5 基于案例一节课贯通Spark Streaming流计算框架的运行源码-原创手记-慕课网

本期内容：

1、在线动态计算分类最热门商品案例回顾与演示
2、基于案例贯通Spark Streaming的运行源码

第一部分案例：

package com.dt.spark.sparkstreamingimport com.robinspark.utils.ConnectionPoolimport org.apache.spark.SparkConfimport org.apache.spark.sql.Rowimport org.apache.spark.sql.hive.HiveContextimport org.apache.spark.sql.types.{IntegerType, StringType, StructField, StructType}import org.apache.spark.streaming.{Seconds, StreamingContext}/**
  * 使用Spark Streaming+Spark SQL来在线动态计算电商中不同类别中最热门的商品排名，例如手机这个类别下面最热门的三种手机、电视这个类别
  * 下最热门的三种电视，该实例在实际生产环境下具有非常重大的意义；
  *
  *
  *
  *   实现技术：Spark Streaming+Spark SQL，之所以Spark Streaming能够使用ML、sql、graphx等功能是因为有foreachRDD和Transform
  * 等接口，这些接口中其实是基于RDD进行操作，所以以RDD为基石，就可以直接使用Spark其它所有的功能，就像直接调用API一样简单。
  *  假设说这里的数据的格式：user item category，例如Rocky Samsung Android
  */object OnlineTheTop3ItemForEachCategory2DB {
  def main(args: Array[String]){    /**
      * 第1步：创建Spark的配置对象SparkConf，设置Spark程序的运行时的配置信息，
      * 例如说通过setMaster来设置程序要链接的Spark集群的Master的URL,如果设置
      * 为local，则代表Spark程序在本地运行，特别适合于机器配置条件非常差（例如
      * 只有1G的内存）的初学者       *
      */
    val conf = new SparkConf() //创建SparkConf对象
    conf.setAppName("OnlineTheTop3ItemForEachCategory2DB") //设置应用程序的名称，在程序运行的监控界面可以看到名称
    conf.setMaster("spark://Master:7077") //此时，程序在Spark集群
    //conf.setMaster("local[2]")
    //设置batchDuration时间间隔来控制Job生成的频率并且创建Spark Streaming执行的入口
    val ssc = new StreamingContext(conf, Seconds(5))

    ssc.checkpoint("/root/Documents/SparkApps/checkpoint")

    val userClickLogsDStream = ssc.socketTextStream("Master", 9999)

    val formattedUserClickLogsDStream = userClickLogsDStream.map(clickLog =>
        (clickLog.split(" ")(2) + "_" + clickLog.split(" ")(1), 1))//    val categoryUserClickLogsDStream = formattedUserClickLogsDStream.reduceByKeyAndWindow((v1:Int, v2: Int) => v1 + v2,//      (v1:Int, v2: Int) => v1 - v2, Seconds(60), Seconds(20))

    val categoryUserClickLogsDStream = formattedUserClickLogsDStream.reduceByKeyAndWindow(_+_,
      _-_, Seconds(60), Seconds(20))

    categoryUserClickLogsDStream.foreachRDD { rdd => {      if (rdd.isEmpty()) {
        println("No data inputted!!!")
      } else {
        val categoryItemRow = rdd.map(reducedItem => {
          val category = reducedItem._1.split("_")(0)
          val item = reducedItem._1.split("_")(1)
          val click_count = reducedItem._2
          Row(category, item, click_count)
        })

        val structType = StructType(Array(
          StructField("category", StringType, true),
          StructField("item", StringType, true),
          StructField("click_count", IntegerType, true)
        ))

        val hiveContext = new HiveContext(rdd.context)
        val categoryItemDF = hiveContext.createDataFrame(categoryItemRow, structType)

        categoryItemDF.registerTempTable("categoryItemTable")

        val reseltDataFram = hiveContext.sql("SELECT category,item,click_count FROM (SELECT category,item,click_count,row_number()" +          " OVER (PARTITION BY category ORDER BY click_count DESC) rank FROM categoryItemTable) subquery " +          " WHERE rank <= 3")
        reseltDataFram.show()

        val resultRowRDD = reseltDataFram.rdd

        resultRowRDD.foreachPartition { partitionOfRecords => {          if (partitionOfRecords.isEmpty){
            println("This RDD is not null but partition is null")
          } else {            // ConnectionPool is a static, lazily initialized pool of connections
            val connection = ConnectionPool.getConnection()
            partitionOfRecords.foreach(record => {
              val sql = "insert into categorytop3(category,item,client_count) values('" + record.getAs("category") + "','" +
                record.getAs("item") + "'," + record.getAs("click_count") + ")"
              val stmt = connection.createStatement();
              stmt.executeUpdate(sql);

