【Spark三十一】SparkSubmit兼谈Spark的集群模式

Spark的部署和Application提交方式会按照Cluster Manager和Deploy Mode进行划分。。。以前只知道Spark的Cluster部署分为Standalone、YARN和Mesos。。。

 

关于Cluster Manager和Deploy Mode的组合在SparkSubmit.scala的createLaunchEnv中有比较详细的逻辑。

Cluster Manager基本上有Standalone，YARN和Mesos三种情况，说明Cluster Manager用来指明集群的资源管理器。这就是说不管是Client还是Cluster部署方式(deployMode的两种可能)，都会使用它们做集群管理器，也就是说Client也是一种集群部署方式？？？

 

 

  /**
   * @return a tuple containing
   *           (1) the arguments for the child process,
   *           (2) a list of classpath entries for the child,
   *           (3) a list of system properties and env vars, and
   *           (4) the main class for the child
   */
//createLaunchEnv的方法返回值
//1.子进程的参数,字符串数组，ArrayBuffer[String]
//2.子进程JVM的classpath路径列表,字符串数组，ArrayBuffer[String]
//3.子进程的系统变量和环境变量 ，HashMap类型
//4.子进程JVM的main class
private[spark] def createLaunchEnv(args: SparkSubmitArguments)
      : (ArrayBuffer[String], ArrayBuffer[String], Map[String, String], String) = {

    // Values to return
    val childArgs = new ArrayBuffer[String]()
    val childClasspath = new ArrayBuffer[String]()
    val sysProps = new HashMap[String, String]()
    var childMainClass = ""

    // Set the cluster manager
    //集群管理器，这里指定了四种：YARN，STANDALONE，MESON和LOCAL
    //需要注意的是，为什么LOCAL也是一种集群管理器，它的集群含义是什么？
    //根据args.master参数值决定clusterManager，注意，区分大小写
    //这里只检查master是否以yarn, spark, mesos或者local开头，实际中，以yarn开头的master值可能是yarn-client，yarn-cluster，yarn-standalone，所以代码后面对master做了更进一步的检查
    val clusterManager: Int = args.master match {
      case m if m.startsWith("yarn") => YARN
      case m if m.startsWith("spark") => STANDALONE
      case m if m.startsWith("mesos") => MESOS
      case m if m.startsWith("local") => LOCAL
      //如果master不以这四个开头，提示出错信息是***Master***必须以yarn, spark, mesos, or local开头
      case _ => printErrorAndExit("Master must start with yarn, spark, mesos, or local"); -1
    }

    // Set the deploy mode; default is client mode
    //设置部署模式，有两种模式，client和cluster模式
    //如果没有设置deployMode(取值null)，则默认认为是client模式
    var deployMode: Int = args.deployMode match {
      case "client" | null => CLIENT
      case "cluster" => CLUSTER
      case _ => printErrorAndExit("Deploy mode must be either client or cluster"); -1
    }

    // Because "yarn-cluster" and "yarn-client" encapsulate both the master
    // and deploy mode, we have some logic to infer the master and deploy mode
    // from each other if only one is specified, or exit early if they are at odds.

