ((taskEnd.taskInfo, metrics, failureInfo)) + executorToTaskInfos(eid) = taskList + } + } +} diff --git a/core/src/main/scala/spark/ui/jobs/IndexPage.scala b/core/src/main/scala/spark/ui/jobs/IndexPage.scala index 04651e9c60..4ad787565d 100644 --- a/core/src/main/scala/spark/ui/jobs/IndexPage.scala +++ b/core/src/main/scala/spark/ui/jobs/IndexPage.scala @@ -1,18 +1,35 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.ui.jobs import java.util.Date import javax.servlet.http.HttpServletRequest -import scala.collection.mutable.HashMap -import scala.collection.mutable.HashSet import scala.Some import scala.xml.{NodeSeq, Node} +import spark.scheduler.cluster.TaskInfo import spark.scheduler.Stage import spark.storage.StorageLevel -import spark.ui.UIUtils._ import spark.ui.Page._ +import spark.ui.UIUtils._ +import spark.Utils /** Page showing list of all ongoing and recently finished stages and pools*/ private[spark] class IndexPage(parent: JobProgressUI) { @@ -22,23 +39,46 @@ private[spark] class IndexPage(parent: JobProgressUI) { val activeStages = listener.activeStages.toSeq val completedStages = listener.completedStages.reverse.toSeq val failedStages = listener.failedStages.reverse.toSeq + val now = System.currentTimeMillis() + + var activeTime = 0L + for (tasks <- listener.stageToTasksActive.values; t <- tasks) { + activeTime += t.timeRunning(now) + } val activeStagesTable = new StageTable(activeStages, parent) val completedStagesTable = new StageTable(completedStages, parent) val failedStagesTable = new StageTable(failedStages, parent) val poolTable = new PoolTable(parent.stagePagePoolSource, listener) + val summary: NodeSeq = + <div> + <ul class="unstyled"> + <li> + <strong>CPU time: </strong> + {parent.formatDuration(listener.totalTime + activeTime)} + </li> + {if (listener.totalShuffleRead > 0) + <li> + <strong>Shuffle read: </strong> + {Utils.memoryBytesToString(listener.totalShuffleRead)} + </li> + } + {if (listener.totalShuffleWrite > 0) + <li> + <strong>Shuffle write: </strong> + {Utils.memoryBytesToString(listener.totalShuffleWrite)} + </li> + } + <li><strong>Active Stages Number:</strong> {activeStages.size} </li> + <li><strong>Completed Stages Number:</strong> {completedStages.size} </li> + <li><strong>Failed Stages Number:</strong> {failedStages.size} </li> + <li><strong>Scheduling Mode:</strong> {parent.sc.getSchedulingMode}</li> + + </ul> + </div> - val content = <div class="row"> - <div class="span12"> - <ul class="unstyled"> - <li><strong>Active Stages Number:</strong> {activeStages.size} </li> - <li><strong>Completed Stages Number:</strong> {completedStages.size} </li> - <li><strong>Failed Stages Number:</strong> {failedStages.size} </li> - <li><strong>Scheduling Mode:</strong> {parent.sc.getSchedulingMode}</li> - </ul> - </div> - </div> ++ + val content = summary ++ <h3>Pools </h3> ++ poolTable.toNodeSeq ++ <h3>Active Stages : {activeStages.size}</h3> ++ activeStagesTable.toNodeSeq++ @@ -47,6 +87,6 @@ private[spark] class IndexPage(parent: JobProgressUI) { <h3>Failed Stages : {failedStages.size}</h3> ++ failedStagesTable.toNodeSeq - headerSparkPage(content, parent.sc, "Spark Stages/Pools", Jobs) + headerSparkPage(content, parent.sc, "Spark Stages", Jobs) } } diff --git a/core/src/main/scala/spark/ui/jobs/JobProgressListener.scala b/core/src/main/scala/spark/ui/jobs/JobProgressListener.scala index da767b3c0a..200e13cf99 100644 --- a/core/src/main/scala/spark/ui/jobs/JobProgressListener.scala +++ b/core/src/main/scala/spark/ui/jobs/JobProgressListener.scala @@ -21,6 +21,15 @@ private[spark] class JobProgressListener(val sc: SparkContext) extends SparkList val completedStages = ListBuffer[Stage]() val failedStages = ListBuffer[Stage]() + // Total metrics reflect metrics only for completed tasks + var totalTime = 0L + var totalShuffleRead = 0L + var totalShuffleWrite = 0L + + val stageToTime = HashMap[Int, Long]() + val stageToShuffleRead = HashMap[Int, Long]() + val stageToShuffleWrite = HashMap[Int, Long]() + val stageToTasksActive = HashMap[Int, HashSet[TaskInfo]]() val stageToTasksComplete = HashMap[Int, Int]() val stageToTasksFailed = HashMap[Int, Int]() val stageToTaskInfos = @@ -42,6 +51,12 @@ private[spark] class JobProgressListener(val sc: SparkContext) extends SparkList val toRemove = RETAINED_STAGES / 10 stages.takeRight(toRemove).foreach( s => { stageToTaskInfos.remove(s.id) + stageToTime.remove(s.id) + stageToShuffleRead.remove(s.id) + stageToShuffleWrite.remove(s.id) + stageToTasksActive.remove(s.id) + stageToTasksComplete.remove(s.id) + stageToTasksFailed.remove(s.id) }) stages.trimEnd(toRemove) } @@ -60,9 +75,21 @@ private[spark] class JobProgressListener(val sc: SparkContext) extends SparkList val stages = poolToActiveStages.getOrElseUpdate(poolName, new HashSet[Stage]()) stages += stage } - + + override def onTaskStart(taskStart: SparkListenerTaskStart) { + val sid = taskStart.task.stageId + val tasksActive = stageToTasksActive.getOrElseUpdate(sid, new HashSet[TaskInfo]()) + tasksActive += taskStart.taskInfo + val taskList = stageToTaskInfos.getOrElse( + sid, ArrayBuffer[(TaskInfo, Option[TaskMetrics], Option[ExceptionFailure])]()) + taskList += ((taskStart.taskInfo, None, None)) + stageToTaskInfos(sid) = taskList + } + override def onTaskEnd(taskEnd: SparkListenerTaskEnd) { val sid = taskEnd.task.stageId + val tasksActive = stageToTasksActive.getOrElseUpdate(sid, new HashSet[TaskInfo]()) + tasksActive -= taskEnd.taskInfo val (failureInfo, metrics): (Option[ExceptionFailure], Option[TaskMetrics]) = taskEnd.reason match { case e: ExceptionFailure => @@ -70,10 +97,29 @@ private[spark] class JobProgressListener(val sc: SparkContext) extends SparkList (Some(e), e.metrics) case _ => stageToTasksComplete(sid) = stageToTasksComplete.getOrElse(sid, 0) + 1 - (None, Some(taskEnd.taskMetrics)) + (None, Option(taskEnd.taskMetrics)) } + + stageToTime.getOrElseUpdate(sid, 0L) + val time = metrics.map(m => m.executorRunTime).getOrElse(0) + stageToTime(sid) += time + totalTime += time + + stageToShuffleRead.getOrElseUpdate(sid, 0L) + val shuffleRead = metrics.flatMap(m => m.shuffleReadMetrics).map(s => + s.remoteBytesRead).getOrElse(0L) + stageToShuffleRead(sid) += shuffleRead + totalShuffleRead += shuffleRead + + stageToShuffleWrite.getOrElseUpdate(sid, 0L) + val shuffleWrite = metrics.flatMap(m => m.shuffleWriteMetrics).map(s => + s.shuffleBytesWritten).getOrElse(0L) + stageToShuffleWrite(sid) += shuffleWrite + totalShuffleWrite += shuffleWrite + val taskList = stageToTaskInfos.getOrElse( sid, ArrayBuffer[(TaskInfo, Option[TaskMetrics], Option[ExceptionFailure])]()) + taskList -= ((taskEnd.taskInfo, None, None)) taskList += ((taskEnd.taskInfo, metrics, failureInfo)) stageToTaskInfos(sid) = taskList } diff --git a/core/src/main/scala/spark/ui/jobs/JobProgressUI.scala b/core/src/main/scala/spark/ui/jobs/JobProgressUI.scala index 5703b146df..3832c5d33c 100644 --- a/core/src/main/scala/spark/ui/jobs/JobProgressUI.scala +++ b/core/src/main/scala/spark/ui/jobs/JobProgressUI.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.ui.jobs import akka.util.Duration diff --git a/core/src/main/scala/spark/ui/jobs/StagePage.scala b/core/src/main/scala/spark/ui/jobs/StagePage.scala index 51b82b6a8c..e327cb3947 100644 --- a/core/src/main/scala/spark/ui/jobs/StagePage.scala +++ b/core/src/main/scala/spark/ui/jobs/StagePage.