Question

我正在尝试设置flink以从检查点恢复。在大多数情况下，这似乎可行，但是在将其部署到我们的临时环境大约一周后，由于尝试启动作业的“作业主”时超时，作业管理器已开始崩溃循环。

我使用的是在Zookeeper 3.4.9-1757313中以高可用性模式部署的flink 1.7.2，仅仅是为了方便检查点恢复。我在kubernetes上仅部署了一个有状态的作业管理器。一定是导致服务器崩溃的原因，并且在重新启动服务器时，它似乎无法启动（大概）已恢复作业的作业主服务器的代码。

我以前见过一次，清除了所有flink zookeeper条目（zk cli中的rmr /flink），然后重新启动flink集群可“修复”该问题。

这是flink配置

    blob.server.port: 6124
    blob.storage.directory: s3://...
    high-availability: zookeeper
    high-availability.zookeeper.quorum: zookeeper:2181
    high-availability.zookeeper.path.root: /flink
    high-availability.storageDir: s3://...
    high-availability.jobmanager.port: 6070
    jobmanager.archive.fs.dir: s3://...
    state.backend: rocksdb
    state.backend.fs.checkpointdir: s3://...
    state.checkpoints.dir: s3://...
    state.checkpoints.num-retained: 2
    web.log.path: /var/log/flink.log
    web.upload.dir: /var/flink-recovery/flink-web-upload
    zookeeper.sasl.disable: true
    s3.access-key: __S3_ACCESS_KEY_ID__
    s3.secret-key: __S3_SECRET_KEY__

这是flink-jobmaster状态集上的容器端口：

ports:
- containerPort: 8081
  name: ui
- containerPort: 6123
  name: rpc
- containerPort: 6124
  name: blob
- containerPort: 6125
  name: query
- containerPort: 9249
  name: prometheus
- containerPort: 6070
  name: ha