    ///因为yarn-cluster和yarn-client封装了master和deployMode, 这里对yarn-cluster和yarn-client两种集群管理器和部署模式的组合进行了特殊处理
    ///只要知道一个就可以推倒出另一个we have some logic to infer the master and deploy mode from each other
    //如果args.master以yarn开头(导致clusterManager == YARN为true)
    if (clusterManager == YARN) { ///如果clusterManager是YARN，那么说明master是以yarn开头
      if (args.master == "yarn-standalone") { //如果master是yarn-standalone,则提示说master的值yarn-standalone已经不推荐使用，
        printWarning("\"yarn-standalone\" is deprecated. Use \"yarn-cluster\" instead.")
        args.master = "yarn-cluster" ///将master值改为yarn-cluster
      }
      (args.master, args.deployMode) match { ///对args.master和args.deployMode的组合进行检查和整理
        case ("yarn-cluster", null) => ///如果args.deployMode为null，那么在前面的逻辑中，deployMode的值是CLIENT，
          deployMode = CLUSTER //根据args.master的取值yarn-cluster，将deployMode的值改为CLUSTER
        case ("yarn-cluster", "client") =>  ////如果args.master, args.deployMode是yarn-cluster，client，那么认为这两个值是冲突的，退出
          printErrorAndExit("Client deploy mode is not compatible with master \"yarn-cluster\"")
        case ("yarn-client", "cluster") =>  ////如果args.master, args.deployMode是yarn-client，cluster，那么认为这两个值是冲突的，退出
          printErrorAndExit("Cluster deploy mode is not compatible with master \"yarn-client\"")
        case (_, mode) => 
          ///其他情况下，将args.master改为如下值(代码执行到这里的前提是args.master以yarn开头)
          ///通过前面的代码，mode只有三种可能的取值：null，cluster和client
          //如果mode是client，那么args.master是yarn-client
          //如果mode是null，那么args.master是yarn-client
          //如果mode是cluster，args.master只能是yarn-cluster，因为(yarn-client，cluster)情况在前面判断过了
          args.master = "yarn-" + Option(mode).getOrElse("client")
      }

      // Make sure YARN is included in our build if we're trying to use it
      //在YARN的集群模式下检查Spark安装包的完整性，为什么一定要检查org.apache.spark.deploy.yarn.Client类的存在？因为后面的代码中会通过反射调用Client的main方法
      //注意，Client是的包名中有个yarn，说明这个Client用于跟YARN集群模式有关
      if (!Utils.classIsLoadable("org.apache.spark.deploy.yarn.Client") && !Utils.isTesting) {
        printErrorAndExit(
          "Could not load YARN classes. " +
          "This copy of Spark may not have been compiled with YARN support.")
      }
    }

    // The following modes are not supported or applicable
    ///对clusterManager和deployMode不支持的组合进行检查
    (clusterManager, deployMode) match {
      case (MESOS, CLUSTER) =>///MESOS和CLUSTER不共存
        printErrorAndExit("Cluster deploy mode is currently not supported for Mesos clusters.")
      case (_, CLUSTER) if args.isPython =>///Python应用程序不支持集群部署模式
        printErrorAndExit("Cluster deploy mode is currently not supported for python applications.")
      case (_, CLUSTER) if isShell(args.primaryResource) => ///Spark Shell不支持集群部署方式？通过Spark Shell提交Application是一种Client方式？？？
        printErrorAndExit("Cluster deploy mode is not applicable to Spark shells.")
      case _ =>
    }

    // If we're running a python app, set the main class to our specific python runner
    if (args.isPython) { ///如果python应用程序，我表示不关心，忽略之
      if (args.primaryResource == PYSPARK_SHELL) {
        args.mainClass = "py4j.GatewayServer"
        args.childArgs = ArrayBuffer("--die-on-broken-pipe", "0")
      } else {
        // If a python file is provided, add it to the child arguments and list of files to deploy.
        // Usage: PythonAppRunner <main python file> <extra python files> [app arguments]
        args.mainClass = "org.apache.spark.deploy.PythonRunner"
        args.childArgs = ArrayBuffer(args.primaryResource, args.pyFiles) ++ args.childArgs
        args.files = mergeFileLists(args.files, args.primaryResource)
      }
      args.files = mergeFileLists(args.files, args.pyFiles)
      if (args.pyFiles != null) {
        sysProps("spark.submit.pyFiles") = args.pyFiles
      }
    }

    // Special flag to avoid deprecation warnings at the client
    sysProps("SPARK_SUBMIT") = "true" ///SPARK_SUBMIT这个参数用来做什么的？注意sysProps是个Scala的HashMap

    // A list of rules to map each argument to system properties or command-line options in
    // each deploy mode; we iterate through these below
    ///这是在做什么？？一系列规则（用于在各种部署模式下，把每个argument映射为system properties或者command-line options的规则）
    val options = List[OptionAssigner](

      // All cluster managers
      ///sysProp是指定参数的函数参数传值，此处没有给clOption赋值
      OptionAssigner(args.master, ALL_CLUSTER_MGRS, ALL_DEPLOY_MODES, sysProp = "spark.master"),
      OptionAssigner(args.name, ALL_CLUSTER_MGRS, ALL_DEPLOY_MODES, sysProp = "spark.app.name"),