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.ui.jobs import java.util.Date @@ -20,23 +37,51 @@ private[spark] class StagePage(parent: JobProgressUI) { def render(request: HttpServletRequest): Seq[Node] = { val stageId = request.getParameter("id").toInt + val now = System.currentTimeMillis() if (!listener.stageToTaskInfos.contains(stageId)) { val content = <div> - <h2>Summary Metrics</h2> No tasks have finished yet - <h2>Tasks</h2> No tasks have finished yet + <h2>Summary Metrics</h2> No tasks have started yet + <h2>Tasks</h2> No tasks have started yet </div> return headerSparkPage(content, parent.sc, "Stage Details: %s".format(stageId), Jobs) } val tasks = listener.stageToTaskInfos(stageId) - val shuffleRead = listener.hasShuffleRead(stageId) - val shuffleWrite = listener.hasShuffleWrite(stageId) + val shuffleRead = listener.stageToShuffleRead(stageId) > 0 + val shuffleWrite = listener.stageToShuffleWrite(stageId) > 0 + + var activeTime = 0L + listener.stageToTasksActive(stageId).foreach { t => + activeTime += t.timeRunning(now) + } + + val summary = + <div> + <ul class="unstyled"> + <li> + <strong>CPU time: </strong> + {parent.formatDuration(listener.stageToTime(stageId) + activeTime)} + </li> + {if (shuffleRead) + <li> + <strong>Shuffle read: </strong> + {Utils.memoryBytesToString(listener.stageToShuffleRead(stageId))} + </li> + } + {if (shuffleWrite) + <li> + <strong>Shuffle write: </strong> + {Utils.memoryBytesToString(listener.stageToShuffleWrite(stageId))} + </li> + } + </ul> + </div> val taskHeaders: Seq[String] = - Seq("Task ID", "Duration", "Locality Level", "Worker", "Launch Time") ++ + Seq("Task ID", "Status", "Duration", "Locality Level", "Worker", "Launch Time") ++ {if (shuffleRead) Seq("Shuffle Read") else Nil} ++ {if (shuffleWrite) Seq("Shuffle Write") else Nil} ++ Seq("Details") @@ -44,7 +89,7 @@ private[spark] class StagePage(parent: JobProgressUI) { val taskTable = listingTable(taskHeaders, taskRow, tasks) // Excludes tasks which failed and have incomplete metrics - val validTasks = tasks.filter(t => Option(t._2).isDefined) + val validTasks = tasks.filter(t => t._1.status == "SUCCESS" && (Option(t._2).isDefined)) val summaryTable: Option[Seq[Node]] = if (validTasks.size == 0) { @@ -81,7 +126,8 @@ private[spark] class StagePage(parent: JobProgressUI) { } val content = - <h2>Summary Metrics</h2> ++ summaryTable.getOrElse(Nil) ++ <h2>Tasks</h2> ++ taskTable; + summary ++ <h2>Summary Metrics</h2> ++ summaryTable.getOrElse(Nil) ++ + <h2>Tasks</h2> ++ taskTable; headerSparkPage(content, parent.sc, "Stage Details: %s".format(stageId), Jobs) } @@ -91,10 +137,17 @@ private[spark] class StagePage(parent: JobProgressUI) { def fmtStackTrace(trace: Seq[StackTraceElement]): Seq[Node] = trace.map(e => <span style="display:block;">{e.toString}</span>) val (info, metrics, exception) = taskData + + val duration = if (info.status == "RUNNING") info.timeRunning(System.currentTimeMillis()) + else metrics.map(m => m.executorRunTime).getOrElse(1) + val formatDuration = if (info.status == "RUNNING") parent.formatDuration(duration) + else metrics.map(m => parent.formatDuration(m.executorRunTime)).getOrElse("") + <tr> <td>{info.taskId}</td> - <td sorttable_customkey={metrics.map{m => m.executorRunTime.toString}.getOrElse("1")}> - {metrics.map{m => parent.formatDuration(m.executorRunTime)}.getOrElse("")} + <td>{info.status}</td> + <td sorttable_customkey={duration.toString}> + {formatDuration} </td> <td>{info.taskLocality}</td> <td>{info.hostPort}</td> diff --git a/core/src/main/scala/spark/ui/jobs/StageTable.scala b/core/src/main/scala/spark/ui/jobs/StageTable.scala index 82fb0bd5cc..e18b70f0b9 100644 --- a/core/src/main/scala/spark/ui/jobs/StageTable.scala +++ b/core/src/main/scala/spark/ui/jobs/StageTable.scala @@ -8,10 +8,13 @@ import javax.servlet.http.HttpServletRequest import scala.Some import scala.xml.{NodeSeq, Node} import scala.collection.mutable.HashMap +import scala.collection.mutable.HashSet +import spark.scheduler.cluster.TaskInfo import spark.scheduler.Stage import spark.ui.UIUtils._ import spark.ui.Page._ +import spark.Utils import spark.storage.StorageLevel /* @@ -20,7 +23,7 @@ import spark.storage.StorageLevel private[spark] trait StagePoolInfo { def getStagePoolName(s: Stage): String - def hasHerf: Boolean + def hasHref: Boolean } /* @@ -29,7 +32,7 @@ private[spark] trait StagePoolInfo { private[spark] class FIFOStagePoolInfo extends StagePoolInfo { def getStagePoolName(s: Stage): String = "N/A" - def hasHerf: Boolean = false + def hasHref: Boolean = false } /* @@ -40,7 +43,7 @@ private[spark] class FairStagePoolInfo(listener: JobProgressListener) extends St listener.stageToPool(s) } - def hasHerf: Boolean = true + def hasHref: Boolean = true } /** Page showing list of all ongoing and recently finished stages */ @@ -58,7 +61,7 @@ private[spark] class StageTable( /** Special table which merges two header cells. */ def stageTable[T](makeRow: T => Seq[Node], rows: Seq[T]): Seq[Node] = { - <table class="table table-bordered table-striped table-condensed sortable"> + <table class="table table-bordered table-striped table-condensed sortable"> <thead> <th>Stage Id</th> <th>Pool Name</th> @@ -66,7 +69,8 @@ private[spark] class StageTable( <th>Submitted</th> <td>Duration</td> <td colspan="2">Tasks: Complete/Total</td> - <td>Shuffle Activity</td> + <td>Shuffle Read</td> + <td>Shuffle Write</td> <td>Stored RDD</td> </thead> <tbody> @@ -82,17 +86,14 @@ private[spark] class StageTable( } } - def makeProgressBar(completed: Int, total: Int): Seq[Node] = { - val width=130 - val height=15 - val completeWidth = (completed.toDouble / total) * width - - <svg width={width.toString} height={height.toString}> - <rect width={width.toString} height={height.toString} - fill="white" stroke="rgb(51,51,51)" stroke-width="1" /> - <rect width={completeWidth.toString} height={height.toString} - fill="rgb(0,136,204)" stroke="black" stroke-width="1" /> - </svg> + def makeProgressBar(started: Int, completed: Int, total: Int): Seq[Node] = { + val completeWidth = "width: %s%%".format((completed.toDouble/total)*100) + val startWidth = "width: %s%%".format((started.toDouble/total)*100) + + <div class="progress" style="height: 15px; margin-bottom: 0px"> + <div class="bar" style={completeWidth}></div> + <div class="bar bar-info" style={startWidth}></div> + </div> } @@ -101,13 +102,17 @@ private[spark] class StageTable( case Some(t) => dateFmt.format(new Date(t)) case None => "Unknown" } - val (read, write) = (listener.hasShuffleRead(s.id), listener.hasShuffleWrite(s.id)) - val shuffleInfo = (read, write) match { - case (true, true) => "Read/Write" - case (true, false) => "Read" - case (false, true) => "Write" - case _ => "" + + val shuffleRead = listener.stageToShuffleRead.getOrElse(s.id, 0L) match { + case 0 => "" + case b => Utils.memoryBytesToString(b) + } + val shuffleWrite = listener.stageToShuffleWrite.getOrElse(s.id, 0L) match { + case 0 => "" + case b => Utils.memoryBytesToString(b) } + + val startedTasks = listener.stageToTasksActive.getOrElse(s.id, HashSet[TaskInfo]()).size val completedTasks = listener.stageToTasksComplete.getOrElse(s.id, 0) val totalTasks = s.numPartitions @@ -115,23 +120,24 @@ private[spark] class StageTable( <tr> <td>{s.id}</td> - <td>{if (stagePoolInfo.hasHerf) { + <td>{if (stagePoolInfo.hasHref) { <a href={"/stages/pool?poolname=%s".format(poolName)}>{poolName}</a> } else { {poolName} }}</td> - <td><a href={"/stages/stage?id=%s".format(s.id)}>{s.origin}</a></td> + <td><a href={"/stages/stage?id=%s".format(s.id)}>{s.name}</a></td> <td>{submissionTime}</td> <td>{getElapsedTime(s.submissionTime, s.completionTime.getOrElse(System.currentTimeMillis()))}</td> - <td class="progress-cell">{makeProgressBar(completedTasks, totalTasks)}</td> + <td class="progress-cell">{makeProgressBar(startedTasks, completedTasks, totalTasks)}</td> <td style="border-left: 0; text-align: center;">{completedTasks} / {totalTasks} {listener.stageToTasksFailed.getOrElse(s.id, 0) match { case f if f > 0 => "(%s failed)".format(f) case _ => }} </td> - <td>{shuffleInfo}</td> + <td>{shuffleRead}</td> + <td>{shuffleWrite}</td> <td>{if (s.rdd.getStorageLevel != StorageLevel.NONE) { <a href={"/storage/rdd?id=%s".format(s.rdd.id)}> {Option(s.rdd.name).getOrElse(s.rdd.id)} diff --git a/core/src/main/scala/spark/ui/storage/BlockManagerUI.scala b/core/src/main/scala/spark/ui/storage/BlockManagerUI.scala index 3d67888e2c..49ed069c75 100644 --- a/core/src/main/scala/spark/ui/storage/BlockManagerUI.scala +++ b/core/src/main/scala/spark/ui/storage/BlockManagerUI.