[
https://issues.apache.org/jira/browse/FLINK-24713?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17438053#comment-17438053
]
Till Rohrmann commented on FLINK-24713:
---------------------------------------
Introducing a configurable reconciliation interval before new containers/pods
are requested can solve the described problem. Note that if the TMs in recovery
cannot fulfill all pending requests, then we will delay the provision of the
required slots. I guess it depends on what one as the operator of Flink prefers
quick resource provisioning or not over provisioning too much for a short time.
> Postpone resourceManager serving after the recovery phase has finished
> ----------------------------------------------------------------------
>
> Key: FLINK-24713
> URL: https://issues.apache.org/jira/browse/FLINK-24713
> Project: Flink
> Issue Type: Improvement
> Components: Runtime / Coordination
> Affects Versions: 1.14.0
> Reporter: Aitozi
> Priority: Major
>
> When ResourceManager started, JobManger will connect to the ResourceManager,
> this means the ResourceManager will begin to try serve the resource requests
> from SlotManager.
> If ResourceManager failover, although it will try to recover the pod /
> container from previous attempt, But new resource requirements may happen
> before the old taskManger register to slotManager.
> In this case, it may double the required taskManager when jobManager
> failover. We may need a mechanism to postpone resourceManager serving after
> the recovery phase has finished
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
