aboutsummaryrefslogtreecommitdiff
path: root/core/src/main/scala/spark/deploy/worker/ExecutorRunner.scala
blob: 8b51ff1c3a4d352762634a6e88ccdf4b9932ceb2 (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
/*
 * Licensed to the Apache Software Foundation (ASF) under one or more
 * contributor license agreements.  See the NOTICE file distributed with
 * this work for additional information regarding copyright ownership.
 * The ASF licenses this file to You under the Apache License, Version 2.0
 * (the "License"); you may not use this file except in compliance with
 * the License.  You may obtain a copy of the License at
 *
 *    http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR 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

/**
 * Manages the execution of one executor process.
 */
private[spark] class ExecutorRunner(
    val appId: String,
    val execId: Int,
    val appDesc: ApplicationDescription,
    val cores: Int,
    val memory: Int,
    val worker: ActorRef,
    val workerId: String,
    val hostPort: String,
    val sparkHome: File,
    val workDir: File)
  extends Logging {

  Utils.checkHostPort(hostPort, "Expected hostport")

  val fullId = appId + "/" + execId
  var workerThread: Thread = null
  var process: Process = null
  var shutdownHook: Thread = null

  def start() {
    workerThread = new Thread("ExecutorRunner for " + fullId) {
      override def run() { fetchAndRunExecutor() }
    }
    workerThread.start()

    // Shutdown hook that kills actors on shutdown.
    shutdownHook = new Thread() {
      override def run() {
        if (process != null) {
          logInfo("Shutdown hook killing child process.")
          process.destroy()
          process.waitFor()
        }
      }
    }
    Runtime.getRuntime.addShutdownHook(shutdownHook)
  }

  /** Stop this executor runner, including killing the process it launched */
  def kill() {
    if (workerThread != null) {
      workerThread.interrupt()
      workerThread = null
      if (process != null) {
        logInfo("Killing process!")
        process.destroy()
        process.waitFor()
      }
      worker ! ExecutorStateChanged(appId, execId, ExecutorState.KILLED, None, None)
      Runtime.getRuntime.removeShutdownHook(shutdownHook)
    }
  }

  /** Replace variables such as {{EXECUTOR_ID}} and {{CORES}} in a command argument passed to us */
  def substituteVariables(argument: String): String = argument match {
    case "{{EXECUTOR_ID}}" => execId.toString
    case "{{HOSTNAME}}" => Utils.parseHostPort(hostPort)._1
    case "{{CORES}}" => cores.toString
    case other => other
  }

  def buildCommandSeq(): Seq[String] = {
    val command = appDesc.command
    val runner = Option(getenv("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) ++
      command.arguments.map(substituteVariables)
  }

  /**
   * Attention: this must always be aligned with the environment variables in the run scripts and
   * the way the JAVA_OPTS are assembled there.
   */
  def buildJavaOpts(): Seq[String] = {
    val libraryOpts = Option(getenv("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 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),
        extraEnvironment=appDesc.command.environment)

    Seq("-cp", classPath) ++ libraryOpts ++ userOpts ++ memoryOpts
  }

  /** Spawn a thread that will redirect a given stream to a file */
  def redirectStream(in: InputStream, file: File) {
    val out = new FileOutputStream(file)
    new Thread("redirect output to " + file) {
      override def run() {
        try {
          Utils.copyStream(in, out, true)
        } catch {
          case e: IOException =>
            logInfo("Redirection to " + file + " closed: " + e.getMessage)
        }
      }
    }.start()
  }

  /**
   * Download and run the executor described in our ApplicationDescription
   */
  def fetchAndRunExecutor() {
    try {
      // Create the executor's working directory
      val executorDir = new File(workDir, appId + "/" + execId)
      if (!executorDir.mkdirs()) {
        throw new IOException("Failed to create directory " + executorDir)
      }

      // Launch the process
      val command = buildCommandSeq()
      val builder = new ProcessBuilder(command: _*).directory(executorDir)
      val env = builder.environment()
      for ((key, value) <- appDesc.command.environment) {
        env.put(key, value)
      }
      // In case we are running this from within the Spark Shell, avoid creating a "scala"
      // parent process for the executor command
      env.put("SPARK_LAUNCH_WITH_SCALA", "0")
      process = builder.start()

      // Redirect its stdout and stderr to files
      redirectStream(process.getInputStream, new File(executorDir, "stdout"))
      redirectStream(process.getErrorStream, new File(executorDir, "stderr"))

      // Wait for it to exit; this is actually a bad thing if it happens, because we expect to run
      // long-lived processes only. However, in the future, we might restart the executor a few
      // times on the same machine.
      val exitCode = process.waitFor()
      val message = "Command exited with code " + exitCode
      worker ! ExecutorStateChanged(appId, execId, ExecutorState.FAILED, Some(message),
                                    Some(exitCode))
    } catch {
      case interrupted: InterruptedException =>
        logInfo("Runner thread for executor " + fullId + " interrupted")

      case e: Exception => {
        logError("Error running executor", e)
        if (process != null) {
          process.destroy()
        }
        val message = e.getClass + ": " + e.getMessage
        worker ! ExecutorStateChanged(appId, execId, ExecutorState.FAILED, Some(message), None)
      }
    }
  }
}