Re: got Warn message - "the expected leader session ID did not equal the received leader session ID " when using LocalFlinkMiniCluster to interpret scala code

Posted by XiangWei Huang on
URL: http://deprecated-apache-flink-user-mailing-list-archive.369.s1.nabble.com/got-Warn-message-the-expected-leader-session-ID-did-not-equal-the-received-leader-session-ID-when-use-tp15542p15893.html

Hi Till,

I’ve found that a StandaloneMiniCluster doesn’t startup web fronted when it is running.so,how can i cancel a running job on it with restful method.

Cheers,

Till

在 2017年9月20日，15:43，Till Rohrmann <[hidden email]> 写道：

Hi XiangWei,

programmatically there is no nice tooling yet to cancel jobs on a dedicated cluster. What you can do is to use Flink's REST API to issue a cancel command [1]. You have to send a GET request to the target URL `/jobs/:jobid/cancel`. In the future we will improve the programmatic job control which will allow you to do these kind of things more easily.

[1] https://ci.apache.org/projects/flink/flink-docs-release-1.3/monitoring/rest_api.html#job-cancellation

Cheers,

Till

On Wed, Sep 20, 2017 at 5:46 AM, XiangWei Huang <[hidden email]> wrote:

Hi Till,

   

     Thanks for your answer,it worked when i use StandaloneMiniCluster,but another problem is that i can’t find a way to cancel

a running Flink job without shutting down the cluster,for LocalFlinkMiniCluster i can do  it with below code :

   for (job <- cluster.getCurrentlyRunningJobsJava()) {
      cluster.stopJob(job)
   }

   Is it possible to cancel a running Flink job without shutting down a StandaloneMiniCluster ?

Best Regards,

XiangWei

在 2017年9月14日，16:58，Till Rohrmann <[hidden email]> 写道：

Hi XiangWei,

the problem is that the LocalFlinkMiniCluster can no longer be used in combination with a RemoteExecutionEnvironment. The reason is that the LocalFlinkMiniCluster uses now an internal leader election service and assigns leader ids to its components. Since this is an internal service it is not possible to retrieve this information like it is the case with the ZooKeeper based leader election services.

Long story short, the Flink Scala shell currently does not work with a LocalFlinkMiniCluster and would have to be fixed to work properly together with a local execution environment. Until then, I recommend starting a local standalone cluster and let the code run there.

Cheers,
Till

​

On Wed, Sep 13, 2017 at 6:21 AM, XiangWei Huang <[hidden email]> wrote:

dear all,

Below is the code i execute:

import java.io._
import java.net.{URL, URLClassLoader}
import java.nio.charset.Charset
import java.util.Collections
import java.util.concurrent.atomic.AtomicBoolean

import com.netease.atom.common.util.logging.Logging
import com.netease.atom.interpreter.Code.Code
import com.netease.atom.interpreter.{Code, Interpreter, InterpreterResult, InterpreterUtils}
import io.netty.buffer._
import org.apache.flink.api.scala.FlinkILoop
import org.apache.flink.client.CliFrontend
import org.apache.flink.client.cli.CliFrontendParser
import org.apache.flink.client.program.ClusterClient
import org.apache.flink.configuration.{QueryableStateOptions, Configuration, ConfigConstants, GlobalConfiguration}
import org.apache.flink.runtime.akka.AkkaUtils
import org.apache.flink.runtime.minicluster.{StandaloneMiniCluster, LocalFlinkMiniCluster}

import scala.Console
import scala.beans.BeanProperty
import scala.collection.JavaConversions._
import scala.collection.mutable
import scala.collection.mutable.{ArrayBuffer, ListBuffer}
import scala.runtime.AbstractFunction0
import scala.tools.nsc.Settings
import scala.tools.nsc.interpreter.{IMain, JPrintWriter, Results}

class FlinkInterpreter extends Interpreter {
  private var bufferedReader: Option[BufferedReader] = None
  private var jprintWriter: JPrintWriter = _
  private val config = new Configuration;
  private var cluster: LocalFlinkMiniCluster = _
  @BeanProperty var imain: IMain = _
  @BeanProperty var flinkILoop: FlinkILoop = _
  private var out: ByteBufOutputStream = null
  private var outBuf: ByteBuf = null
  private var in: ByteBufInputStream = _
  private var isRunning: AtomicBoolean = new AtomicBoolean(false)

  override def isOpen: Boolean = {
    isRunning.get()
  }

  def startLocalMiniCluster(): (String, Int, LocalFlinkMiniCluster) = {
    config.toMap.toMap.foreach(println)
    config.setInteger(ConfigConstants.LOCAL_NUMBER_JOB_MANAGER, 1)
    config.setInteger(ConfigConstants.LOCAL_NUMBER_TASK_MANAGER, 1)
    config.setInteger(ConfigConstants.LOCAL_NUMBER_RESOURCE_MANAGER, 1)
    config.setInteger(ConfigConstants.JOB_MANAGER_IPC_PORT_KEY, 0)
    config.setBoolean(QueryableStateOptions.SERVER_ENABLE.key(), true)
    config.setBoolean(ConfigConstants.LOCAL_START_WEBSERVER, true)
    val localCluster = new LocalFlinkMiniCluster(config, false)
    localCluster.start(true)
    val port = AkkaUtils.getAddress(localCluster.jobManagerActorSystems.get.head).port
    println(s"Starting local Flink cluster (host: localhost,port: ${localCluster.getLeaderRPCPort}).\n")
    ("localhost", localCluster.getLeaderRPCPort, localCluster)
  }