      ///args.jars指定了提交的Application的jars相关的东西
      OptionAssigner(args.jars, ALL_CLUSTER_MGRS, CLIENT, sysProp = "spark.jars"),
      OptionAssigner(args.driverMemory, ALL_CLUSTER_MGRS, CLIENT,
        sysProp = "spark.driver.memory"),
      OptionAssigner(args.driverExtraClassPath, ALL_CLUSTER_MGRS, ALL_DEPLOY_MODES,
        sysProp = "spark.driver.extraClassPath"),
      OptionAssigner(args.driverExtraJavaOptions, ALL_CLUSTER_MGRS, ALL_DEPLOY_MODES,
        sysProp = "spark.driver.extraJavaOptions"),
      OptionAssigner(args.driverExtraLibraryPath, ALL_CLUSTER_MGRS, ALL_DEPLOY_MODES,
        sysProp = "spark.driver.extraLibraryPath"),

      // Standalone cluster only
      ///只适用于Standalone集群，属性无需以spark开头，因为都在spark中
      OptionAssigner(args.jars, STANDALONE, CLUSTER, sysProp = "spark.jars"),
      OptionAssigner(args.driverMemory, STANDALONE, CLUSTER, clOption = "--memory"), //命令行赋值
      OptionAssigner(args.driverCores, STANDALONE, CLUSTER, clOption = "--cores"),

      // Yarn client only
      //只适用于Yarn client，因为由Yarn进行管理，属性名都带有spark
      OptionAssigner(args.queue, YARN, CLIENT, sysProp = "spark.yarn.queue"),
      OptionAssigner(args.numExecutors, YARN, CLIENT, sysProp = "spark.executor.instances"),
      OptionAssigner(args.executorCores, YARN, CLIENT, sysProp = "spark.executor.cores"),
      OptionAssigner(args.files, YARN, CLIENT, sysProp = "spark.yarn.dist.files"),
      OptionAssigner(args.archives, YARN, CLIENT, sysProp = "spark.yarn.dist.archives"),

      // Yarn cluster only
      ///只适用于Yarn cluster，为命令行选项赋值
      OptionAssigner(args.name, YARN, CLUSTER, clOption = "--name"),
      OptionAssigner(args.driverMemory, YARN, CLUSTER, clOption = "--driver-memory"),
      OptionAssigner(args.queue, YARN, CLUSTER, clOption = "--queue"),
      OptionAssigner(args.numExecutors, YARN, CLUSTER, clOption = "--num-executors"),
      OptionAssigner(args.executorMemory, YARN, CLUSTER, clOption = "--executor-memory"),
      OptionAssigner(args.executorCores, YARN, CLUSTER, clOption = "--executor-cores"),
      OptionAssigner(args.files, YARN, CLUSTER, clOption = "--files"),
      OptionAssigner(args.archives, YARN, CLUSTER, clOption = "--archives"),
      OptionAssigner(args.jars, YARN, CLUSTER, clOption = "--addJars"),

      // Other options
      OptionAssigner(args.executorMemory, STANDALONE | MESOS | YARN, ALL_DEPLOY_MODES,
        sysProp = "spark.executor.memory"),
      OptionAssigner(args.totalExecutorCores, STANDALONE | MESOS, ALL_DEPLOY_MODES,
        sysProp = "spark.cores.max"),
      OptionAssigner(args.files, LOCAL | STANDALONE | MESOS, ALL_DEPLOY_MODES,
        sysProp = "spark.files")
    )
    
    ///如果CLIENT部署模式，使用application的jar中指定的可运行的Main Class
    // In client mode, launch the application main class directly
    // In addition, add the main application jar and any added jars (if any) to the classpath
    if (deployMode == CLIENT) {
      childMainClass = args.mainClass
      if (isUserJar(args.primaryResource)) {
        childClasspath += args.primaryResource
      }
      if (args.jars != null) { childClasspath ++= args.jars.split(",") }
      if (args.childArgs != null) { childArgs ++= args.childArgs }
    }


