检索领导者网关和端口失败

我试图从几天前开始使用1.0.0-rc3在每作业YARN会话的上下文中工作JobManager HA，并且遇到有关具有多个网络接口的任务管理器的问题。

手动终止作业管理器进程后，新分配的第二个作业管理器上的jobmanager.log读取：

2016-03-02 18:01:09,635 WARN  Remoting                                                   
  - Tried to associate with unreachable remote address [akka.tcp://[email protected]:34811].
Address is now gated for 5000 ms, all messages to this address will be delivered to dead letters.
Reason: Connection refused: /10.127.68.136:34811
2016-03-02 18:01:09,644 WARN  org.apache.flink.runtime.webmonitor.JobManagerRetriever    
  - Failed to retrieve leader gateway and port.
akka.actor.ActorNotFound: Actor not found for: ActorSelection[Anchor(akka.tcp://[email protected]:34811/),
Path(/user/jobmanager)]
    at akka.actor.ActorSelection$$anonfun$resolveOne$1.apply(ActorSelection.scala:65)
    at akka.actor.ActorSelection$$anonfun$resolveOne$1.apply(ActorSelection.scala:63)
    at scala.concurrent.impl.CallbackRunnable.run(Promise.scala:32)
    at akka.dispatch.BatchingExecutor$Batch$$anonfun$run$1.processBatch$1(BatchingExecutor.scala:67)
    at akka.dispatch.BatchingExecutor$Batch$$anonfun$run$1.apply$mcV$sp(BatchingExecutor.scala:82)
    at akka.dispatch.BatchingExecutor$Batch$$anonfun$run$1.apply(BatchingExecutor.scala:59)
    at akka.dispatch.BatchingExecutor$Batch$$anonfun$run$1.apply(BatchingExecutor.scala:59)
    at scala.concurrent.BlockContext$.withBlockContext(BlockContext.scala:72)
    at akka.dispatch.BatchingExecutor$Batch.run(BatchingExecutor.scala:58)
    at akka.dispatch.ExecutionContexts$sameThreadExecutionContext$.unbatchedExecute(Future.scala:74)
    at akka.dispatch.BatchingExecutor$class.execute(BatchingExecutor.scala:110)
    at akka.dispatch.ExecutionContexts$sameThreadExecutionContext$.execute(Future.scala:73)
    at scala.concurrent.impl.CallbackRunnable.executeWithValue(Promise.scala:40)
    at scala.concurrent.impl.Promise$DefaultPromise.tryComplete(Promise.scala:248)
    at akka.pattern.PromiseActorRef.$bang(AskSupport.scala:267)
    at akka.actor.EmptyLocalActorRef.specialHandle(ActorRef.scala:508)
    at akka.actor.DeadLetterActorRef.specialHandle(ActorRef.scala:541)
    at akka.actor.DeadLetterActorRef.$bang(ActorRef.scala:531)
    at akka.remote.RemoteActorRefProvider$RemoteDeadLetterActorRef.$bang(RemoteActorRefProvider.scala:87)
    at akka.remote.EndpointWriter.postStop(Endpoint.scala:561)
    at akka.actor.Actor$class.aroundPostStop(Actor.scala:475)
    at akka.remote.EndpointActor.aroundPostStop(Endpoint.scala:415)
    at akka.actor.dungeon.FaultHandling$class.akka$actor$dungeon$FaultHandling$$finishTerminate(FaultHandling.scala:210)
    at akka.actor.dungeon.FaultHandling$class.terminate(FaultHandling.scala:172)
    at akka.actor.ActorCell.terminate(ActorCell.scala:369)
    at akka.actor.ActorCell.invokeAll$1(ActorCell.scala:462)
    at akka.actor.ActorCell.systemInvoke(ActorCell.scala:478)
    at akka.dispatch.Mailbox.processAllSystemMessages(Mailbox.scala:279)
    at akka.dispatch.Mailbox.run(Mailbox.scala:220)
    at akka.dispatch.Mailbox.exec(Mailbox.scala:231)
    at scala.concurrent.forkjoin.ForkJoinTask.doExec(ForkJoinTask.java:260)
    at scala.concurrent.forkjoin.ForkJoinPool$WorkQueue.runTask(ForkJoinPool.java:1339)
    at scala.concurrent.forkjoin.ForkJoinPool.runWorker(ForkJoinPool.java:1979)
    at scala.concurrent.forkjoin.ForkJoinWorkerThread.run(ForkJoinWorkerThread.java:107)

未找到IP的位置是旧的作业管理器中。到目前为止，这是预期的行为吗？

然后，问题出现在新的任务管理器上，该任务管理器也尝试不成功地连接到旧的任务管理器。ZooKeeperLeaderRetrievalService通过可用的网络接口开始循环，如相关taskmanager.log中所示：