 
  /**
   * Start flink cluster and create interpreter
   */
  override def open: Unit = {
    outBuf = ByteBufAllocator.DEFAULT.heapBuffer(20480)
    out = new ByteBufOutputStream(outBuf)
    in = new ByteBufInputStream(outBuf)
    //    val (host, port, yarnCluster) = deployNewYarnCluster(YarnConfig(Option(1), None, None, None, Option(1), None))
    val (host, port, localCluster) = startLocalMiniCluster()
    this.cluster = localCluster
    val conf = cluster.configuration
    println(s"Connecting to Flink cluster (host:$host,port:$port)...")
    flinkILoop = new FlinkILoop(host, port, conf, None)
    val settings = new Settings()
    settings.usejavacp.value = true
    settings.Yreplsync.value = true
    flinkILoop.settings_$eq(settings)
    flinkILoop.createInterpreter()
    imain = flinkILoop.intp
    FlinkInterpreter.ourClassloader = imain.classLoader
    val benv = flinkILoop.scalaBenv
    val senv = flinkILoop.scalaSenv
    benv.getConfig.disableSysoutLogging()
    senv.getConfig.disableSysoutLogging()
    // import libraries
    imain.interpret("import scala.tools.nsc.io._")
    //    imain.interpret("import Properties.userHome")
    imain.interpret("import scala.compat.Platform.EOL")
    imain.interpret("import org.apache.flink.api.scala._")
    imain.interpret("import org.apache.flink.api.common.functions._")
    isRunning.set(true)
  }

  override def interpret(line: String): InterpreterResult = {
    if (line == null || line.trim.length == 0) {
      return new InterpreterResult(Code.SUCCESS)
    }
    interpret(line.split("\n"))
  }

  /**
   * Interprete code
   * @param lines
   * @return
   */
  def interpret(lines: Array[String]): InterpreterResult = {
    val imain: IMain = getImain
    val linesToRun: Array[String] = new Array[String](lines.length + 1)
    for (i <- 0 until lines.length) {
      linesToRun(i) = lines(i)
    }
    linesToRun(lines.length) = "print(\"\")"
    System.setOut(new PrintStream(out))
    out.buffer().clear()
    var r: Code = null
    var incomplete: String = ""
    var inComment: Boolean = false
    for (l <- 0 until linesToRun.length) {
      val s: String = linesToRun(l)
      var continuation: Boolean = false
      if (l + 1 < linesToRun.length) {
        val nextLine: String = linesToRun(l + 1).trim
        if (nextLine.isEmpty ||
            nextLine.startsWith("//") ||
            nextLine.startsWith("}") ||
            nextLine.startsWith("object")) {
          continuation = true
        } else if (!inComment && nextLine.startsWith("/*")) {
          inComment = true
          continuation = true
        } else if (!inComment && nextLine.lastIndexOf("*/") >= 0) {
          inComment = false
          continuation = true
        } else if (nextLine.length > 1 &&
            nextLine.charAt(0) == '.' &&
            nextLine.charAt(1) != '.' &&
            nextLine.charAt(1) != '/') {
          continuation = true
        } else if (inComment) {
          continuation = true
        }
        if (continuation) {
          incomplete += s + "\n"
        }
      }
      if (!continuation) {
        val currentCommand: String = incomplete
        var res: Results.Result = null
        try {
          res = Console.withOut(System.out)(new AbstractFunction0[Results.Result] {
            override def apply() = {
              imain.interpret(currentCommand + s)
            }
          }.apply())
        } catch {
          case e: Exception =>
            logError("Interpreter Exception ", e)
            return new InterpreterResult(Code.ERROR, InterpreterUtils.getMostRelevantMessage(e))
        }
        r = getResultCode(res)
        if (r == Code.ERROR) {
          return new InterpreterResult(r, out.toString)
        } else if (r eq Code.INCOMPLETE) {
          incomplete += s + "\n"
        } else {
          incomplete = ""
        }
      }
    }

    if (r eq Code.INCOMPLETE) {
      return new InterpreterResult(r, "Incomplete expression")
    }
    else {
      return new InterpreterResult(r, out.buffer().toString(Charset.forName("utf-8")))
    }
  }

  private def getResultCode(r: Results.Result): Code = {
    if (r.isInstanceOf[Results.Success.type]) {
      return Code.SUCCESS
    }
    else if (r.isInstanceOf[Results.Incomplete.type]) {
      return Code.INCOMPLETE
    }
    else {
      return Code.ERROR
    }
  }

  }
}

object FlinkInterpreter extends Logging {
  var ourClassloader: ClassLoader = _

  def main(args: Array[String]): Unit = {
    val interpreter: FlinkInterpreter = new FlinkInterpreter
    val code =
      """
        |val dataStream = senv.fromElements(1,2,3,4,5)
        |dataStream.countWindowAll(2).sum(0).print()
        |senv.execute("My streaming program")
      """.stripMargin
    interpreter.open
    val result = interpreter.interpret(code)
  }
}

The error messages i got are:

…

…

...

