Spark-HeartbeatReceiver 源码解析

HeartbeatReceiver这个类是一个 endPoint，在driver端才有其对象。他的主要作用是 定时监测 注册到 本dirver的所有的executor 是否存活。
下面来看看源码：

class HeartbeatReceiver

private[spark] class HeartbeatReceiver(sc: SparkContext, clock: Clock)
  extends SparkListener with ThreadSafeRpcEndpoint with Logging {

  def this(sc: SparkContext) {
    this(sc, new SystemClock)
  }

  sc.listenerBus.addToManagementQueue(this)

  override val rpcEnv: RpcEnv = sc.env.rpcEnv

  private[spark] var scheduler: TaskScheduler = null

  // executor ID -> timestamp of when the last heartbeat from this executor was received
  private val executorLastSeen = new mutable.HashMap[String, Long] //保存 executor 的 最新 心跳时间

  // "spark.network.timeout" uses "seconds", while `spark.storage.blockManagerSlaveTimeoutMs` uses
  // "milliseconds"
  private val slaveTimeoutMs =
    sc.conf.getTimeAsMs("spark.storage.blockManagerSlaveTimeoutMs", "120s") //blockManager endPoint 超时时间
  private val executorTimeoutMs =
    sc.conf.getTimeAsSeconds("spark.network.timeout", s"${slaveTimeoutMs}ms") * 1000 //executor 超时时间

  // "spark.network.timeoutInterval" uses "seconds", while
  // "spark.storage.blockManagerTimeoutIntervalMs" uses "milliseconds"
  private val timeoutIntervalMs =
    sc.conf.getTimeAsMs("spark.storage.blockManagerTimeoutIntervalMs", "60s") //blockManager 超时时间
  private val checkTimeoutIntervalMs =
    sc.conf.getTimeAsSeconds("spark.network.timeoutInterval", s"${timeoutIntervalMs}ms") * 1000

  private var timeoutCheckingTask: ScheduledFuture[_] = null

  // "eventLoopThread" is used to run some pretty fast actions. The actions running in it should not
  // block the thread for a long time.
  private val eventLoopThread =
    ThreadUtils.newDaemonSingleThreadScheduledExecutor("heartbeat-receiver-event-loop-thread") //driver 检查 executor心跳 线程

  private val killExecutorThread = ThreadUtils.newDaemonSingleThreadExecutor("kill-executor-thread") //处理 exector 心跳失败 kill 的线程

  override def onStart(): Unit = { //启动 ExpireDeadHosts 定时监测
    timeoutCheckingTask = eventLoopThread.scheduleAtFixedRate(new Runnable {
      override def run(): Unit = Utils.tryLogNonFatalError {
        Option(self).foreach(_.ask[Boolean](ExpireDeadHosts)) //处理在 receiveAndReply 的 case ExpireDeadHosts中
      }
    }, 0, checkTimeoutIntervalMs, TimeUnit.MILLISECONDS)
  }

  override def receiveAndReply(context: RpcCallContext): PartialFunction[Any, Unit] = {

    // Messages sent and received locally
    case ExecutorRegistered(executorId) => //增加一个executor
      executorLastSeen(executorId) = clock.getTimeMillis() //更新 executor 的 最新 心跳时间
      context.reply(true)
    case ExecutorRemoved(executorId) => //溢出一个executor
      executorLastSeen.remove(executorId) //更新 executor 的 最新 心跳时间
      context.reply(true)
    case TaskSchedulerIsSet =>
      scheduler = sc.taskScheduler //为本对象 设置 taskscheduler属性
      context.reply(true)
    case ExpireDeadHosts => //检查 executor 是否超时
      expireDeadHosts() //检查 executor 是否超时 ，和超时后的处理方法
      context.reply(true)

