[
https://issues.apache.org/jira/browse/FLINK-22133?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17324985#comment-17324985
]
Stephan Ewen commented on FLINK-22133:
--------------------------------------
Originally, we were thinking to keep checkpoint IDs out of the
{{SplitEnumerator}} completely, keeping it simple.
We somehow partially went away from this with adding the {{CheckpointListener}}
interface, which gets checkpointIDs for completed checkpoints. In that, I guess
it makes sense to add them also to the {{SplitEnumerator.snapshotState()}}
method.
[~jqin] what do you think?
[~dwysakowicz] [~maguowei] This is a very simple change, but API breaking, so
the sooner we do it the better. Do you think we should put this as a bugfix
into the 1.13 release branch?
> SplitEmumerator does not provide checkpoint id in snapshot
> ----------------------------------------------------------
>
> Key: FLINK-22133
> URL: https://issues.apache.org/jira/browse/FLINK-22133
> Project: Flink
> Issue Type: Bug
> Components: Connectors / Common
> Affects Versions: 1.12.0
> Reporter: Brian Zhou
> Priority: Major
>
> In ExternallyInducedSource API, the checkpoint trigger exposes the checkpoint
> Id for the external client to identify the checkpoint. However, in the
> FLIP-27 source, the SplitEmumerator::snapshot() is a no-arg method. The
> connector cannot track the checkpoint ID from Flink
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
