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

Posted to issues@flink.apache.org by "Matyas Orhidi (Jira)" <ji...@apache.org> on 2022/07/01 13:33:00 UTC

[jira] [Commented] (FLINK-28348) Add configurable flag to disable last-state fallback for savepoint upgrade

    [ https://issues.apache.org/jira/browse/FLINK-28348?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17561456#comment-17561456 ] 

Matyas Orhidi commented on FLINK-28348:
---------------------------------------

fyi [~gyfora] I've started working on this.

> Add configurable flag to disable last-state fallback for savepoint upgrade
> --------------------------------------------------------------------------
>
>                 Key: FLINK-28348
>                 URL: https://issues.apache.org/jira/browse/FLINK-28348
>             Project: Flink
>          Issue Type: Improvement
>          Components: Kubernetes Operator
>            Reporter: Gyula Fora
>            Priority: Critical
>             Fix For: kubernetes-operator-1.1.0
>
>
> Currently if the job is not running and savepoint upgrade mode is configured, the ApplicationReconciler can fall back to last-state upgrade mode if HA was enabled.
> While in most cases this is fine, in some situation the user might only want to allow savepoint ugrades. We should add a flag:
> kubernetes.operator.job.upgrade.last-state-fallback.enabled : true
> Default should be true to match the current behaviour.



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