    // Messages received from executors 处理 executor 的 心跳信息
    case heartbeat @ Heartbeat(executorId, accumUpdates, blockManagerId) =>
      if (scheduler != null) { //taskscheduler 会在 SparkContext 后面 创建 _heartbeatReceiver.ask[Boolean](TaskSchedulerIsSet) 这句 就是对应的操作
        if (executorLastSeen.contains(executorId)) { //原来已经汇报过了
          executorLastSeen(executorId) = clock.getTimeMillis() //更新 executor 的 最新 心跳时间
          eventLoopThread.submit(new Runnable {
            override def run(): Unit = Utils.tryLogNonFatalError {
              val unknownExecutor: Boolean = !scheduler.executorHeartbeatReceived( //unknownExecutor这个如果是 true的话，则会在 executor 中 重新注册blockManager
                // 可以看org.apache.spark.executor.Executor的reportHeartBeat方法
                executorId, accumUpdates, blockManagerId)
              //如果 driver的BlockManagerMasterEndPoint 中已经注册过了 这个 blockManagerId，则返回false，则不需要executor 再次 重新注册 BlockManager
              val response = HeartbeatResponse(reregisterBlockManager = unknownExecutor)
              context.reply(response)
            }
          })
        } else {//还没有 在 driver 注册过
          // This may happen if we get an executor's in-flight heartbeat immediately
          // after we just removed it. It's not really an error condition so we should
          // not log warning here. Otherwise there may be a lot of noise especially if
          // we explicitly remove executors (SPARK-4134).
          logDebug(s"Received heartbeat from unknown executor $executorId")
          context.reply(HeartbeatResponse(reregisterBlockManager = true)) //需要 executor 重新注册 BlockManager
        }
      } else {// TaskScheduler 还没有 在 此类中还没有完成 持有化
        // Because Executor will sleep several seconds before sending the first "Heartbeat", this
        // case rarely happens. However, if it really happens, log it and ask the executor to
        // register itself again.
        logWarning(s"Dropping $heartbeat because TaskScheduler is not ready yet")
        context.reply(HeartbeatResponse(reregisterBlockManager = true))
      }
  }

  /**
   * Send ExecutorRegistered to the event loop to add a new executor. Only for test.
   *
   * @return if HeartbeatReceiver is stopped, return None. Otherwise, return a Some(Future) that
   *         indicate if this operation is successful.
   */
  //add 一个executor
  def addExecutor(executorId: String): Option[Future[Boolean]] = {
    Option(self).map(_.ask[Boolean](ExecutorRegistered(executorId)))
  }

  /**
   * If the heartbeat receiver is not stopped, notify it of executor registrations.
   */
  //add 一个executor
  override def onExecutorAdded(executorAdded: SparkListenerExecutorAdded): Unit = {
    addExecutor(executorAdded.executorId)
  }

  /**
   * Send ExecutorRemoved to the event loop to remove an executor. Only for test.
   *
   * @return if HeartbeatReceiver is stopped, return None. Otherwise, return a Some(Future) that
   *         indicate if this operation is successful.
   */
  //移除一个executor
  def removeExecutor(executorId: String): Option[Future[Boolean]] = {
    Option(self).map(_.ask[Boolean](ExecutorRemoved(executorId)))
  }

  /**
   * If the heartbeat receiver is not stopped, notify it of executor removals so it doesn't
   * log superfluous errors.
   *
   * Note that we must do this after the executor is actually removed to guard against the
   * following race condition: if we remove an executor's metadata from our data structure
   * prematurely, we may get an in-flight heartbeat from the executor before the executor is
   * actually removed, in which case we will still mark the executor as a dead host later
   * and expire it with loud error messages.
   */
  //移除一个executor
  override def onExecutorRemoved(executorRemoved: SparkListenerExecutorRemoved): Unit = {
    removeExecutor(executorRemoved.executorId)
  }
  //检查 executor 是否超时 ，和超时后的处理方法
  private def expireDeadHosts(): Unit = {
    logTrace("Checking for hosts with no recent heartbeats in HeartbeatReceiver.")
    val now = clock.getTimeMillis()
    for ((executorId, lastSeenMs) <- executorLastSeen) { // executorLastSeen 是保存 executor 的 最新 心跳时间
      if (now - lastSeenMs > executorTimeoutMs) {//executor 超过 超时时间设置
        logWarning(s"Removing executor $executorId with no recent heartbeats: " +
          s"${now - lastSeenMs} ms exceeds timeout $executorTimeoutMs ms")
        //taskScheduler 执行 executor lost 相应的 操作
        scheduler.executorLost(executorId, SlaveLost("Executor heartbeat " +
          s"timed out after ${now - lastSeenMs} ms"))
          // Asynchronously kill the executor to avoid blocking the current thread
        killExecutorThread.submit(new Runnable {// kill executor 线程 单独处理  kill 和 replace 这个 executor
          override def run(): Unit = Utils.tryLogNonFatalError {
            // Note: we want to get an executor back after expiring this one,
            // so do not simply call `sc.killExecutor` here (SPARK-8119)
            sc.killAndReplaceExecutor(executorId) //schedulerBackend 的处理方法
          }
        })
        executorLastSeen.remove(executorId) //executorLastSeen 是保存 executor 的 最新 心跳时间 移除这个 executor
      }
    }
  }

  override def onStop(): Unit = {
    if (timeoutCheckingTask != null) {
      timeoutCheckingTask.cancel(true)
    }
    eventLoopThread.shutdownNow()
    killExecutorThread.shutdownNow()
  }
}

object HeartbeatReceiver

private[spark] object HeartbeatReceiver {
  val ENDPOINT_NAME = "HeartbeatReceiver"
}