            })
            ConnectionPool.returnConnection(connection) // return to the pool for future reuse

          }
        }
        }
      }
    }
    }    /**
      * 在StreamingContext调用start方法的内部其实是会启动JobScheduler的Start方法，进行消息循环，在JobScheduler
      * 的start内部会构造JobGenerator和ReceiverTacker，并且调用JobGenerator和ReceiverTacker的start方法：
      *   1，JobGenerator启动后会不断的根据batchDuration生成一个个的Job
      *   2，ReceiverTracker启动后首先在Spark Cluster中启动Receiver（其实是在Executor中先启动ReceiverSupervisor），在Receiver收到
      *   数据后会通过ReceiverSupervisor存储到Executor并且把数据的Metadata信息发送给Driver中的ReceiverTracker，在ReceiverTracker
      *   内部会通过ReceivedBlockTracker来管理接受到的元数据信息
      * 每个BatchInterval会产生一个具体的Job，其实这里的Job不是Spark Core中所指的Job，它只是基于DStreamGraph而生成的RDD
      * 的DAG而已，从Java角度讲，相当于Runnable接口实例，此时要想运行Job需要提交给JobScheduler，在JobScheduler中通过线程池的方式找到一个
      * 单独的线程来提交Job到集群运行（其实是在线程中基于RDD的Action触发真正的作业的运行），为什么使用线程池呢？
      *   1，作业不断生成，所以为了提升效率，我们需要线程池；这和在Executor中通过线程池执行Task有异曲同工之妙；
      *   2，有可能设置了Job的FAIR公平调度的方式，这个时候也需要多线程的支持；
      *
      */
    ssc.start()
    ssc.awaitTermination()
  }
}

第二部分源码解析：

构建StreamingContext时传递SparkConf参数（或者自己Configuration）在内部创建SparkContext

def this(conf: SparkConf, batchDuration: Duration) = {  this(StreamingContext.createNewSparkContext(conf), null, batchDuration)
}

事实说明SparkStreaming就是SparkCore上的一个应用程序

private[streaming] def createNewSparkContext(conf: SparkConf): SparkContext = {  new SparkContext(conf)
}

创建Socket输入流

def socketTextStream(
    hostname: String,    port: Int,    storageLevel: StorageLevel = StorageLevel.MEMORY_AND_DISK_SER_2
  ): ReceiverInputDStream[String] = withNamedScope("socket text stream") {
  socketStream[String](hostname, port, SocketReceiver.bytesToLines, storageLevel)
}

创建SocketInputDStream

def socketStream[T: ClassTag](
    hostname: String,    port: Int,    converter: (InputStream) => Iterator[T],    storageLevel: StorageLevel
  ): ReceiverInputDStream[T] = {  new SocketInputDStream[T](this, hostname, port, converter, storageLevel)
}

SocketInputDstream继承ReceiverInputDStream，通过构建Receiver来接收数据

private[streaming]class SocketInputDStream[T: ClassTag](
    ssc_ : StreamingContext,    host: String,    port: Int,    bytesToObjects: InputStream => Iterator[T],    storageLevel: StorageLevel
  ) extends ReceiverInputDStream[T](ssc_) {  def getReceiver(): Receiver[T] = {
    new SocketReceiver(host, port, bytesToObjects, storageLevel)
  }
}

5.1

  ReceiverInputDStream
  abstract class ReceiverInputDStream[T: ClassTag](ssc_ :   StreamingContext)
    extends InputDStream[T](ssc_) {

  abstract class InputDStream[T: ClassTag] (ssc_ : StreamingContext)
    extends DStream[T](ssc_) {

