Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | Merge pull request #516 from squito/fix_local_metrics | Matei Zaharia | 2013-03-15 | 8 | -12/+104 |
|\ | | | | | Fix local metrics | ||||
| * | increase sleep time | Imran Rashid | 2013-03-10 | 1 | -1/+1 |
| | | |||||
| * | add a small wait to one task to make sure some task runtime really is non-zero | Imran Rashid | 2013-03-10 | 1 | -4/+10 |
| | | |||||
| * | enable task metrics in local mode, add tests | Imran Rashid | 2013-03-09 | 2 | -2/+88 |
| | | |||||
| * | rename remoteFetchWaitTime to fetchWaitTime, since it also includes time ↵ | Imran Rashid | 2013-03-09 | 6 | -10/+10 |
| | | | | | | | | from local fetches | ||||
* | | Add a log4j compile dependency to fix build in IntelliJ | Mikhail Bautin | 2013-03-15 | 1 | -1/+5 |
| | | | | | | | | | | Also rename parent project to spark-parent (otherwise it shows up as "parent" in IntelliJ, which is very confusing). | ||||
* | | Merge pull request #521 from stephenh/earlyclose | Matei Zaharia | 2013-03-13 | 4 | -51/+153 |
|\ \ | | | | | | | Close the reader in HadoopRDD as soon as iteration end. | ||||
| * | | Add a test for NextIterator. | Stephen Haberman | 2013-03-13 | 1 | -0/+68 |
| | | | |||||
| * | | Add NextIterator.closeIfNeeded. | Stephen Haberman | 2013-03-13 | 2 | -2/+16 |
| | | | |||||
| * | | Remove NextIterator.close default implementation. | Stephen Haberman | 2013-03-12 | 2 | -4/+7 |
| | | | |||||
| * | | More quickly call close in HadoopRDD. | Stephen Haberman | 2013-03-11 | 3 | -52/+69 |
| | | | | | | | | | | | | | | | This also refactors out the common "gotNext" iterator pattern into a shared utility class. | ||||
* | | | Send block sizes as longs. | Charles Reiss | 2013-03-11 | 1 | -4/+4 |
|/ / | |||||
* | | Merge remote-tracking branch 'woggling/dag-sched-driver-port' | Matei Zaharia | 2013-03-10 | 1 | -5/+4 |
|\ \ | | | | | | | | | | | | | Conflicts: core/src/test/scala/spark/scheduler/DAGSchedulerSuite.scala | ||||
| * | | Prevent DAGSchedulerSuite from corrupting driver.port. | Charles Reiss | 2013-03-09 | 1 | -4/+5 |
| |/ | | | | | | | | | Use the LocalSparkContext abstraction to properly manage clearing spark.driver.port. | ||||
* | | Merge pull request #512 from patelh/fix-kryo-serializer | Matei Zaharia | 2013-03-10 | 2 | -11/+19 |
|\ \ | | | | | | | Fix reference bug in Kryo serializer, add test, update version | ||||
| * | | Fix reference bug in Kryo serializer, add test, update version | Hiral Patel | 2013-03-07 | 2 | -11/+19 |
| |/ | |||||
* | | Merge pull request #515 from woggling/deploy-app-death | Matei Zaharia | 2013-03-10 | 3 | -7/+17 |
|\ \ | | | | | | | Notify standalone deploy client of application death. | ||||
| * | | Notify standalone deploy client of application death. | Charles Reiss | 2013-03-09 | 3 | -7/+17 |
| |/ | | | | | | | | | | | | | | | | | Usually, this isn't necessary since the application will be removed as a result of the deploy client disconnecting, but occassionally, the standalone deploy master removes an application otherwise. Also mark applications as FAILED instead of FINISHED when they are killed as a result of their executors failing too many times. | ||||
* | | Merge remote-tracking branch 'stephenh/nomocks' | Matei Zaharia | 2013-03-10 | 9 | -536/+275 |
|\ \ | |/ |/| | | | | | | | Conflicts: core/src/main/scala/spark/storage/BlockManagerMaster.scala core/src/test/scala/spark/scheduler/DAGSchedulerSuite.scala | ||||
| * | Fix MapOutputTrackerSuite. | Stephen Haberman | 2013-02-26 | 1 | -2/+4 |
| | | |||||
| * | Override DAGScheduler.runLocally so we can remove the Thread.sleep. | Stephen Haberman | 2013-02-25 | 2 | -19/+27 |
| | | |||||
| * | Merge branch 'master' into nomocks | Stephen Haberman | 2013-02-25 | 95 | -890/+1411 |
| |\ | | | | | | | | | | | | | Conflicts: core/src/test/scala/spark/scheduler/DAGSchedulerSuite.scala | ||||
| * | | Use stubs instead of mocks for DAGSchedulerSuite. | Stephen Haberman | 2013-02-09 | 8 | -527/+253 |
| | | | |||||
* | | | Fix TaskMetrics not being serializable | Matei Zaharia | 2013-03-04 | 1 | -6/+13 |
| | | | |||||
* | | | Merge pull request #506 from rxin/spark-706 | Matei Zaharia | 2013-03-03 | 2 | -51/+120 |
|\ \ \ | | | | | | | | | Fixed SPARK-706: Failures in block manager put leads to read task hanging. | ||||
| * | | | Fixed SPARK-706: Failures in block manager put leads to read task | Reynold Xin | 2013-02-28 | 2 | -51/+120 |
| | | | | | | | | | | | | | | | | hanging. | ||||