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.ui.storage import akka.util.Duration diff --git a/core/src/main/scala/spark/ui/storage/IndexPage.scala b/core/src/main/scala/spark/ui/storage/IndexPage.scala index d284134391..f76192eba8 100644 --- a/core/src/main/scala/spark/ui/storage/IndexPage.scala +++ b/core/src/main/scala/spark/ui/storage/IndexPage.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.ui.storage import javax.servlet.http.HttpServletRequest @@ -16,10 +33,6 @@ private[spark] class IndexPage(parent: BlockManagerUI) { def render(request: HttpServletRequest): Seq[Node] = { val storageStatusList = sc.getExecutorStorageStatus // Calculate macro-level statistics - val maxMem = storageStatusList.map(_.maxMem).reduce(_+_) - val remainingMem = storageStatusList.map(_.memRemaining).reduce(_+_) - val diskSpaceUsed = storageStatusList.flatMap(_.blocks.values.map(_.diskSize)) - .reduceOption(_+_).getOrElse(0L) val rddHeaders = Seq( "RDD Name", @@ -29,19 +42,7 @@ private[spark] class IndexPage(parent: BlockManagerUI) { "Size in Memory", "Size on Disk") val rdds = StorageUtils.rddInfoFromStorageStatus(storageStatusList, sc) - val rddTable = listingTable(rddHeaders, rddRow, rdds) - - val content = - <div class="row"> - <div class="span12"> - <ul class="unstyled"> - <li><strong>Memory:</strong> - {Utils.memoryBytesToString(maxMem - remainingMem)} Used - ({Utils.memoryBytesToString(remainingMem)} Available) </li> - <li><strong>Disk:</strong> {Utils.memoryBytesToString(diskSpaceUsed)} Used </li> - </ul> - </div> - </div> ++ {rddTable}; + val content = listingTable(rddHeaders, rddRow, rdds) headerSparkPage(content, parent.sc, "Spark Storage ", Storage) } diff --git a/core/src/main/scala/spark/ui/storage/RDDPage.scala b/core/src/main/scala/spark/ui/storage/RDDPage.scala index 0cb1e47ea5..003be54ad8 100644 --- a/core/src/main/scala/spark/ui/storage/RDDPage.scala +++ b/core/src/main/scala/spark/ui/storage/RDDPage.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.ui.storage import javax.servlet.http.HttpServletRequest @@ -26,8 +43,14 @@ private[spark] class RDDPage(parent: BlockManagerUI) { val workers = filteredStorageStatusList.map((prefix, _)) val workerTable = listingTable(workerHeaders, workerRow, workers) - val blockHeaders = Seq("Block Name", "Storage Level", "Size in Memory", "Size on Disk") - val blocks = filteredStorageStatusList.flatMap(_.blocks).toArray.sortWith(_._1 < _._1) + val blockHeaders = Seq("Block Name", "Storage Level", "Size in Memory", "Size on Disk", + "Locations") + + val blockStatuses = filteredStorageStatusList.flatMap(_.blocks).toArray.sortWith(_._1 < _._1) + val blockLocations = StorageUtils.blockLocationsFromStorageStatus(filteredStorageStatusList) + val blocks = blockStatuses.map { + case(id, status) => (id, status, blockLocations.get(id).getOrElse(Seq("UNKNOWN"))) + } val blockTable = listingTable(blockHeaders, blockRow, blocks) val content = @@ -74,8 +97,8 @@ private[spark] class RDDPage(parent: BlockManagerUI) { headerSparkPage(content, parent.sc, "RDD Info: " + rddInfo.name, Jobs) } - def blockRow(blk: (String, BlockStatus)): Seq[Node] = { - val (id, block) = blk + def blockRow(row: (String, BlockStatus, Seq[String])): Seq[Node] = { + val (id, block, locations) = row <tr> <td>{id}</td> <td> @@ -87,6 +110,9 @@ private[spark] class RDDPage(parent: BlockManagerUI) { <td sorttable_customkey={block.diskSize.toString}> {Utils.memoryBytesToString(block.diskSize)} </td> + <td> + {locations.map(l => <span>{l}<br/></span>)} + </td> </tr> } diff --git a/core/src/main/scala/spark/util/AkkaUtils.scala b/core/src/main/scala/spark/util/AkkaUtils.scala index 18fc9673bf..9233277bdb 100644 --- a/core/src/main/scala/spark/util/AkkaUtils.scala +++ b/core/src/main/scala/spark/util/AkkaUtils.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.util import akka.actor.{ActorSystem, ExtendedActorSystem} diff --git a/core/src/main/scala/spark/util/BoundedPriorityQueue.scala b/core/src/main/scala/spark/util/BoundedPriorityQueue.scala index 4bc5db8bb7..0575497f5d 100644 --- a/core/src/main/scala/spark/util/BoundedPriorityQueue.scala +++ b/core/src/main/scala/spark/util/BoundedPriorityQueue.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.util import java.io.Serializable diff --git a/core/src/main/scala/spark/util/ByteBufferInputStream.scala b/core/src/main/scala/spark/util/ByteBufferInputStream.scala index d7e67497fe..47a28e2f76 100644 --- a/core/src/main/scala/spark/util/ByteBufferInputStream.scala +++ b/core/src/main/scala/spark/util/ByteBufferInputStream.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.util import java.io.InputStream diff --git a/core/src/main/scala/spark/util/CompletionIterator.scala b/core/src/main/scala/spark/util/CompletionIterator.scala index 8139183780..210450892b 100644 --- a/core/src/main/scala/spark/util/CompletionIterator.scala +++ b/core/src/main/scala/spark/util/CompletionIterator.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.util /** @@ -22,4 +39,4 @@ object CompletionIterator { def completion() = completionFunction } } -}
\ No newline at end of file +} diff --git a/core/src/main/scala/spark/util/Distribution.scala b/core/src/main/scala/spark/util/Distribution.scala index 24738b4307..5d4d7a6c50 100644 --- a/core/src/main/scala/spark/util/Distribution.scala +++ b/core/src/main/scala/spark/util/Distribution.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.util import java.io.PrintStream @@ -62,4 +79,4 @@ object Distribution { quantiles.foreach{q => out.print(q + "\t")} out.println } -}
\ No newline at end of file +} diff --git a/core/src/main/scala/spark/util/IdGenerator.scala b/core/src/main/scala/spark/util/IdGenerator.scala index b6e309fe1a..3422280559 100644 --- a/core/src/main/scala/spark/util/IdGenerator.scala +++ b/core/src/main/scala/spark/util/IdGenerator.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.util import java.util.concurrent.atomic.AtomicInteger diff --git a/core/src/main/scala/spark/util/IntParam.scala b/core/src/main/scala/spark/util/IntParam.scala index 0427646747..daf0d58fa2 100644 --- a/core/src/main/scala/spark/util/IntParam.scala +++ b/core/src/main/scala/spark/util/IntParam.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.util /** diff --git a/core/src/main/scala/spark/util/MemoryParam.scala b/core/src/main/scala/spark/util/MemoryParam.scala index 3726738842..298562323a 100644 --- a/core/src/main/scala/spark/util/MemoryParam.scala +++ b/core/src/main/scala/spark/util/MemoryParam.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.util import spark.Utils diff --git a/core/src/main/scala/spark/util/MetadataCleaner.scala b/core/src/main/scala/spark/util/MetadataCleaner.scala index dafa906712..92909e0959 100644 --- a/core/src/main/scala/spark/util/MetadataCleaner.scala +++ b/core/src/main/scala/spark/util/MetadataCleaner.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.util import java.util.concurrent.{TimeUnit, ScheduledFuture, Executors} diff --git a/core/src/main/scala/spark/util/NextIterator.scala b/core/src/main/scala/spark/util/NextIterator.scala index 48b5018ddd..22163ece8d 100644 --- a/core/src/main/scala/spark/util/NextIterator.scala +++ b/core/src/main/scala/spark/util/NextIterator.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.util /** Provides a basic/boilerplate Iterator implementation. */ @@ -68,4 +85,4 @@ private[spark] abstract class NextIterator[U] extends Iterator[U] { gotNext = false nextValue } -}
