[
https://issues.apache.org/jira/browse/FLINK-17012?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17308796#comment-17308796
]
Piotr Nowojski commented on FLINK-17012:
----------------------------------------
With some offline discussions with [~trohrmann], we agreed upon adding new
state to recovering/initialising, which would also encompass recovering from
unaligned checkpoints.
For the technical solution, we decided to not introduce factories for a time
being and just stick with exposing recovering as a separate method from the
{{AbstractInvocable}}.
> Expose stage of task initialization
> -----------------------------------
>
> Key: FLINK-17012
> URL: https://issues.apache.org/jira/browse/FLINK-17012
> Project: Flink
> Issue Type: Improvement
> Components: Runtime / Task
> Reporter: Wenlong Lyu
> Assignee: Anton Kalashnikov
> Priority: Major
> Labels: pull-request-available
>
> Currently a task switches to running before fully initialized, does not take
> state initialization and operator initialization(#open ) in to account, which
> may take long time to finish. As a result, there would be a weird phenomenon
> that all tasks are running but throughput is 0.
> I think it could be good if we can expose the initialization stage of tasks.
> What to you think?
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
