# Run on a Spark standalone cluster
./bin/spark-submit \--class org.apache.spark.examples.SparkPi \--master spark://207.184.161.138:7077 \--executor-memory 20G \--total-executor-cores 100 \/path/to/examples.jar \1000

def main(args: Array[String]) {val appArgs = new SparkSubmitArguments(args)val (childArgs, classpath, sysProps, mainClass) = createLaunchEnv(appArgs)launch(childArgs, classpath, sysProps, mainClass, appArgs.verbose)
}

//创建一个ActorSystem
val (actorSystem, _) = AkkaUtils.createActorSystem("driverClient",Utils.localHostName(),0,　　conf, new SecurityManager(conf))
//执行ClientActor的preStart方法和receive方法
actorSystem.actorOf(Props(classOf[ClientActor], driverArgs, conf))
//等待运行结束
actorSystem.awaitTermination()

//定义一个case class用来传递参数case class Greeting(who: String)
//定义Actor，比较重要的一个方法是receive方法，用来接收信息的
class GreetingActor extends Actor with ActorLogging {def receive = {case Greeting(who) ⇒ log.info("Hello " + who)}
}
//创建一个ActorSystem
val system = ActorSystem("MySystem")//给ActorSystem设置Actor
val greeter = system.actorOf(Props[GreetingActor], name = "greeter")//向greeter发送信息，用Greeting来传递
greeter ! Greeting("Charlie Parker")

override def preStart() = {    // 这里需要把master的地址转换成akka的地址，然后通过这个akka地址获得指定的actor    // 它的格式是"akka.tcp://%s@%s:%s/user/%s".format(systemName, host, port, actorName)    masterActor = context.actorSelection(Master.toAkkaUrl(driverArgs.master))// 把自身设置成远程生命周期的事件    context.system.eventStream.subscribe(self, classOf[RemotingLifecycleEvent])driverArgs.cmd match {case "launch" =>// 此处省略100个字        val mainClass = "org.apache.spark.deploy.worker.DriverWrapper"// 此处省略100个字// 向master发送提交Driver的请求，把driverDescription传过去，RequestSubmitDriver前面说过了，是个case classmasterActor ! RequestSubmitDriver(driverDescription)case "kill" =>val driverId = driverArgs.driverIdval killFuture = masterActor ! RequestKillDriver(driverId)}
}

val (actorSystem, boundPort) = AkkaUtils.createActorSystem(systemName, host, port, conf = conf, 　　securityManager = securityMgr)
val actor = actorSystem.actorOf(Props(classOf[Master], host, boundPort, webUiPort, securityMgr), actorName)
val timeout = AkkaUtils.askTimeout(conf)
val respFuture = actor.ask(RequestWebUIPort)(timeout)
val resp = Await.result(respFuture, timeout).asInstanceOf[WebUIPortResponse]

    case RequestSubmitDriver(description) => {val driver = createDriver(description)persistenceEngine.addDriver(driver)waitingDrivers += driverdrivers.add(driver)        // 调度schedule()// 告诉client，提交成功了，把driver.id告诉它sender ! SubmitDriverResponse(true, Some(driver.id), s"Driver successfully submitted as ${driver.id}")}

private def schedule() {if (state != RecoveryState.ALIVE) { return }// 首先调度Driver程序，从workers里面随机抽一些出来val shuffledWorkers = Random.shuffle(workers)     for (worker <- shuffledWorkers if worker.state == WorkerState.ALIVE) {for (driver <- waitingDrivers) {        // 判断内存和cpu够不够，够的就执行了哈        if (worker.memoryFree >= driver.desc.mem && worker.coresFree >= driver.desc.cores) {launchDriver(worker, driver)waitingDrivers -= driver}}}// 这里是按照先进先出的，spreadOutApps是由spark.deploy.spreadOut参数来决定的，默认是true    if (spreadOutApps) {// 遍历一下app      for (app <- waitingApps if app.coresLeft > 0) {        // canUse里面判断了worker的内存是否够用，并且该worker是否已经包含了该app的Executorval usableWorkers = workers.toArray.filter(_.state == WorkerState.ALIVE).filter(canUse(app, _)).sortBy(_.coresFree).reverseval numUsable = usableWorkers.lengthval assigned = new Array[Int](numUsable)         // 记录每个节点的核心数var toAssign = math.min(app.coresLeft, usableWorkers.map(_.coresFree).sum)var pos = 0        // 遍历直到分配结束while (toAssign > 0) {          // 从0开始遍历可用的work，如果可用的cpu减去已经分配的>0,就可以分配给它          if (usableWorkers(pos).coresFree - assigned(pos) > 0) {toAssign -= 1            // 这个位置的work的可分配的cpu数+1            assigned(pos) += 1}pos = (pos + 1) % numUsable}// 给刚才标记的worker分配任务        for (pos <- 0 until numUsable) {if (assigned(pos) > 0) {val exec = app.addExecutor(usableWorkers(pos), assigned(pos))launchExecutor(usableWorkers(pos), exec)app.state = ApplicationState.RUNNING}}}} else {// 这种方式和上面的方式的区别是，这种方式尽可能用少量的节点来完成这个任务for (worker <- workers if worker.coresFree > 0 && worker.state == WorkerState.ALIVE) {for (app <- waitingApps if app.coresLeft > 0) {          // 判断条件是worker的内存比app需要的内存多          if (canUse(app, worker)) {val coresToUse = math.min(worker.coresFree, app.coresLeft)if (coresToUse > 0) {val exec = app.addExecutor(worker, coresToUse)launchExecutor(worker, exec)app.state = ApplicationState.RUNNING}}}}}}

  def launchDriver(worker: WorkerInfo, driver: DriverInfo) {logInfo("Launching driver " + driver.id + " on worker " + worker.id)worker.addDriver(driver)driver.worker = Some(worker)worker.actor ! LaunchDriver(driver.id, driver.desc)driver.state = DriverState.RUNNING}

  def launchExecutor(worker: WorkerInfo, exec: ExecutorInfo) {logInfo("Launching executor " + exec.fullId + " on worker " + worker.id)worker.addExecutor(exec)worker.actor ! LaunchExecutor(masterUrl,exec.application.id, exec.id, exec.application.desc, exec.cores, exec.memory)exec.application.driver ! ExecutorAdded(exec.id, worker.id, worker.hostPort, exec.cores, exec.memory)}

case LaunchDriver(driverId, driverDesc) => {logInfo(s"Asked to launch driver $driverId")val driver = new DriverRunner(driverId, workDir, sparkHome, driverDesc, self, akkaUrl)drivers(driverId) = driverdriver.start()coresUsed += driverDesc.coresmemoryUsed += driverDesc.mem
}