    // Map all arguments to command-line options or system properties for our chosen mode
    for (opt <- options) {
      if (opt.value != null &&
          (deployMode & opt.deployMode) != 0 &&
          (clusterManager & opt.clusterManager) != 0) {
        if (opt.clOption != null) { childArgs += (opt.clOption, opt.value) }
        if (opt.sysProp != null) { sysProps.put(opt.sysProp, opt.value) }
      }
    }

    // Add the application jar automatically so the user doesn't have to call sc.addJar
    // For YARN cluster mode, the jar is already distributed on each node as "app.jar"
    // For python files, the primary resource is already distributed as a regular file
    val isYarnCluster = clusterManager == YARN && deployMode == CLUSTER
    if (!isYarnCluster && !args.isPython) {
      var jars = sysProps.get("spark.jars").map(x => x.split(",").toSeq).getOrElse(Seq.empty)
      if (isUserJar(args.primaryResource)) {
        jars = jars ++ Seq(args.primaryResource)
      }
      sysProps.put("spark.jars", jars.mkString(","))
    }

    // In standalone-cluster mode, use Client as a wrapper around the user class
    if (clusterManager == STANDALONE && deployMode == CLUSTER) {
      childMainClass = "org.apache.spark.deploy.Client"
      if (args.supervise) {
        childArgs += "--supervise"
      }
      childArgs += "launch"
      childArgs += (args.master, args.primaryResource, args.mainClass)
      if (args.childArgs != null) {
        childArgs ++= args.childArgs
      }
    }

    // In yarn-cluster mode, use yarn.Client as a wrapper around the user class
    ///在Yarn
    if (isYarnCluster) {
      childMainClass = "org.apache.spark.deploy.yarn.Client"
      if (args.primaryResource != SPARK_INTERNAL) {
        childArgs += ("--jar", args.primaryResource)
      }
      childArgs += ("--class", args.mainClass)
      if (args.childArgs != null) {
        args.childArgs.foreach { arg => childArgs += ("--arg", arg) }
      }
    }

    // Load any properties specified through --conf and the default properties file
    for ((k, v) <- args.sparkProperties) {
      sysProps.getOrElseUpdate(k, v)
    }

    // Ignore invalid spark.driver.host in cluster modes.
    if (deployMode == CLUSTER) {
      sysProps -= ("spark.driver.host")
    }

    // Resolve paths in certain spark properties
    val pathConfigs = Seq(
      "spark.jars",
      "spark.files",
      "spark.yarn.jar",
      "spark.yarn.dist.files",
      "spark.yarn.dist.archives")
    pathConfigs.foreach { config =>
      // Replace old URIs with resolved URIs, if they exist
      sysProps.get(config).foreach { oldValue =>
        sysProps(config) = Utils.resolveURIs(oldValue)
      }
    }

    // Resolve and format python file paths properly before adding them to the PYTHONPATH.
    // The resolving part is redundant in the case of --py-files, but necessary if the user
    // explicitly sets `spark.submit.pyFiles` in his/her default properties file.
    sysProps.get("spark.submit.pyFiles").foreach { pyFiles =>
      val resolvedPyFiles = Utils.resolveURIs(pyFiles)
      val formattedPyFiles = PythonRunner.formatPaths(resolvedPyFiles).mkString(",")
      sysProps("spark.submit.pyFiles") = formattedPyFiles
    }

    (childArgs, childClasspath, sysProps, childMainClass)
  }

 

OptionAssigner

 

/**
 * Provides an indirection layer for passing arguments as system properties or flags to
 * the user's driver program or to downstream launcher tools.
 */
private[spark] case class OptionAssigner(
    value: String,
    clusterManager: Int,
    deployMode: Int,
    clOption: String = null,
    sysProp: String = null)

 

 

 

 

 

 

 

 

 

 

 

 

 

评论

1 楼 wjhdtx 2015-05-28  

拜读了楼主好多spark的文章，受益匪浅，想请教两个问题：

①对于yarn或者standalone来说，楼主怎么考虑如何选哪种？

②如果选择了standalone，那么是选择cluster还是client？从哪些方面考量？