Spark 之 deploy

ExecutorRunner

org.apache.spark.deploy.worker.ExecutorRunner

复制代码
  private[worker] def start(): Unit = {
    workerThread = new Thread("ExecutorRunner for " + fullId) {
      override def run(): Unit = { fetchAndRunExecutor() }
    }
    workerThread.start()
    // Shutdown hook that kills actors on shutdown.
    shutdownHook = ShutdownHookManager.addShutdownHook { () =>
      // It's possible that we arrive here before calling `fetchAndRunExecutor`, then `state` will
      // be `ExecutorState.LAUNCHING`. In this case, we should set `state` to `FAILED`.
      if (state == ExecutorState.LAUNCHING || state == ExecutorState.RUNNING) {
        state = ExecutorState.FAILED
      }
      killProcess(Some("Worker shutting down")) }
  }

  private[worker] def start(): Unit = {
    workerThread = new Thread("ExecutorRunner for " + fullId) {
      override def run(): Unit = { fetchAndRunExecutor() }
    }
    workerThread.start()
    // Shutdown hook that kills actors on shutdown.
    shutdownHook = ShutdownHookManager.addShutdownHook { () =>
      // It's possible that we arrive here before calling `fetchAndRunExecutor`, then `state` will
      // be `ExecutorState.LAUNCHING`. In this case, we should set `state` to `FAILED`.
      if (state == ExecutorState.LAUNCHING || state == ExecutorState.RUNNING) {
        state = ExecutorState.FAILED
      }
      killProcess(Some("Worker shutting down")) }
  }

fetchAndRunExecutor 作为线程的主体内容。

也就是说,val exitCode = process.waitFor() 这一阻塞过程,完全放在了线程里。

复制代码
  /**
   * Download and run the executor described in our ApplicationDescription
   */
  private def fetchAndRunExecutor(): Unit = {
    try {
      val resourceFileOpt = prepareResourcesFile(SPARK_EXECUTOR_PREFIX, resources, executorDir)
      // Launch the process
      val arguments = appDesc.command.arguments ++ resourceFileOpt.map(f =>
        Seq("--resourcesFile", f.getAbsolutePath)).getOrElse(Seq.empty)
      val subsOpts = appDesc.command.javaOpts.map {
        Utils.substituteAppNExecIds(_, appId, execId.toString)
      }
      val subsCommand = appDesc.command.copy(arguments = arguments, javaOpts = subsOpts)
      val builder = CommandUtils.buildProcessBuilder(subsCommand, new SecurityManager(conf),
        memory, sparkHome.getAbsolutePath, substituteVariables)
      val command = builder.command()
      val redactedCommand = Utils.redactCommandLineArgs(conf, command.asScala.toSeq)
        .mkString("\"", "\" \"", "\"")
      logInfo(s"Launch command: $redactedCommand")

      builder.directory(executorDir)
      builder.environment.put("SPARK_EXECUTOR_DIRS", appLocalDirs.mkString(File.pathSeparator))
      // In case we are running this from within the Spark Shell, avoid creating a "scala"
      // parent process for the executor command
      builder.environment.put("SPARK_LAUNCH_WITH_SCALA", "0")

      // Add webUI log urls
      val baseUrl =
        if (conf.get(UI_REVERSE_PROXY)) {
          conf.get(UI_REVERSE_PROXY_URL.key, "").stripSuffix("/") +
            s"/proxy/$workerId/logPage/?appId=$appId&executorId=$execId&logType="
        } else {
          s"$webUiScheme$publicAddress:$webUiPort/logPage/?appId=$appId&executorId=$execId&logType="
        }
      builder.environment.put("SPARK_LOG_URL_STDERR", s"${baseUrl}stderr")
      builder.environment.put("SPARK_LOG_URL_STDOUT", s"${baseUrl}stdout")

      process = builder.start()
      val header = "Spark Executor Command: %s\n%s\n\n".format(
        redactedCommand, "=" * 40)

