[
https://issues.apache.org/jira/browse/YARN-10459?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17485386#comment-17485386
]
Íñigo Goiri commented on YARN-10459:
------------------------------------
Thanks [~jianliang.wu] for bringing this up and [~minni31] for driving the PR.
PR 3195 merged to trunk.
> containerLaunchedOnNode method not need to hold schedulerApptemt lock
> ----------------------------------------------------------------------
>
> Key: YARN-10459
> URL: https://issues.apache.org/jira/browse/YARN-10459
> Project: Hadoop YARN
> Issue Type: Improvement
> Affects Versions: 3.2.0, 3.1.3
> Reporter: Ryan Wu
> Assignee: Minni Mittal
> Priority: Major
> Labels: pull-request-available
> Attachments: YARN-10459.v1.patch
>
> Time Spent: 1h 20m
> Remaining Estimate: 0h
>
>
> Now, the containerLaunchedOnNode method hold the SchedulerApplicationAttempt
> writelock, but looking at the method, it does not change any field. And more
> seriously, this will affect the scheduler.
> {code:java}
> public void containerLaunchedOnNode(ContainerId containerId, NodeId nodeId)
> {
> // Inform the container
> writelock.lock
> try {
> RMContainer rmContainer = getRMContainer(containerId);
> if (rmContainer == null) {
> // Some unknown container sneaked into the system. Kill it.
> rmContext.getDispatcher().getEventHandler().handle( new
> RMNodeCleanContainerEvent(nodeId, containerId)); return;
> }
> rmContainer.handle( new RMContainerEvent(containerId,
> RMContainerEventType.LAUNCHED));
> }finally {
> writeLock.unlock();
> }
> }
> {code}
>
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
