[
https://issues.apache.org/jira/browse/YARN-10467?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17222235#comment-17222235
]
Jim Brennan commented on YARN-10467:
------------------------------------
Thanks for reporting this and for the solution [~haibochen]! Everything
looks good to me.
I hesitate to mention one minor nit, a typo in this comment:
{quote}// there might be some completed containers that *are have* not been
pulled
{quote}
It's up to you whether you want to fix this.
[~jhung] were you planning to commit this?
> ContainerIdPBImpl objects can be leaked in RMNodeImpl.completedContainers
> -------------------------------------------------------------------------
>
> Key: YARN-10467
> URL: https://issues.apache.org/jira/browse/YARN-10467
> Project: Hadoop YARN
> Issue Type: Bug
> Components: resourcemanager
> Affects Versions: 2.10.0, 3.0.3, 3.2.1, 3.1.4
> Reporter: Haibo Chen
> Assignee: Haibo Chen
> Priority: Major
> Attachments: YARN-10467.00.patch, YARN-10467.01.patch,
> YARN-10467.branch-2.10.00.patch, YARN-10467.branch-2.10.01.patch,
> YARN-10467.branch-2.10.02.patch
>
>
> In one of our recent heap analysis, we found that the majority of the heap is
> occupied by {{RMNodeImpl.completedContainers}}<ContainerIdPBImp>, which
> accounts for 19GB, out of 24.3 GB. There are over 86 million
> ContainerIdPBImpl objects, in contrast, only 161,601 RMContainerImpl objects
> which represent the # of active containers that RM is still tracking.
> Inspecting some ContainerIdPBImpl objects, they belong to applications that
> have long finished. This indicates some sort of memory leak of
> ContainerIdPBImpl objects in RMNodeImpl.
>
> Right now, when a container is reported by a NM as completed, it is
> immediately added to RMNodeImpl.completedContainers and later cleaned up
> after the AM has been notified of its completion in the AM-RM heartbeat. The
> cleanup can be broken into a few steps.
> * Step 1: the completed container is first added to
> RMAppAttemptImpl.justFinishedContainers (this is asynchronous to being added
> to {{RMNodeImpl.completedContainers}}).
> * Step 2: During the heartbeat AM-RM heartbeat, the container is removed
> from RMAppAttemptImpl.justFinishedContainers and added to
> RMAppAttemptImpl.finishedContainersSentToAM
> Once a completed container gets added to
> RMAppAttemptImpl.finishedContainersSentToAM, it is guaranteed to be cleaned
> up from {{RMNodeImpl.completedContainers}}
>
> However, if the AM exits (regardless of failure or success) before some
> recently completed containers can be added to
> RMAppAttemptImpl.finishedContainersSentToAM in previous heartbeats, there
> won’t be any future AM-RM heartbeat to perform aforementioned step 2. Hence,
> these objects stay in RMNodeImpl.completedContainers forever.
> We have observed in MR that AMs can decide to exit upon success of all it
> tasks without waiting for notification of the completion of every container,
> or AM may just die suddenly (e.g. OOM). Spark and other framework may just
> be similar.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
