[
https://issues.apache.org/jira/browse/FLINK-27793?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17542479#comment-17542479
]
JasonLee commented on FLINK-27793:
----------------------------------
hi maybe You can upgrade hadoop to above 2.8.5 and try it out Because the
lowest hadoop version supported by Flink version 1.15.0 is 2.8.5
[https://nightlies.apache.org/flink/flink-docs-release-1.15/release-notes/flink-1.15/#upgrade-the-minimal-supported-hadoop-version-to-285]
My Hadoop version is 2.9.0 and I did not experience this problem when I
upgraded flink to 1.15.0.
> After upgrading to flink-1.15.0, flink jobmanager cannot shutdown
> -----------------------------------------------------------------
>
> Key: FLINK-27793
> URL: https://issues.apache.org/jira/browse/FLINK-27793
> Project: Flink
> Issue Type: Bug
> Environment: Flink 1.15.0
> Hadoop 2.7.3.2.6.5.0-292
> Start the task with yarn per job
> Reporter: jinshuangxian
> Priority: Major
> Fix For: 1.15.0
>
> Attachments: image-2022-05-26-20-43-49-784.png
>
>
> Under normal circumstances, after the taskmanager fails, it will retry
> multiple times. After reaching the number of failures, the task will exit.
> After upgrading to flink-1.15.0, the test found that when the taskmanager was
> abnormal, the job was completed, but the task jobmanager was still running
> normally, and the yarn application -list commmand found that the task was
> still running.
> Another problem, when the command "flink stop" is executed, the task stop
> success, i found that the task is still running normally on yarn.
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
