[
https://issues.apache.org/jira/browse/FLINK-19010?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17732058#comment-17732058
]
Yanfei Lei commented on FLINK-19010:
------------------------------------
[~zncheng] Since FLINK-23976 added standardized metrics for capturing how much
time we spend in each JobStatus, I think we can reuse
{{initializingTime/initializingTotalTime}} as the job level restoring time
directly, WDYT?
And I open a PR to introduce `checkpointRestoreTime` for the task level
restoring, could you please help take a look?
> Add a system metric to show the checkpoint restore time
> -------------------------------------------------------
>
> Key: FLINK-19010
> URL: https://issues.apache.org/jira/browse/FLINK-19010
> Project: Flink
> Issue Type: Improvement
> Components: Runtime / Checkpointing, Runtime / Metrics
> Reporter: Zhinan Cheng
> Assignee: Yanfei Lei
> Priority: Minor
> Labels: pull-request-available
>
> Now the system metric only shows the downtime when failure happens. It would
> be interesting to see the time to restore the checkpoint, so users can better
> understand the bottleneck of failure recovery.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
