[
https://issues.apache.org/jira/browse/YARN-3091?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16244191#comment-16244191
]
Jason Lowe commented on YARN-3091:
----------------------------------
bq. I'm looking at reverting the read/write lock changes within the fair
scheduler at least. Thoughts?
+1, we've also seen a number of problems around the scheduler's read/write
locks and have done some short-term fixes to work around them like YARN-6680.
They are significantly more expensive to acquire than a standard mutex if
nobody is holding the lock, and there are lots of places where the scheduler
needs to acquire them during a scheduling pass.
> [Umbrella] Improve and fix locks of RM scheduler
> ------------------------------------------------
>
> Key: YARN-3091
> URL: https://issues.apache.org/jira/browse/YARN-3091
> Project: Hadoop YARN
> Issue Type: Task
> Components: capacityscheduler, fairscheduler, resourcemanager,
> scheduler
> Reporter: Wangda Tan
>
> In existing YARN RM scheduler, there're some issues of using locks. For
> example:
> - Many unnecessary synchronized locks, we have seen several cases recently
> that too frequent access of scheduler makes scheduler hang. Which could be
> addressed by using read/write lock. Components include scheduler, CS queues,
> apps
> - Some fields not properly locked (Like clusterResource)
> We can address them together in this ticket.
> (More details see comments below)
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
---------------------------------------------------------------------
To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
