[
https://issues.apache.org/jira/browse/FLINK-17128?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17083013#comment-17083013
]
Kurt Young commented on FLINK-17128:
------------------------------------
Yes, I think [~lzljs3620320] is right.
One of the essential requirements here is to distinguish the possible failure
recovery model here, i.e. state+checkpoint for typical streaming cases and
buffer+commit after job finished successfully for some batch cases.
Only have information about boundness is not enough. For example, a bounded
streaming job which will do a per-record emit might still rely on the
state+global snapshot failure recovery model.
> Add isBounded to TableSinkFactory#Context
> -----------------------------------------
>
> Key: FLINK-17128
> URL: https://issues.apache.org/jira/browse/FLINK-17128
> Project: Flink
> Issue Type: Sub-task
> Components: Table SQL / API
> Affects Versions: 1.11.0
> Reporter: Jingsong Lee
> Assignee: Jingsong Lee
> Priority: Major
> Labels: pull-request-available
> Fix For: 1.11.0
>
> Time Spent: 10m
> Remaining Estimate: 0h
>
> For current table environment, the difference between batch and streaming is
> very clear.
> We can use this bool to distinguish batch and streaming.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
