[
https://issues.apache.org/jira/browse/FLINK-9480?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16495140#comment-16495140
]
Till Rohrmann commented on FLINK-9480:
--------------------------------------
I can totally see the benefits of speeding up rescaling operation by first
trying to read local state and then falling back to remote state. In the first
iteration, it could be a best effort approach as suggested by [~sihuazhou].
Next we could try to make the scheduling a bit smarter and eventually it could
mean that we first load the required state to a TM before deploying tasks.
I also agree with you two about the priorities wrt rescalable timers and state
ttl.
> Let local recovery support rescaling
> ------------------------------------
>
> Key: FLINK-9480
> URL: https://issues.apache.org/jira/browse/FLINK-9480
> Project: Flink
> Issue Type: Improvement
> Components: State Backends, Checkpointing
> Affects Versions: 1.5.0
> Reporter: Sihua Zhou
> Priority: Major
>
> Currently, local recovery only support restore from checkpoint and without
> rescaling. Maybe we should enable it to support rescaling.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
