[
https://issues.apache.org/jira/browse/FLINK-22401?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17329213#comment-17329213
]
Till Rohrmann commented on FLINK-22401:
---------------------------------------
[~kevin.cyj] you are right that {{taskmanager.slot.timeout}} has been
introduced via FLINK-21428 in 1.13. Before the {{akka.ask.timeout}} was used as
the timeout (default value 10s). Hence, if you did not change the
{{akka.ask.timeout}}, then you should have the same behaviour as before.
I also couldn't find anything suspicious in the logs. Could it be that the
machine had a pause of 10 seconds? If the underlying hardware is massively
oversubscribed this could be a cause for it.
> Python tests fail with "Slot ... is not allocated by job ..."
> -------------------------------------------------------------
>
> Key: FLINK-22401
> URL: https://issues.apache.org/jira/browse/FLINK-22401
> Project: Flink
> Issue Type: Bug
> Components: API / Python, Runtime / Coordination
> Affects Versions: 1.14.0
> Reporter: Dawid Wysakowicz
> Priority: Major
> Labels: test-stability
>
> https://dev.azure.com/apache-flink/apache-flink/_build/results?buildId=16965&view=logs&j=c88eea3b-64a0-564d-0031-9fdcd7b8abee&t=ff888d9b-cd34-53cc-d90f-3e446d355529&l=22519
> {code}
> Apr 21 15:22:58 Caused by: org.apache.flink.util.FlinkException: Slot 0 on
> TaskExecutor 10.1.0.4:41973-2f5e95 is not allocated by job
> 7eb1e50ddf9e3cfecbcf6b0eb15df0cb.
> Apr 21 15:22:58 at
> org.apache.flink.runtime.taskexecutor.TaskExecutor.failNoLongerAllocatedSlots(TaskExecutor.java:1993)
> Apr 21 15:22:58 at
> org.apache.flink.runtime.taskexecutor.TaskExecutor.syncSlotsWithSnapshotFromJobMaster(TaskExecutor.java:1977)
> Apr 21 15:22:58 at
> org.apache.flink.runtime.taskexecutor.TaskExecutor.access$3500(TaskExecutor.java:181)
> Apr 21 15:22:58 at
> org.apache.flink.runtime.taskexecutor.TaskExecutor$JobManagerHeartbeatListener.lambda$reportPayload$1(TaskExecutor.java:2343)
> Apr 21 15:22:58 at java.util.Optional.ifPresent(Optional.java:159)
> Apr 21 15:22:58 at
> org.apache.flink.util.OptionalConsumer.ifPresent(OptionalConsumer.java:45)
> Apr 21 15:22:58 at
> org.apache.flink.runtime.taskexecutor.TaskExecutor$JobManagerHeartbeatListener.reportPayload(TaskExecutor.java:2341)
> Apr 21 15:22:58 at
> org.apache.flink.runtime.taskexecutor.TaskExecutor$JobManagerHeartbeatListener.reportPayload(TaskExecutor.java:2317)
> Apr 21 15:22:58 at
> org.apache.flink.runtime.heartbeat.HeartbeatManagerImpl.requestHeartbeat(HeartbeatManagerImpl.java:214)
> Apr 21 15:22:58 at
> org.apache.flink.runtime.taskexecutor.TaskExecutor.heartbeatFromJobManager(TaskExecutor.java:903)
> Apr 21 15:22:58 at sun.reflect.NativeMethodAccessorImpl.invoke0(Native
> Method)
> Apr 21 15:22:58 at
> sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
> Apr 21 15:22:58 at
> sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
> Apr 21 15:22:58 at java.lang.reflect.Method.invoke(Method.java:498)
> Apr 21 15:22:58 at
> org.apache.flink.runtime.rpc.akka.AkkaRpcActor.handleRpcInvocation(AkkaRpcActor.java:301)
> Apr 21 15:22:58 at
> org.apache.flink.runtime.rpc.akka.AkkaRpcActor.handleRpcMessage(AkkaRpcActor.java:212)
> Apr 21 15:22:58 ... 19 more
> {code}
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
