[
https://issues.apache.org/jira/browse/FLINK-21084?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Dawid Wysakowicz closed FLINK-21084.
------------------------------------
Resolution: Implemented
Implemented in
d9cc268be61c1789e5cb32d0897d962721be9c01..c46b20343d0e7dcd7ce2ac5fc6e71d02b07df1be
> Allows tasks to report finish state
> -----------------------------------
>
> Key: FLINK-21084
> URL: https://issues.apache.org/jira/browse/FLINK-21084
> Project: Flink
> Issue Type: Sub-task
> Components: API / DataStream, Runtime / Checkpointing
> Reporter: Yun Gao
> Assignee: Yun Gao
> Priority: Major
> Labels: pull-request-available
> Fix For: 1.14.0
>
>
> After we have modified the operator interface to change close to finish and
> dispose to close, it would become reasonable for operators to continue taking
> checkpoint after finished() and before closed(). In fact, we would not have
> checkpoints after operators get closed, thus it would be reasonable to view
> operator subtasks as FINISHED once the tasks are finished.
> However, there is a special case, if an operator is finished in the restored
> checkpoints, although in this case we would skip the lifecycle of the
> operators, the task still need to report a snapshot. Since it is reasonable
> to keep the operators still be finished in all the following checkpoints,
> thus the task should tell CheckpointCoordinator about that. Thus we would
> report a TaskStateSnapshot\{isFinished = true} in this case.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
