[
https://issues.apache.org/jira/browse/FLINK-13245?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16885371#comment-16885371
]
zhijiang edited comment on FLINK-13245 at 7/15/19 4:22 PM:
-----------------------------------------------------------
Thanks for finding this potential issue and the investigation! [~Zentol]
[~azagrebin]
I think the idea of above modifications makes sense, because the
`availableReader` is not always equivalent to `allReaders`, then it is proper
to find the canceled view reader from `allReaders` instead.
This issue also exists in previous {{SpillableSubpartition}} which actually
uses memory type in {{SlotCountExceedingParallelismTest,}} so we could not find
this potential bug then.
In detail, we should also call `toRelease.notifySubpartitionConsumed` before
calling `toRelease.releaseAllResources` in above modifications. Otherwise the
reference counter in {{ReleaseOnConsumptionResultPartition}} would not decrease
to zero and really release partition via {{ResultPartitionManager}}.
I would submit the PR and add some unite tests later tomorrow.
was (Author: zjwang):
Thanks for finding this potential issue and the investigation! [~Zentol]
[~azagrebin]
I think the above idea of above modifications makes sense, because the
`availableReader` is not always equivalent to `allReaders`, then it is proper
to find the canceled view reader from `allReaders` instead.
This issue also exists in previous {{SpillableSubpartition}} which actually
uses memory type in {{SlotCountExceedingParallelismTest,}} so we could not find
this potential bug then.
In detail, we should also call `toRelease.notifySubpartitionConsumed` before
calling `toRelease.releaseAllResources` in above modifications. Otherwise the
reference counter in {{ReleaseOnConsumptionResultPartition}} would not decrease
to zero and really release partition via {{ResultPartitionManager}}.
I would submit the PR and add some unite tests later tomorrow.
> Network stack is leaking files
> ------------------------------
>
> Key: FLINK-13245
> URL: https://issues.apache.org/jira/browse/FLINK-13245
> Project: Flink
> Issue Type: Bug
> Components: Runtime / Network
> Affects Versions: 1.9.0
> Reporter: Chesnay Schepler
> Assignee: zhijiang
> Priority: Blocker
> Fix For: 1.9.0
>
>
> There's file leak in the network stack / shuffle service.
> When running the {{SlotCountExceedingParallelismTest}} on Windows a large
> number of {{.channel}} files continue to reside in a
> {{flink-netty-shuffle-XXX}} directory.
> From what I've gathered so far these files are still being used by a
> {{BoundedBlockingSubpartition}}. The cleanup logic in this class uses
> ref-counting to ensure we don't release data while a reader is still present.
> However, at the end of the job this count has not reached 0, and thus nothing
> is being released.
> The same issue is also present on the {{ResultPartition}} level; the
> {{ReleaseOnConsumptionResultPartition}} also are being released while the
> ref-count is greater than 0.
> Overall it appears like there's some issue with the notifications for
> partitions being consumed.
> It is feasible that this issue has recently caused issues on Travis where the
> build were failing due to a lack of disk space.
--
This message was sent by Atlassian JIRA
(v7.6.14#76016)
