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

Posted to issues@flink.apache.org by "Roman Khachatryan (Jira)" <ji...@apache.org> on 2022/04/28 18:27:00 UTC

[jira] [Resolved] (FLINK-25511) Pre-emptively uploaded changelog not discarded up if materialized before checkpoint

     [ https://issues.apache.org/jira/browse/FLINK-25511?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Roman Khachatryan resolved FLINK-25511.
---------------------------------------
    Resolution: Fixed

Merged into master as 5b58ef66a8835bf22db23e1d9836a1e9f4d94045..c5430e2e5d4eeb0aba14ce3ea8401747afe0182d

> Pre-emptively uploaded changelog not discarded up if materialized before checkpoint
> -----------------------------------------------------------------------------------
>
>                 Key: FLINK-25511
>                 URL: https://issues.apache.org/jira/browse/FLINK-25511
>             Project: Flink
>          Issue Type: Sub-task
>          Components: Runtime / Checkpointing, Runtime / State Backends
>    Affects Versions: 1.15.0
>            Reporter: Yuan Mei
>            Assignee: Roman Khachatryan
>            Priority: Major
>              Labels: pull-request-available
>             Fix For: 1.16.0
>
>
> This is related to Garbage Collection. 
> The problem is more severe than FLINK-25512 because there is leftover after materialization truncation each time (if pre-emptive uploads are enabled).
> However, the probability can still be not high because of grouping of backends into a single file (so only if ALL the backends on TM materialize semi-simultaneously then it happen).



--
This message was sent by Atlassian Jira
(v8.20.7#820007)