[
https://issues.apache.org/jira/browse/FLINK-17447?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17099843#comment-17099843
]
Piotr Nowojski commented on FLINK-17447:
----------------------------------------
I would doubt that this is a network stack issue. This error shouldn't happen,
unless there is some serious bug, which I would expect to pop up in our ITCases
sooner or later.
One thing that maybe could cause something like this, if someone/something is
collecting/producing records from different threads, not using mailbox or not
acquiring checkpoint lock. [~chun111111], have you implemented some custom
source function, or a custom operator? Could you share your pipeline? A minimal
example that could reproduce the issue would be also helpful.
> Flink CEPOperator StateException
> --------------------------------
>
> Key: FLINK-17447
> URL: https://issues.apache.org/jira/browse/FLINK-17447
> Project: Flink
> Issue Type: Bug
> Components: Runtime / Network
> Affects Versions: 1.10.0
> Reporter: chun111111
> Priority: Critical
> Attachments: image-2020-04-29-10-21-10-717.png,
> image-2020-04-29-10-22-20-645.png, image-2020-04-29-10-23-14-306.png,
> image-2020-04-29-11-55-12-995.png, image-2020-04-30-11-37-20-411.png
>
>
> !image-2020-04-29-11-55-12-995.png!!image-2020-04-29-10-21-10-717.png!
>
> !image-2020-04-29-10-23-14-306.png!
>
> !image-2020-04-29-10-22-20-645.png!
>
>
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