\ No newline at end of file +} diff --git a/core/src/main/scala/spark/util/RateLimitedOutputStream.scala b/core/src/main/scala/spark/util/RateLimitedOutputStream.scala index e3f00ea8c7..00f782bbe7 100644 --- a/core/src/main/scala/spark/util/RateLimitedOutputStream.scala +++ b/core/src/main/scala/spark/util/RateLimitedOutputStream.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.util import scala.annotation.tailrec diff --git a/core/src/main/scala/spark/util/SerializableBuffer.scala b/core/src/main/scala/spark/util/SerializableBuffer.scala index 09d588fe1c..7e6842628a 100644 --- a/core/src/main/scala/spark/util/SerializableBuffer.scala +++ b/core/src/main/scala/spark/util/SerializableBuffer.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.util import java.nio.ByteBuffer diff --git a/core/src/main/scala/spark/util/StatCounter.scala b/core/src/main/scala/spark/util/StatCounter.scala index 2b980340b7..76358d4151 100644 --- a/core/src/main/scala/spark/util/StatCounter.scala +++ b/core/src/main/scala/spark/util/StatCounter.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.util /** diff --git a/core/src/main/scala/spark/util/TimeStampedHashMap.scala b/core/src/main/scala/spark/util/TimeStampedHashMap.scala index 92dfaa6e6f..07772a0afb 100644 --- a/core/src/main/scala/spark/util/TimeStampedHashMap.scala +++ b/core/src/main/scala/spark/util/TimeStampedHashMap.scala @@ -1,8 +1,26 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.util import java.util.concurrent.ConcurrentHashMap import scala.collection.JavaConversions import scala.collection.mutable.Map +import scala.collection.immutable import spark.scheduler.MapStatus /** @@ -82,6 +100,8 @@ class TimeStampedHashMap[A, B] extends Map[A, B]() with spark.Logging { } } + def toMap: immutable.Map[A, B] = iterator.toMap + /** * Removes old key-value pairs that have timestamp earlier than `threshTime` */ diff --git a/core/src/main/scala/spark/util/TimeStampedHashSet.scala b/core/src/main/scala/spark/util/TimeStampedHashSet.scala index 5f1cc93752..41e3fd8cba 100644 --- a/core/src/main/scala/spark/util/TimeStampedHashSet.scala +++ b/core/src/main/scala/spark/util/TimeStampedHashSet.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.util import scala.collection.mutable.Set diff --git a/core/src/main/scala/spark/util/Vector.scala b/core/src/main/scala/spark/util/Vector.scala index 835822edb2..ed49386f18 100644 --- a/core/src/main/scala/spark/util/Vector.scala +++ b/core/src/main/scala/spark/util/Vector.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.util class Vector(val elements: Array[Double]) extends Serializable { diff --git a/core/src/test/resources/fairscheduler.xml b/core/src/test/resources/fairscheduler.xml index 6e573b1883..996ffb1864 100644 --- a/core/src/test/resources/fairscheduler.xml +++ b/core/src/test/resources/fairscheduler.xml @@ -1,4 +1,21 @@ <?xml version="1.0"?> +<!-- + ~ Licensed to the Apache Software Foundation (ASF) under one or more + ~ contributor license agreements. See the NOTICE file distributed with + ~ this work for additional information regarding copyright ownership. + ~ The ASF licenses this file to You under the Apache License, Version 2.0 + ~ (the "License"); you may not use this file except in compliance with + ~ the License. You may obtain a copy of the License at + ~ + ~ http://www.apache.org/licenses/LICENSE-2.0 + ~ + ~ Unless required by applicable law or agreed to in writing, software + ~ distributed under the License is distributed on an "AS IS" BASIS, + ~ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + ~ See the License for the specific language governing permissions and + ~ limitations under the License. + --> + <allocations> <pool name="1"> <minShare>2</minShare> diff --git a/core/src/test/resources/log4j.properties b/core/src/test/resources/log4j.properties index d05cf3dec1..f6fef03689 100644 --- a/core/src/test/resources/log4j.properties +++ b/core/src/test/resources/log4j.properties @@ -1,4 +1,21 @@ -# Set everything to be logged to the file core/target/unit-tests.log +# +# Licensed to the Apache Software Foundation (ASF) under one or more +# contributor license agreements. See the NOTICE file distributed with +# this work for additional information regarding copyright ownership. +# The ASF licenses this file to You under the Apache License, Version 2.0 +# (the "License"); you may not use this file except in compliance with +# the License. You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +# + +# Set everything to be logged to the file core/target/unit-tests.log log4j.rootCategory=INFO, file log4j.appender.file=org.apache.log4j.FileAppender log4j.appender.file.append=false diff --git a/core/src/test/resources/test_metrics_config.properties b/core/src/test/resources/test_metrics_config.properties new file mode 100644 index 0000000000..2b31ddf2eb --- /dev/null +++ b/core/src/test/resources/test_metrics_config.properties @@ -0,0 +1,6 @@ +*.sink.console.period = 10 +*.sink.console.unit = seconds +*.source.jvm.class = spark.metrics.source.JvmSource +master.sink.console.period = 20 +master.sink.console.unit = minutes + diff --git a/core/src/test/resources/test_metrics_system.properties b/core/src/test/resources/test_metrics_system.properties new file mode 100644 index 0000000000..d5479f0298 --- /dev/null +++ b/core/src/test/resources/test_metrics_system.properties @@ -0,0 +1,7 @@ +*.sink.console.period = 10 +*.sink.console.unit = seconds +test.sink.console.class = spark.metrics.sink.ConsoleSink +test.sink.dummy.class = spark.metrics.sink.DummySink +test.source.dummy.class = spark.metrics.source.DummySource +test.sink.console.period = 20 +test.sink.console.unit = minutes diff --git a/core/src/test/scala/spark/AccumulatorSuite.scala b/core/src/test/scala/spark/AccumulatorSuite.scala index ac8ae7d308..0af175f316 100644 --- a/core/src/test/scala/spark/AccumulatorSuite.scala +++ b/core/src/test/scala/spark/AccumulatorSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import org.scalatest.FunSuite diff --git a/core/src/test/scala/spark/BroadcastSuite.scala b/core/src/test/scala/spark/BroadcastSuite.scala index 362a31fb0d..785721ece8 100644 --- a/core/src/test/scala/spark/BroadcastSuite.scala +++ b/core/src/test/scala/spark/BroadcastSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import org.scalatest.FunSuite diff --git a/core/src/test/scala/spark/CheckpointSuite.scala b/core/src/test/scala/spark/CheckpointSuite.scala index 28a7b21b92..a84c89e3c9 100644 --- a/core/src/test/scala/spark/CheckpointSuite.scala +++ b/core/src/test/scala/spark/CheckpointSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import org.scalatest.FunSuite diff --git a/core/src/test/scala/spark/ClosureCleanerSuite.scala b/core/src/test/scala/spark/ClosureCleanerSuite.scala index b2d0dd4627..7d2831e19c 100644 --- a/core/src/test/scala/spark/ClosureCleanerSuite.scala +++ b/core/src/test/scala/spark/ClosureCleanerSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import java.io.NotSerializableException diff --git a/core/src/test/scala/spark/DistributedSuite.scala b/core/src/test/scala/spark/DistributedSuite.scala index 0024ede828..e11efe459c 100644 --- a/core/src/test/scala/spark/DistributedSuite.scala +++ b/core/src/test/scala/spark/DistributedSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import network.ConnectionManagerId diff --git a/core/src/test/scala/spark/DriverSuite.scala b/core/src/test/scala/spark/DriverSuite.scala index 5e84b3a66a..ed16b9d8ef 100644 --- a/core/src/test/scala/spark/DriverSuite.scala +++ b/core/src/test/scala/spark/DriverSuite.scala @@ -1,7 +1,27 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import java.io.File +import org.apache.log4j.Logger +import org.apache.log4j.Level + import org.scalatest.FunSuite import org.scalatest.concurrent.Timeouts import org.scalatest.prop.TableDrivenPropertyChecks._ @@ -27,6 +47,7 @@ class DriverSuite extends FunSuite with Timeouts { */ object DriverWithoutCleanup { def main(args: Array[String]) { + Logger.getRootLogger().setLevel(Level.WARN) val sc = new SparkContext(args(0), "DriverWithoutCleanup") sc.parallelize(1 to 100, 4).count() } diff --git a/core/src/test/scala/spark/FailureSuite.scala b/core/src/test/scala/spark/FailureSuite.scala index 8c1445a465..5b133cdd6e 100644 --- a/core/src/test/scala/spark/FailureSuite.scala +++ b/core/src/test/scala/spark/FailureSuite.scala @@ -1,9 +1,23 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import org.scalatest.FunSuite -import org.scalatest.prop.Checkers - -import scala.collection.mutable.ArrayBuffer import SparkContext._ @@ -23,7 +37,7 @@ object FailureSuiteState { } class FailureSuite extends FunSuite with LocalSparkContext { - + // Run a 3-task map job in which task 1 deterministically fails once, and check // whether the job completes successfully and we ran 4 tasks in total. test("failure in a single-stage job") { @@ -49,7 +63,7 @@ class FailureSuite extends FunSuite with LocalSparkContext { test("failure in a two-stage job") { sc = new SparkContext("local[1,1]", "test") val results = sc.makeRDD(1 to 3).map(x => (x, x)).groupByKey(3).map { - case (k, v) => + case (k, v) => FailureSuiteState.synchronized { FailureSuiteState.tasksRun += 1 if (k == 1 && FailureSuiteState.tasksFailed == 0) { @@ -70,15 +84,43 @@ class FailureSuite extends FunSuite with LocalSparkContext { sc = new SparkContext("local[1,1]", "test") val results = sc.makeRDD(1 to 3).map(x => new NonSerializable) - val thrown = intercept[spark.SparkException] { + val thrown = intercept[SparkException] { results.collect() } - assert(thrown.getClass === classOf[spark.SparkException]) + assert(thrown.getClass === classOf[SparkException]) assert(thrown.getMessage.contains("NotSerializableException")) FailureSuiteState.clear() } + test("failure because task closure is not serializable") { + sc = new SparkContext("local[1,1]", "test") + val a = new NonSerializable + + // Non-serializable closure in the final result stage + val thrown = intercept[SparkException] { + sc.parallelize(1 to 10, 2).map(x => a).count() + } + assert(thrown.getClass === classOf[SparkException]) + assert(thrown.getMessage.contains("NotSerializableException")) + + // Non-serializable closure in an earlier stage + val thrown1 = intercept[SparkException] { + sc.parallelize(1 to 10, 2).map(x => (x, a)).partitionBy(new HashPartitioner(3)).count() + } + assert(thrown1.getClass === classOf[SparkException]) + assert(thrown1.getMessage.contains("NotSerializableException")) + + // Non-serializable closure in foreach function + val thrown2 = intercept[SparkException] { + sc.parallelize(1 to 10, 2).foreach(x => println(a)) + } + assert(thrown2.getClass === classOf[SparkException]) + assert(thrown2.getMessage.contains("NotSerializableException")) + + FailureSuiteState.clear() + } + // TODO: Need to add tests with shuffle fetch failures. } diff --git a/core/src/test/scala/spark/FileServerSuite.scala b/core/src/test/scala/spark/FileServerSuite.scala index f1a35bced3..242ae971f8 100644 --- a/core/src/test/scala/spark/FileServerSuite.scala +++ b/core/src/test/scala/spark/FileServerSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import com.google.common.io.Files @@ -35,26 +52,26 @@ class FileServerSuite extends FunSuite with LocalSparkContext { val testData = Array((1,1), (1,1), (2,1), (3,5), (2,2), (3,0)) val result = sc.parallelize(testData).reduceByKey { val path = SparkFiles.get("FileServerSuite.txt") - val in = new BufferedReader(new FileReader(path)) + val in = new BufferedReader(new FileReader(path)) val fileVal = in.readLine().toInt in.close() _ * fileVal + _ * fileVal - }.collect + }.collect() assert(result.toSet === Set((1,200), (2,300), (3,500))) } test("Distributing files locally using URL as input") { // addFile("file:///....") sc = new SparkContext("local[4]", "test") - sc.addFile((new File(tmpFile.toString)).toURL.toString) + sc.addFile(new File(tmpFile.toString).toURI.toString) val testData = Array((1,1), (1,1), (2,1), (3,5), (2,2), (3,0)) val result = sc.parallelize(testData).reduceByKey { val path = SparkFiles.get("FileServerSuite.txt") - val in = new BufferedReader(new FileReader(path)) + val in = new BufferedReader(new FileReader(path)) val fileVal = in.readLine().toInt in.close() _ * fileVal + _ * fileVal - }.collect + }.collect() assert(result.toSet === Set((1,200), (2,300), (3,500))) } @@ -80,12 +97,11 @@ class FileServerSuite extends FunSuite with LocalSparkContext { val testData = Array((1,1), (1,1), (2,1), (3,5), (2,2), (3,0)) val result = sc.parallelize(testData).reduceByKey { val path = SparkFiles.get("FileServerSuite.txt") - val in = new BufferedReader(new FileReader(path)) + val in = new BufferedReader(new FileReader(path)) val fileVal = in.readLine().toInt in.close() _ * fileVal + _ * fileVal - }.collect - println(result) + }.collect() assert(result.toSet === Set((1,200), (2,300), (3,500))) } diff --git a/core/src/test/scala/spark/FileSuite.scala b/core/src/test/scala/spark/FileSuite.scala index e61ff7793d..1e2c257c4b 100644 --- a/core/src/test/scala/spark/FileSuite.scala +++ b/core/src/test/scala/spark/FileSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import java.io.{FileWriter, PrintWriter, File} diff --git a/core/src/test/scala/spark/JavaAPISuite.java b/core/src/test/scala/spark/JavaAPISuite.java index d306124fca..5e2bf2d231 100644 --- a/core/src/test/scala/spark/JavaAPISuite.java +++ b/core/src/test/scala/spark/JavaAPISuite.java @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark; import java.io.File; @@ -309,7 +326,9 @@ public class JavaAPISuite implements Serializable { Assert.assertEquals(20/6.0, rdd.mean(), 0.01); Assert.assertEquals(20/6.0, rdd.mean(), 0.01); Assert.assertEquals(6.22222, rdd.variance(), 0.01); + Assert.assertEquals(7.46667, rdd.sampleVariance(), 0.01); Assert.assertEquals(2.49444, rdd.stdev(), 0.01); + Assert.assertEquals(2.73252, rdd.sampleStdev(), 0.01); Double first = rdd.first(); List<Double> take = rdd.take(5); diff --git a/core/src/test/scala/spark/KryoSerializerSuite.scala b/core/src/test/scala/spark/KryoSerializerSuite.scala index 327e2ff848..30d2d5282b 100644 --- a/core/src/test/scala/spark/KryoSerializerSuite.scala +++ b/core/src/test/scala/spark/KryoSerializerSuite.scala @@ -1,13 +1,27 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import scala.collection.mutable -import scala.collection.immutable import org.scalatest.FunSuite import com.esotericsoftware.kryo._ -import SparkContext._ - class KryoSerializerSuite extends FunSuite { test("basic types") { val ser = (new KryoSerializer).newInstance() @@ -36,6 +50,7 @@ class KryoSerializerSuite extends FunSuite { check(Array(true, false, true)) check(Array('a', 'b', 'c')) check(Array[Int]()) + check(Array(Array("1", "2"), Array("1", "2", "3", "4"))) } test("pairs") { @@ -86,7 +101,7 @@ class KryoSerializerSuite extends FunSuite { } test("custom registrator") { - import spark.test._ + import KryoTest._ System.setProperty("spark.kryo.registrator", classOf[MyRegistrator].getName) val ser = (new KryoSerializer).newInstance() @@ -106,14 +121,14 @@ class KryoSerializerSuite extends FunSuite { val hashMap = new java.util.HashMap[String, String] hashMap.put("foo", "bar") check(hashMap) - + System.clearProperty("spark.kryo.registrator") } } -package test { +object KryoTest { case class CaseClass(i: Int, s: String) {} - + class ClassWithNoArgConstructor { var x: Int = 0 override def equals(other: Any) = other match { @@ -137,4 +152,4 @@ package test { k.register(classOf[java.util.HashMap[_, _]]) } } -} +}
