[
https://issues.apache.org/jira/browse/FLINK-29369?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Jingsong Lee updated FLINK-29369:
---------------------------------
Fix Version/s: table-store-0.2.1
> Commit delete file failure due to Checkpoint aborted
> ----------------------------------------------------
>
> Key: FLINK-29369
> URL: https://issues.apache.org/jira/browse/FLINK-29369
> Project: Flink
> Issue Type: Bug
> Components: Table Store
> Reporter: Jingsong Lee
> Assignee: Jingsong Lee
> Priority: Blocker
> Labels: pull-request-available
> Fix For: table-store-0.3.0, table-store-0.2.1
>
>
> After checkpoint abort, the files in cp5 may fall into cp6, because the
> compaction commit is deleted first and then added, which may lead to:
> -Delete a file
> -Add the same file again
> This causes the deleted file not to be found.
> We need to properly process the merge of the compression files.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
