[
https://issues.apache.org/jira/browse/AURORA-1593?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15113311#comment-15113311
]
Zameer Manji commented on AURORA-1593:
--------------------------------------
The problem here is that the {{TaskStateChange}} event is created when the
storage is started (when the scheduler becomes the leader), but the
{{TaskHistoryPruner}} is started up after the scheduler is registered with
Mesos.
It might be possible to avoid this error by delaying pubsub registration until
the scheduler is registred with Mesos. Kevin left a comment suggesting this:
{noformat}
// TODO(ksweeney): Would this be better as a scheduler active service?
SchedulerServicesModule.addAppStartupServiceBinding(binder()).to(RegisterSubscribers.class);
{noformat}
> PubSubEventModule fails to dispatch events to TaskHistoryPruner on startup
> --------------------------------------------------------------------------
>
> Key: AURORA-1593
> URL: https://issues.apache.org/jira/browse/AURORA-1593
> Project: Aurora
> Issue Type: Bug
> Reporter: Zameer Manji
> Assignee: John Sirois
>
> On latest master I see several exceptions that look like:
> {noformat}
> E0122 22:59:19.272 [AsyncProcessor-7, PubsubEventModule:84] Failed to
> dispatch event to public void
> org.apache.aurora.scheduler.pruning.TaskHistoryPruner.recordStateChange(org.apache.aurora.scheduler.events.PubsubEvent$TaskStateChange):
> java.lang.IllegalStateException j
> ava.lang.IllegalStateException: null
> at
> com.google.common.base.Preconditions.checkState(Preconditions.java:159)
> ~[guava-19.0.jar:na]
> at
> org.apache.aurora.scheduler.pruning.TaskHistoryPruner.recordStateChange(TaskHistoryPruner.java:117)
> ~[aurora-116.jar:na]
> at sun.reflect.GeneratedMethodAccessor116.invoke(Unknown Source)
> ~[na:na]
> at
> sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
> ~[na:1.8.0_66-Tw8r9b2]
> at java.lang.reflect.Method.invoke(Method.java:497)
> ~[na:1.8.0_66-Tw8r9b2]
> at
> com.google.common.eventbus.Subscriber.invokeSubscriberMethod(Subscriber.java:95)
> ~[guava-19.0.jar:na]
> at
> com.google.common.eventbus.Subscriber$SynchronizedSubscriber.invokeSubscriberMethod(Subscriber.java:154)
> ~[guava-19.0.jar:na]
> at com.google.common.eventbus.Subscriber$1.run(Subscriber.java:80)
> ~[guava-19.0.jar:na]
> at
> java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
> ~[na:1.8.0_66-Tw8r9b2]
> at java.util.concurrent.FutureTask.run(FutureTask.java:266)
> ~[na:1.8.0_66-Tw8r9b2]
> at
> java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$201(ScheduledThreadPoolExecutor.java:180)
> ~[na:1.8.0_66-Tw8r9b2]
> at
> java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:293)
> ~[na:1.8.0_66-Tw8r9b2]
> at
> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
> ~[na:1.8.0_66-Tw8r9b2]
> at
> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
> ~[na:1.8.0_66-Tw8r9b2]
> at java.lang.Thread.run(Thread.java:745) ~[na:1.8.0_66-Tw8r9b2]
> {noformat}
> The problem is that {{TaskHistoryPruner}} assumes it is started before the
> event bus starts sending events to the service. This appears to not be the
> case.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