\ No newline at end of file diff --git a/core/src/test/scala/spark/LocalSparkContext.scala b/core/src/test/scala/spark/LocalSparkContext.scala index 76d5258b02..ddc212d290 100644 --- a/core/src/test/scala/spark/LocalSparkContext.scala +++ b/core/src/test/scala/spark/LocalSparkContext.scala @@ -1,13 +1,39 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import org.scalatest.Suite import org.scalatest.BeforeAndAfterEach +import org.scalatest.BeforeAndAfterAll + +import org.jboss.netty.logging.InternalLoggerFactory +import org.jboss.netty.logging.Slf4JLoggerFactory /** Manages a local `sc` {@link SparkContext} variable, correctly stopping it after each test. */ -trait LocalSparkContext extends BeforeAndAfterEach { self: Suite => +trait LocalSparkContext extends BeforeAndAfterEach with BeforeAndAfterAll { self: Suite => @transient var sc: SparkContext = _ + override def beforeAll() { + InternalLoggerFactory.setDefaultFactory(new Slf4JLoggerFactory()); + super.beforeAll() + } + override def afterEach() { resetSparkContext() super.afterEach() diff --git a/core/src/test/scala/spark/MapOutputTrackerSuite.scala b/core/src/test/scala/spark/MapOutputTrackerSuite.scala index 6e585e1c3a..ce6cec0451 100644 --- a/core/src/test/scala/spark/MapOutputTrackerSuite.scala +++ b/core/src/test/scala/spark/MapOutputTrackerSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import org.scalatest.FunSuite diff --git a/core/src/test/scala/spark/PairRDDFunctionsSuite.scala b/core/src/test/scala/spark/PairRDDFunctionsSuite.scala index 682d2745bf..b102eaf4e6 100644 --- a/core/src/test/scala/spark/PairRDDFunctionsSuite.scala +++ b/core/src/test/scala/spark/PairRDDFunctionsSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import scala.collection.mutable.ArrayBuffer diff --git a/core/src/test/scala/spark/PartitionPruningRDDSuite.scala b/core/src/test/scala/spark/PartitionPruningRDDSuite.scala new file mode 100644 index 0000000000..88352b639f --- /dev/null +++ b/core/src/test/scala/spark/PartitionPruningRDDSuite.scala @@ -0,0 +1,28 @@ +package spark + +import org.scalatest.FunSuite +import spark.SparkContext._ +import spark.rdd.PartitionPruningRDD + + +class PartitionPruningRDDSuite extends FunSuite with SharedSparkContext { + + test("Pruned Partitions inherit locality prefs correctly") { + class TestPartition(i: Int) extends Partition { + def index = i + } + val rdd = new RDD[Int](sc, Nil) { + override protected def getPartitions = { + Array[Partition]( + new TestPartition(1), + new TestPartition(2), + new TestPartition(3)) + } + def compute(split: Partition, context: TaskContext) = {Iterator()} + } + val prunedRDD = PartitionPruningRDD.create(rdd, {x => if (x==2) true else false}) + val p = prunedRDD.partitions(0) + assert(p.index == 2) + assert(prunedRDD.partitions.length == 1) + } +} diff --git a/core/src/test/scala/spark/PartitioningSuite.scala b/core/src/test/scala/spark/PartitioningSuite.scala index 99e433e3bd..b1e0b2b4d0 100644 --- a/core/src/test/scala/spark/PartitioningSuite.scala +++ b/core/src/test/scala/spark/PartitioningSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import org.scalatest.FunSuite diff --git a/core/src/test/scala/spark/PipedRDDSuite.scala b/core/src/test/scala/spark/PipedRDDSuite.scala index 1c9ca50811..35c04710a3 100644 --- a/core/src/test/scala/spark/PipedRDDSuite.scala +++ b/core/src/test/scala/spark/PipedRDDSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import org.scalatest.FunSuite @@ -67,7 +84,7 @@ class PipedRDDSuite extends FunSuite with SharedSparkContext { test("pipe with non-zero exit status") { val nums = sc.makeRDD(Array(1, 2, 3, 4), 2) - val piped = nums.pipe("cat nonexistent_file") + val piped = nums.pipe(Seq("cat nonexistent_file", "2>", "/dev/null")) intercept[SparkException] { piped.collect() } diff --git a/core/src/test/scala/spark/RDDSuite.scala b/core/src/test/scala/spark/RDDSuite.scala index e41ae385c0..cbddf4e523 100644 --- a/core/src/test/scala/spark/RDDSuite.scala +++ b/core/src/test/scala/spark/RDDSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import scala.collection.mutable.HashMap @@ -240,7 +257,7 @@ class RDDSuite extends FunSuite with SharedSparkContext { val ints = sc.makeRDD(scala.util.Random.shuffle(nums), 2) val topK = ints.top(5) assert(topK.size === 5) - assert(topK.sorted === nums.sorted.takeRight(5)) + assert(topK === nums.reverse.take(5)) } test("top with custom ordering") { @@ -252,6 +269,24 @@ class RDDSuite extends FunSuite with SharedSparkContext { assert(topK.sorted === Array("b", "a")) } + test("takeOrdered with predefined ordering") { + val nums = Array(1, 2, 3, 4, 5, 6, 7, 8, 9, 10) + val rdd = sc.makeRDD(nums, 2) + val sortedLowerK = rdd.takeOrdered(5) + assert(sortedLowerK.size === 5) + assert(sortedLowerK === Array(1, 2, 3, 4, 5)) + } + + test("takeOrdered with custom ordering") { + val nums = Array(1, 2, 3, 4, 5, 6, 7, 8, 9, 10) + implicit val ord = implicitly[Ordering[Int]].reverse + val rdd = sc.makeRDD(nums, 2) + val sortedTopK = rdd.takeOrdered(5) + assert(sortedTopK.size === 5) + assert(sortedTopK === Array(10, 9, 8, 7, 6)) + assert(sortedTopK === nums.sorted(ord).take(5)) + } + test("takeSample") { val data = sc.parallelize(1 to 100, 2) for (seed <- 1 to 5) { @@ -284,4 +319,10 @@ class RDDSuite extends FunSuite with SharedSparkContext { assert(sample.toSet.size < 100, "sampling with replacement returned all distinct elements") } } + + test("runJob on an invalid partition") { + intercept[IllegalArgumentException] { + sc.runJob(sc.parallelize(1 to 10, 2), {iter: Iterator[Int] => iter.size}, Seq(0, 1, 2), false) + } + } } diff --git a/core/src/test/scala/spark/SharedSparkContext.scala b/core/src/test/scala/spark/SharedSparkContext.scala index 1da79f9824..70c24515be 100644 --- a/core/src/test/scala/spark/SharedSparkContext.scala +++ b/core/src/test/scala/spark/SharedSparkContext.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import org.scalatest.Suite diff --git a/core/src/test/scala/spark/ShuffleNettySuite.scala b/core/src/test/scala/spark/ShuffleNettySuite.scala index bfaffa953e..6bad6c1d13 100644 --- a/core/src/test/scala/spark/ShuffleNettySuite.scala +++ b/core/src/test/scala/spark/ShuffleNettySuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import org.scalatest.BeforeAndAfterAll diff --git a/core/src/test/scala/spark/ShuffleSuite.scala b/core/src/test/scala/spark/ShuffleSuite.scala index 950218fa28..752e4b85e6 100644 --- a/core/src/test/scala/spark/ShuffleSuite.scala +++ b/core/src/test/scala/spark/ShuffleSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import scala.collection.mutable.ArrayBuffer @@ -16,9 +33,9 @@ import spark.rdd.ShuffledRDD import spark.SparkContext._ class ShuffleSuite extends FunSuite with ShouldMatchers with LocalSparkContext { - test("groupByKey with compression") { + test("groupByKey without compression") { try { - System.setProperty("spark.shuffle.compress", "true") + System.setProperty("spark.shuffle.compress", "false") sc = new SparkContext("local", "test") val pairs = sc.parallelize(Array((1, 1), (1, 2), (1, 3), (2, 1)), 4) val groups = pairs.groupByKey(4).collect() @@ -28,7 +45,7 @@ class ShuffleSuite extends FunSuite with ShouldMatchers with LocalSparkContext { val valuesFor2 = groups.find(_._1 == 2).get._2 assert(valuesFor2.toList.sorted === List(1)) } finally { - System.setProperty("spark.blockManager.compress", "false") + System.setProperty("spark.shuffle.compress", "true") } } diff --git a/core/src/test/scala/spark/SizeEstimatorSuite.scala b/core/src/test/scala/spark/SizeEstimatorSuite.scala index b5c8525f91..1ef812dfbd 100644 --- a/core/src/test/scala/spark/SizeEstimatorSuite.scala +++ b/core/src/test/scala/spark/SizeEstimatorSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import org.scalatest.FunSuite diff --git a/core/src/test/scala/spark/SortingSuite.scala b/core/src/test/scala/spark/SortingSuite.scala index f7bf207c68..b933c4aab8 100644 --- a/core/src/test/scala/spark/SortingSuite.scala +++ b/core/src/test/scala/spark/SortingSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import org.scalatest.FunSuite diff --git a/core/src/test/scala/spark/SparkContextInfoSuite.scala b/core/src/test/scala/spark/SparkContextInfoSuite.scala new file mode 100644 index 0000000000..6d50bf5e1b --- /dev/null +++ b/core/src/test/scala/spark/SparkContextInfoSuite.scala @@ -0,0 +1,60 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +package spark + +import org.scalatest.FunSuite +import spark.SparkContext._ + +class SparkContextInfoSuite extends FunSuite with LocalSparkContext { + test("getPersistentRDDs only returns RDDs that are marked as cached") { + sc = new SparkContext("local", "test") + assert(sc.getPersistentRDDs.isEmpty === true) + + val rdd = sc.makeRDD(Array(1, 2, 3, 4), 2) + assert(sc.getPersistentRDDs.isEmpty === true) + + rdd.cache() + assert(sc.getPersistentRDDs.size === 1) + assert(sc.getPersistentRDDs.values.head === rdd) + } + + test("getPersistentRDDs returns an immutable map") { + sc = new SparkContext("local", "test") + val rdd1 = sc.makeRDD(Array(1, 2, 3, 4), 2).cache() + + val myRdds = sc.getPersistentRDDs + assert(myRdds.size === 1) + assert(myRdds.values.head === rdd1) + + val rdd2 = sc.makeRDD(Array(5, 6, 7, 8), 1).cache() + + // getPersistentRDDs should have 2 RDDs, but myRdds should not change + assert(sc.getPersistentRDDs.size === 2) + assert(myRdds.size === 1) + } + + test("getRDDStorageInfo only reports on RDDs that actually persist data") { + sc = new SparkContext("local", "test") + val rdd = sc.makeRDD(Array(1, 2, 3, 4), 2).cache() + + assert(sc.getRDDStorageInfo.size === 0) + + rdd.collect() + assert(sc.getRDDStorageInfo.size === 1) + } +}
