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

Posted to issues@flink.apache.org by "Stephan Ewen (Jira)" <ji...@apache.org> on 2020/05/16 16:33:00 UTC

[jira] [Resolved] (FLINK-17670) Savepoint handling of "non-restored" state should also take OperatorCoordinator state into account

     [ https://issues.apache.org/jira/browse/FLINK-17670?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Stephan Ewen resolved FLINK-17670.
----------------------------------
    Resolution: Fixed

Fixed in 1.11.0 via

 (refactor) 90e4708c9a5ab24b5f50423a181d75be375e2d1e
 (feature) 382cf09bbfb4e383e373af8ebf1a9dee13b8a632

> Savepoint handling of "non-restored" state should also take OperatorCoordinator state into account
> --------------------------------------------------------------------------------------------------
>
>                 Key: FLINK-17670
>                 URL: https://issues.apache.org/jira/browse/FLINK-17670
>             Project: Flink
>          Issue Type: Bug
>          Components: Runtime / Checkpointing
>            Reporter: Stephan Ewen
>            Assignee: Stephan Ewen
>            Priority: Major
>             Fix For: 1.11.0
>
>
> The savepoint restore should fail when there is unmatched state from an OperatorCoordinator only, and non-restored state is not allowed.
> The relevant validation logic is in the {{org.apache.flink.runtime.checkpoint.Checkpoints}} class.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)