[
https://issues.apache.org/jira/browse/FLINK-29195?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17600790#comment-17600790
]
chenyuzhi edited comment on FLINK-29195 at 9/6/22 2:21 PM:
-----------------------------------------------------------
Thanks for your tips! We could do get the latest checkpoint id from the REST
API.
However, in a general scenior, we may collect metrics of flink application by
various metric reporters(like jmx/prometheus) and build some alert basing on
them. In this case maybe we can't report checkpoint id using REST API directly.
[~masteryhx]
was (Author: stupid_pig):
Thanks for your tips! We could do get the latest checkpoint id from the REST
API.
However, in a general scenior, we may collect metrics of flink application by
various metric reporters(like jmx/prometheus), and build some alert basing on
them. In this case maybe we can't report checkpoint id using REST API directly.
[~masteryhx]
> Expose lastCheckpointId metric
> -------------------------------
>
> Key: FLINK-29195
> URL: https://issues.apache.org/jira/browse/FLINK-29195
> Project: Flink
> Issue Type: New Feature
> Components: Runtime / Checkpointing, Runtime / Metrics
> Reporter: chenyuzhi
> Priority: Major
>
> May be we could expose the lastCheckpointId metric.
> With this metric, we could monitor for flink application and get notice when
> the job restoring from an unexpect checkpoint, instead of extract
> checkpointId from the metric lastCheckpointExternalPath
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