5.2. DStream

    - 依赖于其他DStream
    -  什么时候依据DStream，依赖关系的模板，构成RDD之间的依赖
    -  基于DStream它有一个Function，Function 基于Batch     Interval（time Interval）生成RDD，这个和定时器有关系

  abstract class DStream[T: ClassTag] (     @transient private[streaming] var ssc: StreamingContext) extends Serializable with Logging {

SocketReceiver对象在onStart中创建Thread启动run方法调用执行receive接收数据。
def onStart() {
// Start the thread that receives data over a connection
new Thread("Socket Receiver") {
setDaemon(true)
override def run() { receive() }
}.start()
}
创建一个Socket connection连接接收数据

/** Create a socket connection and receive data until receiver is stopped */
  def receive() {    var socket: Socket = null
    try {
      logInfo("Connecting to " + host + ":" + port)
      socket = new Socket(host, port)
      logInfo("Connected to " + host + ":" + port)
      val iterator = bytesToObjects(socket.getInputStream())      while(!isStopped && iterator.hasNext) {
        store(iterator.next)
      }      if (!isStopped()) {
        restart("Socket data stream had no more data")
      } else {
        logInfo("Stopped receiving")
      }

总体流程：在StreamingContext调用start方法的内部其实是会启动JobScheduler的Start方法，进行消息循环，在JobScheduler的start内部会构造JobGenerator和ReceiverTacker，并且调用JobGenerator和ReceiverTacker的start方法：
1）、JobGenerator启动后会不断的根据batchDuration生成一个个的Job
2）、ReceiverTracker启动后首先在Spark Cluster中启动Receiver（其实是在Executor中先启动ReceiverSupervisor），
在Receiver收到数据后会通过ReceiverSupervisor存储到Executor并且把数据的Metadata信息发送给Driver中的ReceiverTracker，
在ReceiverTracker 内部会通过ReceivedBlockTracker来管理接受到的元数据信息每个BatchInterval会产生一个具体的Job（这里的Job主要是封装了业务逻辑例如上面实例中的代码），其实这里的Job不是Spark Core中所指的Job，它只是基于DStreamGraph而生成的RDD 的DAG而已，
从Java角度讲，相当于Runnable接口实例，此时要想运行Job需要提交给JobScheduler，在JobScheduler中通过线程池的方式找到一个单独的线程来提交Job到集群运行（其实是在线程中基于RDD的Action触发真正的作业的运行）,
为什么使用线程池呢？
a）、作业不断生成，所以为了提升效率，我们需要线程池；这和在Executor中通过线程池执行Task有异曲同工之妙；
b）、有可能设置了Job的FAIR公平调度的方式，这个时候也需要多线程的支持；

8.1、StreamingContext.start

// Start the streaming scheduler in a new thread, so that thread local properties// like call sites and job groups can be reset without affecting those of the// current thread.ThreadUtils.runInNewThread("streaming-start") {
  sparkContext.setCallSite(startSite.get)
  sparkContext.clearJobGroup()
  sparkContext.setLocalProperty(SparkContext.SPARK_JOB_INTERRUPT_ON_CANCEL, "false")
  scheduler.start()
}

补充：线程本地存储，线程ThreadLocal每个线程有自己的私有属性，设置线程的私有属性不会影响当前线程或其他线程

JobScheduler.start 创建EventLoop消息线程并启动

def start(): Unit = synchronized {  if (eventLoop != null) return // scheduler has already been started

  logDebug("Starting JobScheduler")
  eventLoop = new EventLoop[JobSchedulerEvent]("JobScheduler") {    override protected def onReceive(event: JobSchedulerEvent): Unit = processEvent(event)    override protected def onError(e: Throwable): Unit = reportError("Error in job scheduler", e)
  }
  eventLoop.start()

9.1 EventLoop中创建Thread线程接收和发送消息，调用JobScheduler中的processEvent方法

private[spark] abstract class EventLoop[E](name: String) extends Logging {  private val eventQueue: BlockingQueue[E] = new LinkedBlockingDeque[E]()  private val stopped = new AtomicBoolean(false)  private val eventThread = new Thread(name) {
    setDaemon(true)

    override def run(): Unit = {      try {        while (!stopped.get) {
          val event = eventQueue.take()          try {
            onReceive(event)
          } catch {