[WARN] [17/09/13 12:04:52] [org.apache.flink.runtime.jobmanager.JobManager] Discard message LeaderSessionMessage(00000000-0000-0000-0000-000000000000,SubmitJob(JobGraph(jobId: 1b923a7a54c06ffa1c91d276a45be826),EXECUTION_RESULT_AND_STATE_CHANGES)) because the expected leader session ID 678ef53b-ff25-4298-b566-9c2d9e7371c7 did not equal the received leader session ID 00000000-0000-0000-0000-000000000000.

[INFO] [17/09/13 12:05:52] [org.apache.flink.runtime.client.JobSubmissionClientActor] Terminate JobClientActor.

[INFO] [17/09/13 12:05:52] [org.apache.flink.runtime.client.JobSubmissionClientActor] Disconnect from JobManager Actor[akka.tcp://flink@localhost:63522/user/jobmanager#82627940].

[INFO] [17/09/13 12:05:52] [akka.remote.RemoteActorRefProvider$RemotingTerminator] Shutting down remote daemon.

[INFO] [17/09/13 12:05:52] [akka.remote.RemoteActorRefProvider$RemotingTerminator] Remote daemon shut down; proceeding with flushing remote transports.

[INFO] [17/09/13 12:05:52] [akka.remote.RemoteActorRefProvider$RemotingTerminator] Remoting shut down.

org.apache.flink.client.program.ProgramInvocationException: The program execution failed: Couldn't retrieve the JobExecutionResult from the JobManager.

  at org.apache.flink.client.program.ClusterClient.run(ClusterClient.java:478)

  at org.apache.flink.client.program.StandaloneClusterClient.submitJob(StandaloneClusterClient.java:105)

  at org.apache.flink.client.program.ClusterClient.run(ClusterClient.java:442)

  at org.apache.flink.client.program.ClusterClient.run(ClusterClient.java:434)

  at org.apache.flink.streaming.api.environment.RemoteStreamEnvironment.executeRemotely(RemoteStreamEnvironment.java:212)

  at org.apache.flink.api.java.ScalaShellRemoteStreamEnvironment.executeRemotely(ScalaShellRemoteStreamEnvironment.java:87)

  at org.apache.flink.streaming.api.environment.RemoteStreamEnvironment.execute(RemoteStreamEnvironment.java:176)

  at org.apache.flink.streaming.api.scala.StreamExecutionEnvironment.execute(StreamExecutionEnvironment.scala:638)

  ... 34 elided

Caused by: org.apache.flink.runtime.client.JobExecutionException: Couldn't retrieve the JobExecutionResult from the JobManager.

  at org.apache.flink.runtime.client.JobClient.awaitJobResult(JobClient.java:309)

  at org.apache.flink.runtime.client.JobClient.submitJobAndWait(JobClient.java:396)

  at org.apache.flink.client.program.ClusterClient.run(ClusterClient.java:467)

  ... 41 more

Caused by: org.apache.flink.runtime.client.JobClientActorSubmissionTimeoutException: Job submission to the JobManager timed out. You may increase 'akka.client.timeout' in case the JobManager needs more time to configure and confirm the job submission.

  at org.apache.flink.runtime.client.JobSubmissionClientActor.handleCustomMessage(JobSubmissionClientActor.java:119)

  at org.apache.flink.runtime.client.JobClientActor.handleMessage(JobClientActor.java:251)

  at org.apache.flink.runtime.akka.FlinkUntypedActor.handleLeaderSessionID(FlinkUntypedActor.java:89)

  at org.apache.flink.runtime.akka.FlinkUntypedActor.onReceive(FlinkUntypedActor.java:68)

  at akka.actor.UntypedActor$$anonfun$receive$1.applyOrElse(UntypedActor.scala:167)

  at akka.actor.Actor$class.aroundReceive(Actor.scala:467)

  at akka.actor.UntypedActor.aroundReceive(UntypedActor.scala:97)

  at akka.actor.ActorCell.receiveMessage(ActorCell.scala:516)

  at akka.actor.ActorCell.invoke(ActorCell.scala:487)

  at akka.dispatch.Mailbox.processMailbox(Mailbox.scala:238)

  at akka.dispatch.Mailbox.run(Mailbox.scala:220)

  at akka.dispatch.ForkJoinExecutorConfigurator$AkkaForkJoinTask.exec(AbstractDispatcher.scala:397)

  at scala.concurrent.forkjoin.ForkJoinTask.doExec(ForkJoinTask.java:260)

  at scala.concurrent.forkjoin.ForkJoinPool$WorkQueue.runTask(ForkJoinPool.java:1339)

  at scala.concurrent.forkjoin.ForkJoinPool.runWorker(ForkJoinPool.java:1979)

  at scala.concurrent.forkjoin.ForkJoinWorkerThread.run(ForkJoinWorkerThread.java:107)