[
https://issues.apache.org/jira/browse/FLINK-22638?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17349154#comment-17349154
]
Dawid Wysakowicz edited comment on FLINK-22638 at 5/21/21, 11:07 AM:
---------------------------------------------------------------------
Implemented in:
* master
** f0415160da780a5c8d9670102a3ea03a7e423771
* 1.13.2
** e70fe71f48e453bb84d808761c01ca27abbd7c48
was (Author: dawidwys):
Fixed in:
* master
** f0415160da780a5c8d9670102a3ea03a7e423771
* 1.13.2
** e70fe71f48e453bb84d808761c01ca27abbd7c48
> Keep channels blocked on alignment timeout
> ------------------------------------------
>
> Key: FLINK-22638
> URL: https://issues.apache.org/jira/browse/FLINK-22638
> Project: Flink
> Issue Type: Improvement
> Components: Runtime / Checkpointing
> Reporter: Dawid Wysakowicz
> Assignee: Dawid Wysakowicz
> Priority: Major
> Labels: pull-request-available
> Fix For: 1.14.0, 1.13.2
>
>
> In order to improve checkpointing time we could keep channels blocked in case
> an alignment timeout occurs. That way we prioritize the channels that we have
> not received the barrier yet. This solution is based on the assumption that
> all upstream operators are working with aligned checkpoints and we do not
> mind delaying the subsequent checkpoints on the blocked channels.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
