[
https://issues.apache.org/jira/browse/FLINK-22886?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17367597#comment-17367597
]
Roman Khachatryan commented on FLINK-22886:
-------------------------------------------
Reviewed and merged into 1.12 as 20da2a10dd91d7dbc935a11dcf4964de8d7bef2b
and into 1.13 as 4519089d4a45f223ca3ffd2a88864d02a059f8cd.
Thanks a lot for providing the fix and the test!
> Thread leak in RocksDBStateUploader
> -----------------------------------
>
> Key: FLINK-22886
> URL: https://issues.apache.org/jira/browse/FLINK-22886
> Project: Flink
> Issue Type: Bug
> Components: Runtime / State Backends
> Affects Versions: 1.11.3, 1.13.1, 1.12.4
> Reporter: Jiayi Liao
> Assignee: Yue Ma
> Priority: Major
> Labels: critical, pull-request-available
> Attachments: image-2021-06-06-13-46-34-604.png
>
>
> {{ExecutorService}} in {{RocksDBStateUploader}} is not shut down, which may
> leak thread when tasks fail.
> BTW, we should name the thread group in {{ExecutorService}}, otherwise what
> we see in the stack, is a lot of threads named with pool-m-thread-n like this:
>
> !image-2021-06-06-13-46-34-604.png!
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