9.2 会接受不同的任务，JobScheduler是整个Job的调度器，它本身用了一个线程循环，去监听不同的Job启动、Job完成、Job失败等任务（消息驱动系统）
private def processEvent(event: JobSchedulerEvent) {
try {
event match {
case JobStarted(job, startTime) => handleJobStart(job, startTime)
case JobCompleted(job, completedTime) => handleJobCompletion(job, completedTime)
case ErrorReported(m, e) => handleError(m, e)
}
} catch {

JobScheduler.start

// attach rate controllers of input streams to receive batch completion updatesfor {
  inputDStream <- ssc.graph.getInputStreams
  rateController <- inputDStream.rateController
} ssc.addStreamingListener(rateController)

10.1 多个InputStream
``
inputDStream <- ssc.graph.getInputStreams

  10.2 RateController控制输入的速度// Keep track of the freshest rate for this stream using the rateEstimatorprotected[streaming] val rateController: Option[RateController] = None11. JobScheduler.start

listenerBus.start(ssc.sparkContext)
receiverTracker = new ReceiverTracker(ssc)
inputInfoTracker = new InputInfoTracker(ssc)
receiverTracker.start()
jobGenerator.start()

11.1 StreamingListenerBus

override def onPostEvent(listener: StreamingListener, event: StreamingListenerEvent): Unit = {
event match {
case receiverStarted: StreamingListenerReceiverStarted =>
listener.onReceiverStarted(receiverStarted)
case receiverError: StreamingListenerReceiverError =>
listener.onReceiverError(receiverError)
case receiverStopped: StreamingListenerReceiverStopped =>
listener.onReceiverStopped(receiverStopped)
case batchSubmitted: StreamingListenerBatchSubmitted =>
listener.onBatchSubmitted(batchSubmitted)
case batchStarted: StreamingListenerBatchStarted =>
listener.onBatchStarted(batchStarted)
case batchCompleted: StreamingListenerBatchCompleted =>
listener.onBatchCompleted(batchCompleted)
case outputOperationStarted: StreamingListenerOutputOperationStarted =>
listener.onOutputOperationStarted(outputOperationStarted)
case outputOperationCompleted: StreamingListenerOutputOperationCompleted =>
listener.onOutputOperationCompleted(outputOperationCompleted)
case _ =>
}
}

  11.2 receiverTracker.start()，ReceiveTracker是通过发Job的方式到集群的Executor上启动Receiver

/** Start the endpoint and receiver execution thread. */
def start(): Unit = synchronized {
if (isTrackerStarted) {
throw new SparkException("ReceiverTracker already started")
}

if (!receiverInputStreams.isEmpty) {
endpoint = ssc.env.rpcEnv.setupEndpoint(
"ReceiverTracker", new ReceiverTrackerEndpoint(ssc.env.rpcEnv))
if (!skipReceiverLaunch) launchReceivers()
logInfo("ReceiverTracker started")
trackerState = Started
}
}

11.2.1、创建一个ReceiverTrackerEndpoint消息通信体

override def receive: PartialFunction[Any, Unit] = {
// Local messages
case StartAllReceivers(receivers) =>
val scheduledLocations = schedulingPolicy.scheduleReceivers(receivers, getExecutors)
for (receiver <- receivers) {
val executors = scheduledLocations(receiver.streamId)
updateReceiverScheduledExecutors(receiver.streamId, executors)
receiverPreferredLocations(receiver.streamId) = receiver.preferredLocation
startReceiver(receiver, executors)
}
case RestartReceiver(receiver) =>
// Old scheduled executors minus the ones that are not active any more
val oldScheduledExecutors = getStoredScheduledExecutors(receiver.streamId)
val scheduledLocations = if (oldScheduledExecutors.nonEmpty) {
// Try global scheduling again
oldScheduledExecutors
} else {
val oldReceiverInfo = receiverTrackingInfos(receiver.streamId)
// Clear "scheduledLocations" to indicate we are going to do local scheduling
val newReceiverInfo = oldReceiverInfo.copy(
state = ReceiverState.INACTIVE, scheduledLocations = None)
receiverTrackingInfos(receiver.streamId) = newReceiverInfo
schedulingPolicy.rescheduleReceiver(
receiver.streamId,
receiver.preferredLocation,
receiverTrackingInfos,
getExecutors)
}
// Assume there is one receiver restarting at one time, so we don't need to update
// receiverTrackingInfos
startReceiver(receiver, scheduledLocations)

