[
https://issues.apache.org/jira/browse/FLINK-27433?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17529174#comment-17529174
]
Yun Tang commented on FLINK-27433:
----------------------------------
This is caused by FLINK-24785 and FLINK-26998. We moved the logs of RocksDB
from its inner folder to the flink log folder by default for the easy usage.
The several thousand rocksDB logs means that you have thousand of keyed
operator. If you just want to avoid such logs, maybe you could add option of
{{state.backend.rocksdb.log.dir}} to {{/tmp}} to E2E tests.
BTW, could you share one instance that have flooding logs？
> E2E log artifacts contain several thousand rocksdb logs
> -------------------------------------------------------
>
> Key: FLINK-27433
> URL: https://issues.apache.org/jira/browse/FLINK-27433
> Project: Flink
> Issue Type: Technical Debt
> Components: Build System / Azure Pipelines, Tests
> Affects Versions: 1.16.0
> Reporter: Chesnay Schepler
> Priority: Major
>
> It would be great if we could minimize them somehow because as is they kinda
> break the azure UI and are just a pain (beyond taking up a lot of space).
> Even zipping them before the upload would be fine in my book.
> I also just don't get why so few e2e tests manage to create 5k files.
> [~yunta] Is this caused by FLINK-23791?
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