\ No newline at end of file diff --git a/core/src/test/scala/spark/ThreadingSuite.scala b/core/src/test/scala/spark/ThreadingSuite.scala index ff315b6693..f2acd0bd3c 100644 --- a/core/src/test/scala/spark/ThreadingSuite.scala +++ b/core/src/test/scala/spark/ThreadingSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import java.util.concurrent.Semaphore diff --git a/core/src/test/scala/spark/UnpersistSuite.scala b/core/src/test/scala/spark/UnpersistSuite.scala index 94776e7572..93977d16f4 100644 --- a/core/src/test/scala/spark/UnpersistSuite.scala +++ b/core/src/test/scala/spark/UnpersistSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import org.scalatest.FunSuite diff --git a/core/src/test/scala/spark/UtilsSuite.scala b/core/src/test/scala/spark/UtilsSuite.scala index 4a113e16bf..31c3b25c50 100644 --- a/core/src/test/scala/spark/UtilsSuite.scala +++ b/core/src/test/scala/spark/UtilsSuite.scala @@ -1,7 +1,27 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark +import com.google.common.base.Charsets +import com.google.common.io.Files +import java.io.{ByteArrayOutputStream, ByteArrayInputStream, FileOutputStream, File} import org.scalatest.FunSuite -import java.io.{ByteArrayOutputStream, ByteArrayInputStream} +import org.apache.commons.io.FileUtils import scala.util.Random class UtilsSuite extends FunSuite { @@ -71,5 +91,49 @@ class UtilsSuite extends FunSuite { assert(Utils.splitCommandString("''") === Seq("")) assert(Utils.splitCommandString("\"\"") === Seq("")) } + + test("string formatting of time durations") { + val second = 1000 + val minute = second * 60 + val hour = minute * 60 + def str = Utils.msDurationToString(_) + + assert(str(123) === "123 ms") + assert(str(second) === "1.0 s") + assert(str(second + 462) === "1.5 s") + assert(str(hour) === "1.00 h") + assert(str(minute) === "1.0 m") + assert(str(minute + 4 * second + 34) === "1.1 m") + assert(str(10 * hour + minute + 4 * second) === "10.02 h") + assert(str(10 * hour + 59 * minute + 59 * second + 999) === "11.00 h") + } + + test("reading offset bytes of a file") { + val tmpDir2 = Files.createTempDir() + val f1Path = tmpDir2 + "/f1" + val f1 = new FileOutputStream(f1Path) + f1.write("1\n2\n3\n4\n5\n6\n7\n8\n9\n".getBytes(Charsets.UTF_8)) + f1.close() + + // Read first few bytes + assert(Utils.offsetBytes(f1Path, 0, 5) === "1\n2\n3") + + // Read some middle bytes + assert(Utils.offsetBytes(f1Path, 4, 11) === "3\n4\n5\n6") + + // Read last few bytes + assert(Utils.offsetBytes(f1Path, 12, 18) === "7\n8\n9\n") + + // Read some nonexistent bytes in the beginning + assert(Utils.offsetBytes(f1Path, -5, 5) === "1\n2\n3") + + // Read some nonexistent bytes at the end + assert(Utils.offsetBytes(f1Path, 12, 22) === "7\n8\n9\n") + + // Read some nonexistent bytes on both ends + assert(Utils.offsetBytes(f1Path, -3, 25) === "1\n2\n3\n4\n5\n6\n7\n8\n9\n") + + FileUtils.deleteDirectory(tmpDir2) + } } diff --git a/core/src/test/scala/spark/ZippedPartitionsSuite.scala b/core/src/test/scala/spark/ZippedPartitionsSuite.scala index 96cb295f45..5e6d7b09d8 100644 --- a/core/src/test/scala/spark/ZippedPartitionsSuite.scala +++ b/core/src/test/scala/spark/ZippedPartitionsSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import scala.collection.immutable.NumericRange diff --git a/core/src/test/scala/spark/metrics/MetricsConfigSuite.scala b/core/src/test/scala/spark/metrics/MetricsConfigSuite.scala new file mode 100644 index 0000000000..87cd2ffad2 --- /dev/null +++ b/core/src/test/scala/spark/metrics/MetricsConfigSuite.scala @@ -0,0 +1,64 @@ +package spark.metrics + +import java.util.Properties +import java.io.{File, FileOutputStream} + +import org.scalatest.{BeforeAndAfter, FunSuite} + +import spark.metrics._ + +class MetricsConfigSuite extends FunSuite with BeforeAndAfter { + var filePath: String = _ + + before { + filePath = getClass.getClassLoader.getResource("test_metrics_config.properties").getFile() + } + + test("MetricsConfig with default properties") { + val conf = new MetricsConfig(Option("dummy-file")) + conf.initialize() + + assert(conf.properties.size() === 0) + assert(conf.properties.getProperty("test-for-dummy") === null) + + val property = conf.getInstance("random") + assert(property.size() === 0) + } + + test("MetricsConfig with properties set") { + val conf = new MetricsConfig(Option(filePath)) + conf.initialize() + + val masterProp = conf.getInstance("master") + assert(masterProp.size() === 3) + assert(masterProp.getProperty("sink.console.period") === "20") + assert(masterProp.getProperty("sink.console.unit") === "minutes") + assert(masterProp.getProperty("source.jvm.class") === "spark.metrics.source.JvmSource") + + val workerProp = conf.getInstance("worker") + assert(workerProp.size() === 3) + assert(workerProp.getProperty("sink.console.period") === "10") + assert(workerProp.getProperty("sink.console.unit") === "seconds") + assert(masterProp.getProperty("source.jvm.class") === "spark.metrics.source.JvmSource") + } + + test("MetricsConfig with subProperties") { + val conf = new MetricsConfig(Option(filePath)) + conf.initialize() + + val propCategories = conf.propertyCategories + assert(propCategories.size === 2) + + val masterProp = conf.getInstance("master") + val sourceProps = conf.subProperties(masterProp, MetricsSystem.SOURCE_REGEX) + assert(sourceProps.size === 1) + assert(sourceProps("jvm").getProperty("class") === "spark.metrics.source.JvmSource") + + val sinkProps = conf.subProperties(masterProp, MetricsSystem.SINK_REGEX) + assert(sinkProps.size === 1) + assert(sinkProps.contains("console")) + + val consoleProps = sinkProps("console") + assert(consoleProps.size() === 2) + } +} diff --git a/core/src/test/scala/spark/metrics/MetricsSystemSuite.scala b/core/src/test/scala/spark/metrics/MetricsSystemSuite.scala new file mode 100644 index 0000000000..c189996417 --- /dev/null +++ b/core/src/test/scala/spark/metrics/MetricsSystemSuite.scala @@ -0,0 +1,39 @@ +package spark.metrics + +import java.util.Properties +import java.io.{File, FileOutputStream} + +import org.scalatest.{BeforeAndAfter, FunSuite} + +import spark.metrics._ + +class MetricsSystemSuite extends FunSuite with BeforeAndAfter { + var filePath: String = _ + + before { + filePath = getClass.getClassLoader.getResource("test_metrics_system.properties").getFile() + System.setProperty("spark.metrics.conf", filePath) + } + + test("MetricsSystem with default config") { + val metricsSystem = MetricsSystem.createMetricsSystem("default") + val sources = metricsSystem.sources + val sinks = metricsSystem.sinks + + assert(sources.length === 0) + assert(sinks.length === 0) + } + + test("MetricsSystem with sources add") { + val metricsSystem = MetricsSystem.createMetricsSystem("test") + val sources = metricsSystem.sources + val sinks = metricsSystem.sinks + + assert(sources.length === 0) + assert(sinks.length === 1) + + val source = new spark.deploy.master.MasterSource(null) + metricsSystem.registerSource(source) + assert(sources.length === 1) + } +} diff --git a/core/src/test/scala/spark/rdd/JdbcRDDSuite.scala b/core/src/test/scala/spark/rdd/JdbcRDDSuite.scala index 6afb0fa9bc..dc8ca941c1 100644 --- a/core/src/test/scala/spark/rdd/JdbcRDDSuite.scala +++ b/core/src/test/scala/spark/rdd/JdbcRDDSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark import org.scalatest.