* | | | | minor cleanup based on feedback in review request | Imran Rashid | 2013-03-03 | 4 | -24/+24 |
| | | | | |||||
* | | | | change CleanupIterator to CompletionIterator | Imran Rashid | 2013-03-03 | 3 | -27/+27 |
| | | | | |||||
* | | | | refactoring of TaskMetrics | Imran Rashid | 2013-03-03 | 6 | -59/+110 |
| | | | | |||||
* | | | | Merge branch 'master' into stageInfo | Imran Rashid | 2013-03-03 | 29 | -107/+474 |
|\ \ \ \ | |||||
| * \ \ \ | Merge pull request #504 from mosharaf/master | Matei Zaharia | 2013-03-02 | 2 | -2/+2 |
| |\ \ \ \ | | | | | | | | | | | | | Worker address was getting removed when removing an app. | ||||
| | * | | | | Fixed master datastructure updates after removing an application; and a typo. | Mosharaf Chowdhury | 2013-02-27 | 2 | -2/+2 |
| | | | | | | |||||
| * | | | | | bump version to 0.7.1-SNAPSHOT in the subproject poms to keep the maven ↵ | Mark Hamstra | 2013-02-28 | 1 | -1/+1 |
| |/ / / / | | | | | | | | | | | | | | | | build building. | ||||
| * | | | | Fix a problem with no hosts being counted as alive in the first job | Matei Zaharia | 2013-02-26 | 1 | -3/+3 |
| | | | | | |||||
| * | | | | Fix overly large thread names in PySpark | Matei Zaharia | 2013-02-26 | 1 | -2/+2 |
| | | | | | |||||
| * | | | | Fixed replication bug in BlockManager | Tathagata Das | 2013-02-25 | 2 | -3/+17 |
| | | | | | |||||
| * | | | | Allow passing sparkHome and JARs to StreamingContext constructor | Matei Zaharia | 2013-02-25 | 2 | -3/+3 |
| | | | | | | | | | | | | | | | | | | | | | | | | | Also warns if spark.cleaner.ttl is not set in the version where you pass your own SparkContext. | ||||
| * | | | | Set spark.deploy.spreadOut to true by default in 0.7 (improves locality) | Matei Zaharia | 2013-02-25 | 1 | -1/+1 |
| | | | | | |||||
| * | | | | Add a config property for Akka lifecycle event logging | Matei Zaharia | 2013-02-25 | 1 | -2/+4 |
| | | | | | |||||
| * | | | | Merge pull request #498 from pwendell/shutup-akka | Matei Zaharia | 2013-02-25 | 1 | -1/+1 |
| |\ \ \ \ | | | | | | | | | | | | | Disable remote lifecycle logging from Akka. | ||||
| | * | | | | Disable remote lifecycle logging from Akka. | Patrick Wendell | 2013-02-25 | 1 | -1/+1 |
| | | | | | | | | | | | | | | | | | | | | | | | | This changes the default setting to `off` for remote lifecycle events. When this is on, it is very chatty at the INFO level. It also prints out several ERROR messages sometimes when sc.stop() is called. | ||||
| * | | | | | Get spark.default.paralellism on each call to defaultPartitioner, | Matei Zaharia | 2013-02-25 | 1 | -4/+1 |
| | |_|_|/ | |/| | | | | | | | | | | | | | instead of only once, in case the user changes it across Spark uses | ||||
| * | | | | Merge pull request #459 from stephenh/bettersplits | Matei Zaharia | 2013-02-25 | 8 | -38/+90 |
| |\ \ \ \ | | |/ / / | |/| | | | Change defaultPartitioner to use upstream split size. | ||||
| | * | | | Use default parallelism if its set. | Stephen Haberman | 2013-02-24 | 2 | -6/+19 |
| | | | | | |||||
| | * | | | Merge branch 'master' into bettersplits | Stephen Haberman | 2013-02-24 | 92 | -793/+1260 |
| | |\ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Conflicts: core/src/main/scala/spark/RDD.scala core/src/main/scala/spark/scheduler/cluster/StandaloneSchedulerBackend.scala core/src/test/scala/spark/ShuffleSuite.scala | ||||
| | * | | | | Update more javadocs. | Stephen Haberman | 2013-02-16 | 2 | -15/+17 |
| | | | | | | |||||
| | * | | | | Tweak test names. | Stephen Haberman | 2013-02-16 | 1 | -2/+2 |
| | | | | | | |||||
| | * | | | | Remove fileServerSuite.txt. | Stephen Haberman | 2013-02-16 | 1 | -1/+0 |
| | | | | | | |||||
| | * | | | | Update default.parallelism docs, have StandaloneSchedulerBackend use it. | Stephen Haberman | 2013-02-16 | 7 | -24/+39 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Only brand new RDDs (e.g. parallelize and makeRDD) now use default parallelism, everything else uses their largest parent's partitioner or partition size. | ||||
| | * | | | | Change defaultPartitioner to use upstream split size. | Stephen Haberman | 2013-02-10 | 3 | -6/+29 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Previously it used the SparkContext.defaultParallelism, which occassionally ended up being a very bad guess. Looking at upstream RDDs seems to make better use of the context. Also sorted the upstream RDDs by partition size first, as if we have a hugely-partitioned RDD and tiny-partitioned RDD, it is unlikely we want the resulting RDD to be tiny-partitioned. |