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

Posted to issues@flink.apache.org by "Chesnay Schepler (Jira)" <ji...@apache.org> on 2023/09/14 15:33:00 UTC

[jira] [Closed] (FLINK-33086) Protect failure enrichment against unhandled exceptions

     [ https://issues.apache.org/jira/browse/FLINK-33086?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Chesnay Schepler closed FLINK-33086.
------------------------------------
    Fix Version/s: 1.18.0
       Resolution: Fixed

master: e58a718d04e3ec6e2a43da8a868e5515916c0eea
1.18: 206609e822a8029a78245e6eef7ab5d88a0f370b

> Protect failure enrichment against unhandled exceptions
> -------------------------------------------------------
>
>                 Key: FLINK-33086
>                 URL: https://issues.apache.org/jira/browse/FLINK-33086
>             Project: Flink
>          Issue Type: Sub-task
>          Components: Runtime / Coordination
>            Reporter: Panagiotis Garefalakis
>            Assignee: Panagiotis Garefalakis
>            Priority: Major
>              Labels: pull-request-available
>             Fix For: 1.18.0
>
>
> Existing [labelFailure|https://github.com/apache/flink/blob/603181da811edb47c0d573492639a381fbbedc28/flink-runtime/src/main/java/org/apache/flink/runtime/failure/FailureEnricherUtils.java#L175] async logic is expecting FailureEnricher future to never fail (or do their own exception handling) however there is no way to enforce that as they are loaded as they are implemented as pluggable components. This could result to throwing away labels from other enrichers that successfully completed. 
> A better solution would be to handle the failures and LOG the errors.



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