[
https://issues.apache.org/jira/browse/AURORA-179?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13897330#comment-13897330
]
Suman Karumuri commented on AURORA-179:
---------------------------------------
By binary search narrowed down the bug to the following commit:
commit 28372648a364bcef17b9dffaa258081dbfc5b54f
Author: Bill Farner <[email protected]>
Date: Wed Jan 22 10:23:31 2014 -0800
Only make the SchedulerDriver available after registered() has been called.
I've also removed some functions from the Driver interface, since they
should
be used exclusively by SchedulerLifecycle.
Bugs closed: AURORA-45
Reviewed at https://reviews.apache.org/r/16995/
Looks like IsolatedSchedulerModule is not updated when the schedulerDriver was
updated. Looking into a fix now. Until then my previous comment should atleast
unblock the UI testing.
> Local scheduler run fails to register with mesos
> ------------------------------------------------
>
> Key: AURORA-179
> URL: https://issues.apache.org/jira/browse/AURORA-179
> Project: Aurora
> Issue Type: Bug
> Components: Scheduler
> Reporter: Maxim Khutornenko
> Assignee: Suman Karumuri
> Priority: Critical
>
> $ gradle clean run
> {noformat}
> ...
> I0204 00:21:43.997 THREAD1
> com.twitter.common.application.Lifecycle.awaitShutdown: Awaiting shutdown
> E0204 00:22:44.013 THREAD24
> org.apache.aurora.scheduler.SchedulerLifecycle$5$3.run: Framework has not
> been registered within the tolerated delay.
> I0204 00:22:44.013 THREAD24
> com.twitter.common.util.StateMachine$Builder$1.execute: SchedulerLifecycle
> state machine transition LEADER_AWAITING_REGISTRATION -> DEAD
> E0204 00:22:44.016 THREAD22
> org.apache.aurora.scheduler.SchedulerLifecycle$SchedulerCandidateImpl.onDefeated:
> Lost leadership, committing suicide.
> I0204 00:22:44.016 THREAD24
> com.twitter.common.util.StateMachine$Builder$1.execute: storage state machine
> transition READY -> STOPPED
> I0204 00:22:44.016 THREAD22
> com.twitter.common.util.StateMachine$Builder$1.execute: SchedulerLifecycle
> state machine transition DEAD -> DEAD
> I0204 00:22:44.017 THREAD24
> com.twitter.common.application.Lifecycle.shutdown: Shutting down application
> I0204 00:22:44.017 THREAD22
> org.apache.aurora.scheduler.SchedulerLifecycle$7.execute: Shutdown already
> invoked, ignoring extra call.
> I0204 00:22:44.017 THREAD24
> com.twitter.common.application.ShutdownRegistry$ShutdownRegistryImpl.execute:
> Executing 12 shutdown commands.
> W0204 00:22:44.018 THREAD22
> com.twitter.common.zookeeper.CandidateImpl$4.onGroupChange: All candidates
> have temporarily left the group: Group /twitter/service/mesos/local/scheduler
> ...
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