{ BeforeAndAfter, FunSuite } diff --git a/core/src/test/scala/spark/rdd/ParallelCollectionSplitSuite.scala b/core/src/test/scala/spark/rdd/ParallelCollectionSplitSuite.scala index d27a2538e4..d1276d541f 100644 --- a/core/src/test/scala/spark/rdd/ParallelCollectionSplitSuite.scala +++ b/core/src/test/scala/spark/rdd/ParallelCollectionSplitSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.rdd import scala.collection.immutable.NumericRange diff --git a/core/src/test/scala/spark/scheduler/ClusterSchedulerSuite.scala b/core/src/test/scala/spark/scheduler/ClusterSchedulerSuite.scala index 8e1ad27e14..05afcd6567 100644 --- a/core/src/test/scala/spark/scheduler/ClusterSchedulerSuite.scala +++ b/core/src/test/scala/spark/scheduler/ClusterSchedulerSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler import org.scalatest.FunSuite @@ -16,7 +33,7 @@ class DummyTaskSetManager( initNumTasks: Int, clusterScheduler: ClusterScheduler, taskSet: TaskSet) - extends ClusterTaskSetManager(clusterScheduler,taskSet) { + extends ClusterTaskSetManager(clusterScheduler, taskSet) { parent = null weight = 1 diff --git a/core/src/test/scala/spark/scheduler/DAGSchedulerSuite.scala b/core/src/test/scala/spark/scheduler/DAGSchedulerSuite.scala index da72bfbf89..caaf3209fd 100644 --- a/core/src/test/scala/spark/scheduler/DAGSchedulerSuite.scala +++ b/core/src/test/scala/spark/scheduler/DAGSchedulerSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler import scala.collection.mutable.{Map, HashMap} @@ -6,21 +23,14 @@ import org.scalatest.FunSuite import org.scalatest.BeforeAndAfter import spark.LocalSparkContext - -import spark.storage.BlockManager -import spark.storage.BlockManagerId -import spark.storage.BlockManagerMaster -import spark.{Dependency, ShuffleDependency, OneToOneDependency} -import spark.FetchFailedException import spark.MapOutputTracker import spark.RDD import spark.SparkContext -import spark.SparkException import spark.Partition import spark.TaskContext -import spark.TaskEndReason - -import spark.{FetchFailed, Success} +import spark.{Dependency, ShuffleDependency, OneToOneDependency} +import spark.{FetchFailed, Success, TaskEndReason} +import spark.storage.{BlockManagerId, BlockManagerMaster} import spark.scheduler.cluster.Pool import spark.scheduler.cluster.SchedulingMode diff --git a/core/src/test/scala/spark/scheduler/JobLoggerSuite.scala b/core/src/test/scala/spark/scheduler/JobLoggerSuite.scala index 328e7e7529..bb9e715f95 100644 --- a/core/src/test/scala/spark/scheduler/JobLoggerSuite.scala +++ b/core/src/test/scala/spark/scheduler/JobLoggerSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler import java.util.Properties @@ -37,8 +54,8 @@ class JobLoggerSuite extends FunSuite with LocalSparkContext with ShouldMatchers val parentRdd = makeRdd(4, Nil) val shuffleDep = new ShuffleDependency(parentRdd, null) val rootRdd = makeRdd(4, List(shuffleDep)) - val shuffleMapStage = new Stage(1, parentRdd, Some(shuffleDep), Nil, jobID) - val rootStage = new Stage(0, rootRdd, None, List(shuffleMapStage), jobID) + val shuffleMapStage = new Stage(1, parentRdd, Some(shuffleDep), Nil, jobID, None) + val rootStage = new Stage(0, rootRdd, None, List(shuffleMapStage), jobID, None) joblogger.onStageSubmitted(SparkListenerStageSubmitted(rootStage, 4, null)) joblogger.getRddNameTest(parentRdd) should be (parentRdd.getClass.getName) diff --git a/core/src/test/scala/spark/scheduler/LocalSchedulerSuite.scala b/core/src/test/scala/spark/scheduler/LocalSchedulerSuite.scala index 8bd813fd14..14bb58731b 100644 --- a/core/src/test/scala/spark/scheduler/LocalSchedulerSuite.scala +++ b/core/src/test/scala/spark/scheduler/LocalSchedulerSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler import org.scalatest.FunSuite diff --git a/core/src/test/scala/spark/scheduler/SparkListenerSuite.scala b/core/src/test/scala/spark/scheduler/SparkListenerSuite.scala index 48aa67c543..392d67d67b 100644 --- a/core/src/test/scala/spark/scheduler/SparkListenerSuite.scala +++ b/core/src/test/scala/spark/scheduler/SparkListenerSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler import org.scalatest.FunSuite diff --git a/core/src/test/scala/spark/scheduler/TaskContextSuite.scala b/core/src/test/scala/spark/scheduler/TaskContextSuite.scala index 647bcaf860..95a6eee2fc 100644 --- a/core/src/test/scala/spark/scheduler/TaskContextSuite.scala +++ b/core/src/test/scala/spark/scheduler/TaskContextSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.scheduler import org.scalatest.FunSuite diff --git a/core/src/test/scala/spark/storage/BlockManagerSuite.scala b/core/src/test/scala/spark/storage/BlockManagerSuite.scala index b9d5f9668e..b719d65342 100644 --- a/core/src/test/scala/spark/storage/BlockManagerSuite.scala +++ b/core/src/test/scala/spark/storage/BlockManagerSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.storage import java.nio.ByteBuffer diff --git a/core/src/test/scala/spark/ui/UISuite.scala b/core/src/test/scala/spark/ui/UISuite.scala index e4bb3abc33..56c1fed6ad 100644 --- a/core/src/test/scala/spark/ui/UISuite.scala +++ b/core/src/test/scala/spark/ui/UISuite.scala @@ -1,14 +1,26 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.ui +import scala.util.{Failure, Success, Try} +import java.net.ServerSocket import org.scalatest.FunSuite import org.eclipse.jetty.server.Server -import java.net.ServerSocket -import scala.util.{Failure, Success, Try} -import spark.Utils -import com.google.common.io.Files -import org.apache.commons.io.FileUtils -import java.io.{FileOutputStream, File} -import com.google.common.base.Charsets class UISuite extends FunSuite { test("jetty port increases under contention") { @@ -31,47 +43,4 @@ class UISuite extends FunSuite { case Failure (e) => } } - - test("string formatting of time durations") { - val second = 1000 - val minute = second * 60 - val hour = minute * 60 - def str = Utils.msDurationToString(_) - - assert(str(123) === "123 ms") - assert(str(second) === "1.0 s") - assert(str(second + 462) === "1.5 s") - assert(str(hour) === "1.00 h") - assert(str(minute) === "1.0 m") - assert(str(minute + 4 * second + 34) === "1.1 m") - assert(str(10 * hour + minute + 4 * second) === "10.02 h") - assert(str(10 * hour + 59 * minute + 59 * second + 999) === "11.00 h") - } - - test("reading last n bytes of a file") { - val tmpDir = Files.createTempDir() - - // File smaller than limit - val f1Path = tmpDir + "/f1" - val f1 = new FileOutputStream(f1Path) - f1.write("a\nb\nc\nd".getBytes(Charsets.UTF_8)) - f1.close() - assert(Utils.lastNBytes(f1Path, 1024) === "a\nb\nc\nd") - - // File larger than limit - val f2Path = tmpDir + "/f2" - val f2 = new FileOutputStream(f2Path) - f2.write("1\n2\n3\n4\n5\n6\n7\n8\n".getBytes(Charsets.UTF_8)) - f2.close() - assert(Utils.lastNBytes(f2Path, 8) === "5\n6\n7\n8\n") - - // Request limit too - val f3Path = tmpDir + "/f2" - val f3 = new FileOutputStream(f3Path) - f3.write("1\n2\n3\n4\n5\n6\n7\n8\n".getBytes(Charsets.UTF_8)) - f3.close() - assert(Utils.lastNBytes(f3Path, 8) === "5\n6\n7\n8\n") - - FileUtils.deleteDirectory(tmpDir) - } } diff --git a/core/src/test/scala/spark/util/DistributionSuite.scala b/core/src/test/scala/spark/util/DistributionSuite.scala index cc6249b1dd..6578b55e82 100644 --- a/core/src/test/scala/spark/util/DistributionSuite.scala +++ b/core/src/test/scala/spark/util/DistributionSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.util import org.scalatest.FunSuite diff --git a/core/src/test/scala/spark/util/NextIteratorSuite.scala b/core/src/test/scala/spark/util/NextIteratorSuite.scala index ed5b36da73..fdbd43d941 100644 --- a/core/src/test/scala/spark/util/NextIteratorSuite.scala +++ b/core/src/test/scala/spark/util/NextIteratorSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.util import org.scalatest.FunSuite diff --git a/core/src/test/scala/spark/util/RateLimitedOutputStreamSuite.scala b/core/src/test/scala/spark/util/RateLimitedOutputStreamSuite.scala index 794063fb6d..4c0044202f 100644 --- a/core/src/test/scala/spark/util/RateLimitedOutputStreamSuite.scala +++ b/core/src/test/scala/spark/util/RateLimitedOutputStreamSuite.scala @@ -1,3 +1,20 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + package spark.util import org.scalatest.FunSuite |