[
https://issues.apache.org/jira/browse/FLINK-24120?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17408902#comment-17408902
]
Till Rohrmann edited comment on FLINK-24120 at 9/2/21, 2:46 PM:
----------------------------------------------------------------
[~wind_ljy] could you add this to the documentation of Flink (when to use this,
what it does)? Because our user seem to run into this problem.
Maybe we can add it here:
https://ci.apache.org/projects/flink/flink-docs-release-1.13/docs/deployment/resource-providers/standalone/docker/#switch-memory-allocator
or here
https://ci.apache.org/projects/flink/flink-docs-release-1.13/docs/deployment/memory/mem_trouble/.
was (Author: till.rohrmann):
[~wind_ljy] could you add this to the documentation of Flink (when to use this,
what it does)? Maybe we can add it here:
https://ci.apache.org/projects/flink/flink-docs-release-1.13/docs/deployment/resource-providers/standalone/docker/#switch-memory-allocator
or here
https://ci.apache.org/projects/flink/flink-docs-release-1.13/docs/deployment/memory/mem_trouble/.
> The memory will rise a little after each savepoint ends
> -------------------------------------------------------
>
> Key: FLINK-24120
> URL: https://issues.apache.org/jira/browse/FLINK-24120
> Project: Flink
> Issue Type: Bug
> Components: flink-docker
> Affects Versions: 1.13.2
> Environment: flink on k8s
> flink version : 1.13.2
> state.backend: rocksdb
> state.checkpoints.dir: hdfs_url
> state.savepoints.dir: hdfs_url
> state.backend.rocksdb.memory.managed: true
>
> Reporter: no nO
> Priority: Blocker
> Attachments: image-2021-09-02-10-40-36-043.png, 截屏2021-09-02
> 上午10.44.10.png
>
>
> !image-2021-09-02-10-40-36-043.png!
> My task will do a savepoint every hour, so every hour will do a savepoint.
> From the memory monitoring, it can be seen that the memory of each hour will
> soar up, although the memory will drop a little later, but from every hour
> From the point of view of the memory peak on the whole point, the memory
> continues to rise little by little, and eventually rises to the limited
> memory, which will lead to being killed by k8S
>
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