11.2.1.1、ReceiverSchedulingPolicy.scheduleReceivers，从下面的代码中可以看出来在那些Executor上启动Receiver,以及怎么具体在Executor上启动Receiver

// Firstly, we need to respect "preferredLocation". So if a receiver has "preferredLocation",
// we need to make sure the "preferredLocation" is in the candidate scheduled executor list.
for (i <- 0 until receivers.length) {
// Note: preferredLocation is host but executors are host_executorId
receivers(i).preferredLocation.foreach { host =>
hostToExecutors.get(host) match {
case Some(executorsOnHost) =>
// preferredLocation is a known host. Select an executor that has the least receivers in
// this host
val leastScheduledExecutor =
executorsOnHost.minBy(executor => numReceiversOnExecutor(executor))
scheduledLocations(i) += leastScheduledExecutor
numReceiversOnExecutor(leastScheduledExecutor) =
numReceiversOnExecutor(leastScheduledExecutor) + 1
case None =>
// preferredLocation is an unknown host.
// Note: There are two cases:
// 1. This executor is not up. But it may be up later.
// 2. This executor is dead, or it's not a host in the cluster.
// Currently, simply add host to the scheduled executors.

    // Note: host could be `HDFSCacheTaskLocation`, so use `TaskLocation.apply` to handle
    // this case
    scheduledLocations(i) += TaskLocation(host)
}

}
}

补充：ReceiverTracker本身不直接监管Receiver，它是Driver级别的可间接地，用ReceiverSupervisor监控那台机器上Executor中的Receiver。

11.2.2、if (!skipReceiverLaunch) launchReceivers()

/**

Get the receivers from the ReceiverInputDStreams, distributes them to the
worker nodes as a parallel collection, and runs them.
*/
private def launchReceivers(): Unit = {
val receivers = receiverInputStreams.map(nis => {
val rcvr = nis.getReceiver()
rcvr.setReceiverId(nis.id)
rcvr
})

runDummySparkJob()

logInfo("Starting " + receivers.length + " receivers")
endpoint.send(StartAllReceivers(receivers))
}

11.2.2.1运行了一个Dummy的作业，确保所有的Slaves正常工作，保证所有的Receiver都在一台机器上

/**

Run the dummy Spark job to ensure that all slaves have registered. This avoids all the
receivers to be scheduled on the same node.
TODO Should poll the executor number and wait for executors according to
"spark.scheduler.minRegisteredResourcesRatio" and
"spark.scheduler.maxRegisteredResourcesWaitingTime" rather than running a dummy job.
*/
private def runDummySparkJob(): Unit = {
if (!ssc.sparkContext.isLocal) {
ssc.sparkContext.makeRDD(1 to 50, 50).map(x => (x, 1)).reduceByKey(_ + _, 20).collect()
}
assert(getExecutors.nonEmpty)
}

11.2.2.2、endpoint.send(StartAllReceivers(receivers)

// endpoint is created when generator starts.
// This not being null means the tracker has been started and not stopped
private var endpoint: RpcEndpointRef = null

endpoint = ssc.env.rpcEnv.setupEndpoint(
"ReceiverTracker", new ReceiverTrackerEndpoint(ssc.env.rpcEnv))

