[
https://issues.apache.org/jira/browse/FLINK-33798?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17795576#comment-17795576
]
Hangxiang Yu commented on FLINK-33798:
--------------------------------------
Thanks [~fanrui] pinging me here.
I think you are right. The behavious after relocating is not consistent with
before.
We could make it.
Thanks [~liming] for reporting this and we could go ahead.
> Automatically clean up rocksdb logs when the task failover.
> -----------------------------------------------------------
>
> Key: FLINK-33798
> URL: https://issues.apache.org/jira/browse/FLINK-33798
> Project: Flink
> Issue Type: Improvement
> Components: Runtime / State Backends
> Reporter: Ming Li
> Assignee: Ming Li
> Priority: Major
>
> Since FLINK-24785 relocates rocksdb log, multiple rocksdb logs will be
> created under the flink log directory, but they are not cleaned up during
> task failover, resulting in a large number of rocksdb logs under the flink
> log directory.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