      // Redirect its stdout and stderr to files
      val stdout = new File(executorDir, "stdout")
      stdoutAppender = FileAppender(process.getInputStream, stdout, conf, true)

      val stderr = new File(executorDir, "stderr")
      Files.write(header, stderr, StandardCharsets.UTF_8)
      stderrAppender = FileAppender(process.getErrorStream, stderr, conf, true)

      state = ExecutorState.RUNNING
      worker.send(ExecutorStateChanged(appId, execId, state, None, None))
      // Wait for it to exit; executor may exit with code 0 (when driver instructs it to shutdown)
      // or with nonzero exit code
      val exitCode = process.waitFor()
      state = ExecutorState.EXITED
      val message = "Command exited with code " + exitCode
      worker.send(ExecutorStateChanged(appId, execId, state, Some(message), Some(exitCode)))
    } catch {
      case interrupted: InterruptedException =>
        logInfo("Runner thread for executor " + fullId + " interrupted")
        state = ExecutorState.KILLED
        killProcess(None)
      case e: Exception =>
        logError("Error running executor", e)
        state = ExecutorState.FAILED
        killProcess(Some(e.toString))
    }
  }
killProcess 方案
复制代码
  /**
   * Kill executor process, wait for exit and notify worker to update resource status.
   *
   * @param message the exception message which caused the executor's death
   */
  private def killProcess(message: Option[String]): Unit = {
    var exitCode: Option[Int] = None
    if (process != null) {
      logInfo("Killing process!")
      if (stdoutAppender != null) {
        stdoutAppender.stop()
      }
      if (stderrAppender != null) {
        stderrAppender.stop()
      }
      exitCode = Utils.terminateProcess(process, EXECUTOR_TERMINATE_TIMEOUT_MS)
      if (exitCode.isEmpty) {
        logWarning("Failed to terminate process: " + process +
          ". This process will likely be orphaned.")
      }
    }
    try {
      worker.send(ExecutorStateChanged(appId, execId, state, message, exitCode))
    } catch {
      case e: IllegalStateException => logWarning(e.getMessage(), e)
    }
  }

org.apache.spark.util.Utils.scala

复制代码
  /**
   * Terminates a process waiting for at most the specified duration.
   *
   * @return the process exit value if it was successfully terminated, else None
   */
  def terminateProcess(process: Process, timeoutMs: Long): Option[Int] = {
    // Politely destroy first
    process.destroy()
    if (process.waitFor(timeoutMs, TimeUnit.MILLISECONDS)) {
      // Successful exit
      Option(process.exitValue())
    } else {
      try {
        process.destroyForcibly()
      } catch {
        case NonFatal(e) => logWarning("Exception when attempting to kill process", e)
      }
      // Wait, again, although this really should return almost immediately
      if (process.waitFor(timeoutMs, TimeUnit.MILLISECONDS)) {
        Option(process.exitValue())
      } else {
        logWarning("Timed out waiting to forcibly kill process")
        None
      }
    }
  }
相关推荐
浪子小院23 分钟前
ModelEngine 智能体全流程开发实战:从 0 到 1 搭建多协作办公助手
大数据·人工智能
AEIC学术交流中心1 小时前
【快速EI检索 | ACM出版】2026年大数据与智能制造国际学术会议(BDIM 2026)
大数据·制造
wending-Y1 小时前
记录一次排查Flink一直重启的问题
大数据·flink
UI设计兰亭妙微1 小时前
医疗大数据平台电子病例界面设计
大数据·界面设计
初恋叫萱萱2 小时前
模型瘦身实战:用 `cann-model-compression-toolkit` 实现高效 INT8 量化
大数据
互联网科技看点2 小时前
孕期科学补铁,保障母婴健康-仁合益康蛋白琥珀酸铁口服溶液成为产妇优选方案
大数据
Dxy12393102163 小时前
深度解析 Elasticsearch:从倒排索引到 DSL 查询的实战突围
大数据·elasticsearch·搜索引擎
YongCheng_Liang3 小时前
零基础学大数据:大数据基础与前置技术夯实
大数据·big data
AC赳赳老秦3 小时前
2026国产算力新周期:DeepSeek实战适配英伟达H200,引领大模型训练效率跃升
大数据·前端·人工智能·算法·tidb·memcache·deepseek
鹏说大数据3 小时前
Spark 和 Hive 的关系与区别
大数据·hive·spark