2016-03-02 18:01:13,636 INFO  org.apache.flink.runtime.leaderretrieval.ZooKeeperLeaderRetrievalService
 - Starting ZooKeeperLeaderRetrievalService.
2016-03-02 18:01:13,646 INFO  org.apache.flink.runtime.util.LeaderRetrievalUtils         
  - Trying to select the network interface and address to use by connecting to the leading
JobManager.
2016-03-02 18:01:13,646 INFO  org.apache.flink.runtime.util.LeaderRetrievalUtils         
  - TaskManager will try to connect for 10000 milliseconds before falling back to heuristics
2016-03-02 18:01:13,712 INFO  org.apache.flink.runtime.net.ConnectionUtils               
  - Retrieved new target address /10.127.68.136:34811.
2016-03-02 18:01:14,079 INFO  org.apache.flink.runtime.net.ConnectionUtils               
  - Trying to connect to address /10.127.68.136:34811
2016-03-02 18:01:14,082 INFO  org.apache.flink.runtime.net.ConnectionUtils               
  - Failed to connect from address 'task.manager.eth0.hostname.com/10.127.68.136': Connection
refused
2016-03-02 18:01:14,082 INFO  org.apache.flink.runtime.net.ConnectionUtils               
  - Failed to connect from address '/10.127.68.136': Connection refused
2016-03-02 18:01:14,082 INFO  org.apache.flink.runtime.net.ConnectionUtils               
  - Failed to connect from address '/10.120.193.110': Connection refused
2016-03-02 18:01:14,082 INFO  org.apache.flink.runtime.net.ConnectionUtils               
  - Failed to connect from address '/10.127.68.136': Connection refused
2016-03-02 18:01:14,083 INFO  org.apache.flink.runtime.net.ConnectionUtils               
  - Failed to connect from address '/127.0.0.1': Connection refused
2016-03-02 18:01:14,083 INFO  org.apache.flink.runtime.net.ConnectionUtils               
  - Failed to connect from address '/10.120.193.110': Connection refused
2016-03-02 18:01:14,083 INFO  org.apache.flink.runtime.net.ConnectionUtils               
  - Failed to connect from address '/10.127.68.136': Connection refused
2016-03-02 18:01:14,083 INFO  org.apache.flink.runtime.net.ConnectionUtils               
  - Failed to connect from address '/127.0.0.1': Connection refused

经过五次重复之后，任务管理器停止尝试检索领导者，并且从现在开始使用HEURISTIC策略最终使用eth1（10.120.193.110）：

2016-03-02 18:01:23,650 INFO  org.apache.flink.runtime.leaderretrieval.ZooKeeperLeaderRetrievalService
 - Stopping ZooKeeperLeaderRetrievalService.
2016-03-02 18:01:23,655 INFO  org.apache.zookeeper.ClientCnxn                            
  - EventThread shut down
2016-03-02 18:01:23,655 INFO  org.apache.zookeeper.ZooKeeper                             
  - Session: 0x25229757cff035b closed
2016-03-02 18:01:23,664 INFO  org.apache.flink.runtime.taskmanager.TaskManager           
  - TaskManager will use hostname/address 'task.manager.eth1.hostname.com' (10.120.193.110)
for communication.

发现新的jobmanager之后，taskmanager可以使用eth1在jobmanager上注册。问题是无法连接到eth1。因此，flink应该始终使用eth0。我们稍后看到的异常是：

java.io.IOException: Connecting the channel failed: Connecting to remote task manager + 'other.task.manager.eth1.hostname/10.120.193.111:46620'
has failed. This might indicate that the remote task manager has been lost.
    at org.apache.flink.runtime.io.network.netty.PartitionRequestClientFactory$ConnectingChannel.waitForChannel(PartitionRequestClientFactory.java:196)
    at org.apache.flink.runtime.io.network.netty.PartitionRequestClientFactory$ConnectingChannel.access$000(PartitionRequestClientFactory.java:131)
    at org.apache.flink.runtime.io.network.netty.PartitionRequestClientFactory.createPartitionRequestClient(PartitionRequestClientFactory.java:83)
    at org.apache.flink.runtime.io.network.netty.NettyConnectionManager.createPartitionRequestClient(NettyConnectionManager.java:60)
    at org.apache.flink.runtime.io.network.partition.consumer.RemoteInputChannel.requestSubpartition(RemoteInputChannel.java:115)
    at org.apache.flink.runtime.io.network.partition.consumer.SingleInputGate.requestPartitions(SingleInputGate.java:388)
    at org.apache.flink.runtime.io.network.partition.consumer.SingleInputGate.getNextBufferOrEvent(SingleInputGate.java:411)
    at org.apache.flink.streaming.runtime.io.BarrierBuffer.getNextNonBlocked(BarrierBuffer.java:108)
    at org.apache.flink.streaming.runtime.io.StreamInputProcessor.processInput(StreamInputProcessor.java:175)
    at org.apache.flink.streaming.runtime.tasks.OneInputStreamTask.run(OneInputStreamTask.java:65)
    at org.apache.flink.streaming.runtime.tasks.StreamTask.invoke(StreamTask.java:224)
    at org.apache.flink.runtime.taskmanager.Task.run(Task.java:559)
    at java.lang.Thread.run(Thread.java:744)

根本原因似乎是网络接口选择仍在使用旧的Jobmanager位置，因此无法选择正确的接口。特别是，在HEURISTIC和SLOW策略之间，网络接口上的迭代顺序似乎有所不同，这会导致选择了错误的接口。