You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@tez.apache.org by "Bikas Saha (JIRA)" <ji...@apache.org> on 2014/10/08 21:54:33 UTC
[jira] [Commented] (TEZ-1643) DAGAppMaster kills DAG & shuts down,
when RM is restarted
[ https://issues.apache.org/jira/browse/TEZ-1643?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14164051#comment-14164051 ]
Bikas Saha commented on TEZ-1643:
---------------------------------
What can be done about this other than shutting down? If YARN's own AMRMclient has given up on the RM? Maybe we could add some retries but nothing would stop the AMRMClient from failing again. Without RM HA the client the RM will not be able to resync on the allocation quota/status and the RM (after restart) will ask all containers to be killed (including the AM).
We could continue to run the job with existing containers instead of failing the DAG and hope to finish some (or all work) while the RM is unavailable. Once the RM comes back we will be killed (and restarted).
In HA scenarios the client should wait much longer for the RM to come back up. So this jira may be a wont fix for non-HA cases.
> DAGAppMaster kills DAG & shuts down, when RM is restarted
> ---------------------------------------------------------
>
> Key: TEZ-1643
> URL: https://issues.apache.org/jira/browse/TEZ-1643
> Project: Apache Tez
> Issue Type: Bug
> Reporter: Rajesh Balamohan
> Priority: Critical
>
> Scenario:
> 1. Start a long running job
> 2. Kill RM (recovery is enabled in RM. No RM-HA configured)
> 3. AMRMClientAsyncImpl$HeartbeatThread throws error (EOFException) which internally causes the appmaster to kill DAG.
> 2014-10-08 02:24:06,705 INFO [IPC Server handler 6 on 55291] org.apache.tez.dag.app.dag.impl.TaskImpl: TaskAttempt:attempt_1412734988643_0001_1_00_000000_0 sent events: (0-1)
> 2014-10-08 02:24:12,255 ERROR [AMRM Heartbeater thread] org.apache.hadoop.yarn.client.api.async.impl.AMRMClientAsyncImpl: Exception on heartbeat
> java.io.IOException: Failed on local exception: java.io.EOFException; Host Details : local host is: "m-tez-uns-try-3/1.1.1.1"; destination host is: "
> m-tez-uns-try-3":8030;
> at org.apache.hadoop.net.NetUtils.wrapException(NetUtils.java:764)
> at org.apache.hadoop.ipc.Client.call(Client.java:1472)
> at org.apache.hadoop.ipc.Client.call(Client.java:1399)
> at org.apache.hadoop.ipc.ProtobufRpcEngine$Invoker.invoke(ProtobufRpcEngine.java:232)
> at com.sun.proxy.$Proxy27.allocate(Unknown Source)
> at org.apache.hadoop.yarn.api.impl.pb.client.ApplicationMasterProtocolPBClientImpl.allocate(ApplicationMasterProtocolPBClientImpl.java:77)
> at sun.reflect.GeneratedMethodAccessor3.invoke(Unknown Source)
> at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
> at java.lang.reflect.Method.invoke(Method.java:606)
> at org.apache.hadoop.io.retry.RetryInvocationHandler.invokeMethod(RetryInvocationHandler.java:187)
> at org.apache.hadoop.io.retry.RetryInvocationHandler.invoke(RetryInvocationHandler.java:102)
> at com.sun.proxy.$Proxy28.allocate(Unknown Source)
> at org.apache.hadoop.yarn.client.api.impl.AMRMClientImpl.allocate(AMRMClientImpl.java:278)
> at org.apache.hadoop.yarn.client.api.async.impl.AMRMClientAsyncImpl$HeartbeatThread.run(AMRMClientAsyncImpl.java:224)
> Caused by: java.io.EOFException
> at java.io.DataInputStream.readInt(DataInputStream.java:392)
> at org.apache.hadoop.ipc.Client$Connection.receiveRpcResponse(Client.java:1071)
> at org.apache.hadoop.ipc.Client$Connection.run(Client.java:966)
> 2014-10-08 02:24:12,256 INFO [AMRM Callback Handler Thread] org.apache.hadoop.yarn.client.api.async.impl.AMRMClientAsyncImpl: Interrupted while waiting for queue
> java.lang.InterruptedException
> at java.util.concurrent.locks.AbstractQueuedSynchronizer$ConditionObject.reportInterruptAfterWait(AbstractQueuedSynchronizer.java:2017)
> at java.util.concurrent.locks.AbstractQueuedSynchronizer$ConditionObject.await(AbstractQueuedSynchronizer.java:2052)
> at java.util.concurrent.LinkedBlockingQueue.take(LinkedBlockingQueue.java:442)
> at org.apache.hadoop.yarn.client.api.async.impl.AMRMClientAsyncImpl$CallbackHandlerThread.run(AMRMClientAsyncImpl.java:274)
> 2014-10-08 02:24:12,257 ERROR [AMRM Callback Handler Thread] org.apache.hadoop.yarn.client.api.async.impl.AMRMClientAsyncImpl: Stopping callback due to:
> java.io.IOException: Failed on local exception: java.io.EOFException; Host Details : local host is: "m-tez-uns-try-3/1.1.1.1"; destination host is: "m-tez-uns-try-3":8030;
> at org.apache.hadoop.net.NetUtils.wrapException(NetUtils.java:764)
> at org.apache.hadoop.ipc.Client.call(Client.java:1472)
> at org.apache.hadoop.ipc.Client.call(Client.java:1399)
> at org.apache.hadoop.ipc.ProtobufRpcEngine$Invoker.invoke(ProtobufRpcEngine.java:232)
> at com.sun.proxy.$Proxy27.allocate(Unknown Source)
> at org.apache.hadoop.yarn.api.impl.pb.client.ApplicationMasterProtocolPBClientImpl.allocate(ApplicationMasterProtocolPBClientImpl.java:77)
> at sun.reflect.GeneratedMethodAccessor3.invoke(Unknown Source)
> at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
> at java.lang.reflect.Method.invoke(Method.java:606)
> at org.apache.hadoop.io.retry.RetryInvocationHandler.invokeMethod(RetryInvocationHandler.java:187)
> at org.apache.hadoop.io.retry.RetryInvocationHandler.invoke(RetryInvocationHandler.java:102)
> at com.sun.proxy.$Proxy28.allocate(Unknown Source)
> at org.apache.hadoop.yarn.client.api.impl.AMRMClientImpl.allocate(AMRMClientImpl.java:278)
> at org.apache.hadoop.yarn.client.api.async.impl.AMRMClientAsyncImpl$HeartbeatThread.run(AMRMClientAsyncImpl.java:224)
> Caused by: java.io.EOFException
> at java.io.DataInputStream.readInt(DataInputStream.java:392)
> at org.apache.hadoop.ipc.Client$Connection.receiveRpcResponse(Client.java:1071)
> at org.apache.hadoop.ipc.Client$Connection.run(Client.java:966)
> 2014-10-08 02:24:12,257 INFO [TaskSchedulerAppCaller #0] org.apache.tez.dag.app.rm.TaskSchedulerEventHandler: Error reported by scheduler
> 2014-10-08 02:24:12,258 INFO [AsyncDispatcher event handler] org.apache.tez.common.TezUtilsInternal: Redirecting log file based on addend: dag_1412734988643_0001_1_post
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)