ReceiverTrackerEndpoint
override def receive: PartialFunction[Any, Unit] = {
// Local messages
case StartAllReceivers(receivers) =>
val scheduledLocations = schedulingPolicy.scheduleReceivers(receivers, getExecutors)
for (receiver <- receivers) {
val executors = scheduledLocations(receiver.streamId)
updateReceiverScheduledExecutors(receiver.streamId, executors)
receiverPreferredLocations(receiver.streamId) = receiver.preferredLocation
startReceiver(receiver, executors)
}

startReceiver
// Function to start the receiver on the worker node
val startReceiverFunc: Iterator[Receiver[]] => Unit =
(iterator: Iterator[Receiver[]]) => {
if (!iterator.hasNext) {
throw new SparkException(
"Could not start receiver as object not found.")
}
if (TaskContext.get().attemptNumber() == 0) {
val receiver = iterator.next()
assert(iterator.hasNext == false)
val supervisor = new ReceiverSupervisorImpl(
receiver, SparkEnv.get, serializableHadoopConf.value, checkpointDirOption)
supervisor.start()
supervisor.awaitTermination()
} else {
// It's restarted by TaskScheduler, but we want to reschedule it again. So exit it.
}
}

逆天的设计啊

// Create the RDD using the scheduledLocations to run the receiver in a Spark job
val receiverRDD: RDD[Receiver[]] =
if (scheduledLocations.isEmpty) {
ssc.sc.makeRDD(Seq(receiver), 1)
} else {
val preferredLocations = scheduledLocations.map(.toString).distinct
ssc.sc.makeRDD(Seq(receiver -> preferredLocations))
}
receiverRDD.setName(s"Receiver $receiverId")
ssc.sparkContext.setJobDescription(s"Streaming job running receiver $receiverId")
ssc.sparkContext.setCallSite(Option(ssc.getStartSite()).getOrElse(Utils.getCallSite()))

val future = ssc.sparkContext.submitJob[Receiver[], Unit, Unit](
receiverRDD, startReceiverFunc, Seq(0), (, ) => Unit, ())
// We will keep restarting the receiver job until ReceiverTracker is stopped
future.onComplete {
case Success() =>
if (!shouldStartReceiver) {
onReceiverJobFinish(receiverId)
} else {
logInfo(s"Restarting Receiver $receiverId")
self.send(RestartReceiver(receiver))
}
case Failure(e) =>
if (!shouldStartReceiver) {
onReceiverJobFinish(receiverId)
} else {
logError("Receiver has been stopped. Try to restart it.", e)
logInfo(s"Restarting Receiver $receiverId")
self.send(RestartReceiver(receiver))
}
}(submitJobThreadPool)
logInfo(s"Receiver ${receiver.streamId} started")
}

ReceiverSupervisorImpl.startReceiver
/** Start receiver */
def startReceiver(): Unit = synchronized {
try {
if (onReceiverStart()) {
logInfo("Starting receiver")
receiverState = Started
receiver.onStart()
logInfo("Called receiver onStart")
} else {
// The driver refused us
stop("Registered unsuccessfully because Driver refused to start receiver " + streamId, None)
}

override protected def onReceiverStart(): Boolean = {
val msg = RegisterReceiver(
streamId, receiver.getClass.getSimpleName, host, executorId, endpoint)
trackerEndpoint.askWithRetryBoolean
}

11.3、JobScheduler.start  jobGenerator.start()

/** Start generation of jobs */
def start(): Unit = synchronized {
if (eventLoop != null) return // generator has already been started

// Call checkpointWriter here to initialize it before eventLoop uses it to avoid a deadlock.
// See SPARK-10125
checkpointWriter

eventLoop = new EventLoopJobGeneratorEvent {
override protected def onReceive(event: JobGeneratorEvent): Unit = processEvent(event)

override protected def onError(e: Throwable): Unit = {
  jobScheduler.reportError("Error in job generator", e)
}

}
eventLoop.start()

if (ssc.isCheckpointPresent) {
restart()
} else {
startFirstTime()
}
}

根据时间间隔不断发送消息

/** Processes all events */
private def processEvent(event: JobGeneratorEvent) {
logDebug("Got event " + event)
event match {
case GenerateJobs(time) => generateJobs(time)
case ClearMetadata(time) => clearMetadata(time)
case DoCheckpoint(time, clearCheckpointDataLater) =>
doCheckpoint(time, clearCheckpointDataLater)
case ClearCheckpointData(time) => clearCheckpointData(time)
}
}