我希望flink从s3中的检查点成功恢复，但是作业管理器在启动时会崩溃，并带有以下堆栈跟踪：

2019-06-18 14:02:05,123 ERROR org.apache.flink.runtime.entrypoint.ClusterEntrypoint         - Fatal error occurred in the cluster entrypoint.
org.apache.flink.util.FlinkException: JobMaster for job f13131ca883d6cf92f69a52cff4f1017 failed.
    at org.apache.flink.runtime.dispatcher.Dispatcher.jobMasterFailed(Dispatcher.java:759)
    at org.apache.flink.runtime.dispatcher.Dispatcher.lambda$startJobManagerRunner$6(Dispatcher.java:339)
    at java.util.concurrent.CompletableFuture.uniWhenComplete(CompletableFuture.java:760)
    at java.util.concurrent.CompletableFuture$UniWhenComplete.tryFire(CompletableFuture.java:736)
    at java.util.concurrent.CompletableFuture$Completion.run(CompletableFuture.java:442)
    at org.apache.flink.runtime.rpc.akka.AkkaRpcActor.handleRunAsync(AkkaRpcActor.java:332)
    at org.apache.flink.runtime.rpc.akka.AkkaRpcActor.handleRpcMessage(AkkaRpcActor.java:158)
    at org.apache.flink.runtime.rpc.akka.FencedAkkaRpcActor.handleRpcMessage(FencedAkkaRpcActor.java:70)
    at org.apache.flink.runtime.rpc.akka.AkkaRpcActor.onReceive(AkkaRpcActor.java:142)
    at org.apache.flink.runtime.rpc.akka.FencedAkkaRpcActor.onReceive(FencedAkkaRpcActor.java:40)
    at akka.actor.UntypedActor$$anonfun$receive$1.applyOrElse(UntypedActor.scala:165)
    at akka.actor.Actor$class.aroundReceive(Actor.scala:502)
    at akka.actor.UntypedActor.aroundReceive(UntypedActor.scala:95)
    at akka.actor.ActorCell.receiveMessage(ActorCell.scala:526)
    at akka.actor.ActorCell.invoke(ActorCell.scala:495)
    at akka.dispatch.Mailbox.processMailbox(Mailbox.scala:257)
    at akka.dispatch.Mailbox.run(Mailbox.scala:224)
    at akka.dispatch.Mailbox.exec(Mailbox.scala:234)
    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)
Caused by: org.apache.flink.util.FlinkException: Could not start the job manager.
    at org.apache.flink.runtime.jobmaster.JobManagerRunner.lambda$verifyJobSchedulingStatusAndStartJobManager$2(JobManagerRunner.java:340)
    at java.util.concurrent.CompletableFuture.uniWhenComplete(CompletableFuture.java:760)
    at java.util.concurrent.CompletableFuture$UniWhenComplete.tryFire(CompletableFuture.java:736)
    at java.util.concurrent.CompletableFuture$Completion.run(CompletableFuture.java:442)
    at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
    at java.util.concurrent.FutureTask.run(FutureTask.java:266)
    at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$201(ScheduledThreadPoolExecutor.java:180)
    at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:293)
    at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
    at java.lang.Thread.run(Thread.java:748)
Caused by: akka.pattern.AskTimeoutException: Ask timed out on [Actor[akka://flink/user/jobmanager_2#-806528277]] after [10000 ms]. Sender[null] sent message of type "org.apache.flink.runtime.rpc.messages.UnfencedMessage".
    at akka.pattern.PromiseActorRef$$anonfun$1.apply$mcV$sp(AskSupport.scala:604)
    at akka.actor.Scheduler$$anon$4.run(Scheduler.scala:126)
    at scala.concurrent.Future$InternalCallbackExecutor$.unbatchedExecute(Future.scala:601)
    at scala.concurrent.BatchingExecutor$class.execute(BatchingExecutor.scala:109)
    at scala.concurrent.Future$InternalCallbackExecutor$.execute(Future.scala:599)
    at akka.actor.LightArrayRevolverScheduler$TaskHolder.executeTask(LightArrayRevolverScheduler.scala:329)
    at akka.actor.LightArrayRevolverScheduler$$anon$4.executeBucket$1(LightArrayRevolverScheduler.scala:280)
    at akka.actor.LightArrayRevolverScheduler$$anon$4.nextTick(LightArrayRevolverScheduler.scala:284)
    at akka.actor.LightArrayRevolverScheduler$$anon$4.run(LightArrayRevolverScheduler.scala:236)
    ... 1 more

我真的很茫然。我对flink的内部运作了解不多，所以这个异常对我来说意义不大。任何线索将不胜感激。

编辑：我一直在浏览Flink源代码。当领导者试图从存储在zookeeper中的检查点信息中恢复其工作图后，将引发该异常。要弄清这种异常的确切来源，是很麻烦的，因为它们都包裹在期货和Akka中。我的猜测是，这是在工作经理启动JobMaster子流程以计划工作图之后发生的。有点猜测，但我认为作业管理器正在尝试从其JobMaster获取新作业的状态，但是JobMaster线程进入了死锁状态（也许它也可能死了，尽管我希望那时有堆栈跟踪信息）并且所以要求超时了。似乎是个真正的傻瓜。

注意：要求的UnfencedMessage用于在作业管理器中本地使用（这与接收方作为作业管理器例外）相符，因此我们可以消除JobMaster之间的网络配置错误和任务管理器。

Answer 1

我正在使用/jars/upload端点在flink上执行jars操作。当flink上载了太多jar时，似乎flink的性能会下降。所有端点，包括/jobs/<job_id>端点，都变得无响应。将工作图概述加载到flink UI中需要1-2分钟。我想象这个休息端点使用工作经理做的相同角色。我认为我一定已经到达了一个引爆点，开始引起超时。我将30个奇数的jar数量减少到了4个最新版本，flink再次响应。

尝试为检查点作业启动flink作业主服务器的超时

1 个答案: