You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@flink.apache.org by "Zakelly Lan (Jira)" <ji...@apache.org> on 2025/04/01 07:17:00 UTC

[jira] [Assigned] (FLINK-37597) Use ResourceGuard to make sure snapshot quit before ForSt dispose

     [ https://issues.apache.org/jira/browse/FLINK-37597?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Zakelly Lan reassigned FLINK-37597:
-----------------------------------

    Assignee: Yanfei Lei  (was: Zakelly Lan)

> Use ResourceGuard to make sure snapshot quit before ForSt dispose
> -----------------------------------------------------------------
>
>                 Key: FLINK-37597
>                 URL: https://issues.apache.org/jira/browse/FLINK-37597
>             Project: Flink
>          Issue Type: Bug
>          Components: Runtime / State Backends
>    Affects Versions: 2.0.0
>            Reporter: Zakelly Lan
>            Assignee: Yanfei Lei
>            Priority: Major
>             Fix For: 2.1.0
>
>
> Currently, there might be a chance that the forst state backend dispose when on-going snapshots are still in progress. This would be problematic since the db will coredump when enabling file deletion while the db core is already closed. This happens when the snapshot finishes after forst state backend disposes. We should apply ResourceGuard to protect the disposal order of db and snapshot, like RocksDB does.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)