/** Generate jobs and perform checkpoint for the given time. */
private def generateJobs(time: Time) {
// Set the SparkEnv in this thread, so that job generation code can access the environment
// Example: BlockRDDs are created in this thread, and it needs to access BlockManager
// Update: This is probably redundant after threadlocal stuff in SparkEnv has been removed.
SparkEnv.set(ssc.env)
Try {
jobScheduler.receiverTracker.allocateBlocksToBatch(time) // allocate received blocks to batch
graph.generateJobs(time) // generate jobs using allocated block
} match {
case Success(jobs) =>
val streamIdToInputInfos = jobScheduler.inputInfoTracker.getInfo(time)
jobScheduler.submitJobSet(JobSet(time, jobs, streamIdToInputInfos))
case Failure(e) =>
jobScheduler.reportError("Error generating jobs for time " + time, e)
}
eventLoop.post(DoCheckpoint(time, clearCheckpointDataLater = false))
}

def submitJobSet(jobSet: JobSet) {
if (jobSet.jobs.isEmpty) {
logInfo("No jobs added for time " + jobSet.time)
} else {
listenerBus.post(StreamingListenerBatchSubmitted(jobSet.toBatchInfo))
jobSets.put(jobSet.time, jobSet)
jobSet.jobs.foreach(job => jobExecutor.execute(new JobHandler(job)))
logInfo("Added jobs for time " + jobSet.time)
}
}

/**

Executes the given task sometime in the future. The task
may execute in a new thread or in an existing pooled thread.
If the task cannot be submitted for execution, either because this
executor has been shutdown or because its capacity has been reached,
the task is handled by the current {@code RejectedExecutionHandler}.
@param command the task to execute
@throws RejectedExecutionException at discretion of

    {@code RejectedExecutionHandler}, if the task

    cannot be accepted for execution

@throws NullPointerException if {@code command} is null
/
public void execute(Runnable command) {
if (command == null)
throw new NullPointerException();
/
- Proceed in 3 steps:
- start a new thread with the given command as its first
- task. The call to addWorker atomically checks runState and
- workerCount, and so prevents false alarms that would add
- threads when it shouldn't, by returning false.
- to double-check whether we should have added a thread
- (because existing ones died since last checking) or that
- the pool shut down since entry into this method. So we
- recheck state and if necessary roll back the enqueuing if
- stopped, or start a new thread if there are none.
- thread. If it fails, we know we are shut down or saturated
- and so reject the task.
  */
  int c = ctl.get();
  if (workerCountOf(c) < corePoolSize) {
  if (addWorker(command, true))
  return;
  c = ctl.get();
  }
  if (isRunning(c) && workQueue.offer(command)) {
  int recheck = ctl.get();
  if (! isRunning(recheck) && remove(command))
  reject(command);
  else if (workerCountOf(recheck) == 0)
  addWorker(null, false);
  }
  else if (!addWorker(command, false))
  reject(command);
  }

If we cannot queue task, then we try to add a new

If a task can be successfully queued, then we still need

If fewer than corePoolSize threads are running, try to

private class JobHandler(job: Job) extends Runnable with Logging {
import JobScheduler._

def run() {  try {
    val formattedTime = UIUtils.formatBatchTime(
      job.time.milliseconds, ssc.graph.batchDuration.milliseconds, showYYYYMMSS = false)
    val batchUrl = s"/streaming/batch/?id=${job.time.milliseconds}"
    val batchLinkText = s"[output operation ${job.outputOpId}, batch time ${formattedTime}]"

    ssc.sc.setJobDescription(
      s"""Streaming job from <a href="$batchUrl">$batchLinkText</a>""")
    ssc.sc.setLocalProperty(BATCH_TIME_PROPERTY_KEY, job.time.milliseconds.toString)
    ssc.sc.setLocalProperty(OUTPUT_OP_ID_PROPERTY_KEY, job.outputOpId.toString)    // We need to assign `eventLoop` to a temp variable. Otherwise, because
    // `JobScheduler.stop(false)` may set `eventLoop` to null when this method is running, then
    // it's possible that when `post` is called, `eventLoop` happens to null.
    var _eventLoop = eventLoop    if (_eventLoop != null) {
      _eventLoop.post(JobStarted(job, clock.getTimeMillis()))      // Disable checks for existing output directories in jobs launched by the streaming
      // scheduler, since we may need to write output to an existing directory during checkpoint
      // recovery; see SPARK-4835 for more details.
      PairRDDFunctions.disableOutputSpecValidation.withValue(true) {
        job.run()
      }
      _eventLoop = eventLoop      if (_eventLoop != null) {
        _eventLoop.post(JobCompleted(job, clock.getTimeMillis()))
      }
    } else {      // JobScheduler has been stopped.
    }
  } finally {
    ssc.sc.setLocalProperty(JobScheduler.BATCH_TIME_PROPERTY_KEY, null)
    ssc.sc.setLocalProperty(JobScheduler.OUTPUT_OP_ID_PROPERTY_KEY, null)
  }
}

}
}

private def processEvent(event: JobSchedulerEvent) {
try {
event match {
case JobStarted(job, startTime) => handleJobStart(job, startTime)
case JobCompleted(job, completedTime) => handleJobCompletion(job, completedTime)
case ErrorReported(m, e) => handleError(m, e)
}
} catch {
case e: Throwable =>
reportError("Error in job scheduler", e)
}
}

private def handleJobStart(job: Job, startTime: Long) {
val jobSet = jobSets.get(job.time)
val isFirstJobOfJobSet = !jobSet.hasStarted
jobSet.handleJobStart(job)
if (isFirstJobOfJobSet) {
// "StreamingListenerBatchStarted" should be posted after calling "handleJobStart" to get the
// correct "jobSet.processingStartTime".
listenerBus.post(StreamingListenerBatchStarted(jobSet.toBatchInfo))
}
job.setStartTime(startTime)
listenerBus.post(StreamingListenerOutputOperationStarted(job.toOutputOperationInfo))
logInfo("Starting job " + job.id + " from job set of time " + jobSet.time)
}

private class JobHandler(job: Job) extends Runnable with Logging {
import JobScheduler._

def run() {  try {
    val formattedTime = UIUtils.formatBatchTime(
      job.time.milliseconds, ssc.graph.batchDuration.milliseconds, showYYYYMMSS = false)
    val batchUrl = s"/streaming/batch/?id=${job.time.milliseconds}"
    val batchLinkText = s"[output operation ${job.outputOpId}, batch time ${formattedTime}]"

    ssc.sc.setJobDescription(
      s"""Streaming job from <a href="$batchUrl">$batchLinkText</a>""")
    ssc.sc.setLocalProperty(BATCH_TIME_PROPERTY_KEY, job.time.milliseconds.toString)
    ssc.sc.setLocalProperty(OUTPUT_OP_ID_PROPERTY_KEY, job.outputOpId.toString)    // We need to assign `eventLoop` to a temp variable. Otherwise, because
    // `JobScheduler.stop(false)` may set `eventLoop` to null when this method is running, then
    // it's possible that when `post` is called, `eventLoop` happens to null.
    var _eventLoop = eventLoop    if (_eventLoop != null) {
      _eventLoop.post(JobStarted(job, clock.getTimeMillis()))      // Disable checks for existing output directories in jobs launched by the streaming
      // scheduler, since we may need to write output to an existing directory during checkpoint
      // recovery; see SPARK-4835 for more details.
      PairRDDFunctions.disableOutputSpecValidation.withValue(true) {
        job.run()
      }
      _eventLoop = eventLoop      if (_eventLoop != null) {
        _eventLoop.post(JobCompleted(job, clock.getTimeMillis()))
      }
    } else {      // JobScheduler has been stopped.
    }
  } finally {
    ssc.sc.setLocalProperty(JobScheduler.BATCH_TIME_PROPERTY_KEY, null)
    ssc.sc.setLocalProperty(JobScheduler.OUTPUT_OP_ID_PROPERTY_KEY, null)
  }
}

}
}

作者：海纳百川_spark
链接：https://www.jianshu.com/p/a40078004ced