[
https://issues.apache.org/jira/browse/FLINK-12662?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16857340#comment-16857340
]
vinoyang commented on FLINK-12662:
----------------------------------
Hi [~till.rohrmann] I have thought a bit about the solution. Basically, the
design seems like the {{ArchivedExecutionGraph}} and {{AccessExecutionGraph}}.
IMO, we can reuse the {{AccessExecutionGraph}} for the purpose of rest API. In
addition, we need to introduce a new entity of the execution graph, named e.g.
{{AttemptedExecutionGraph}} to distinguish with {{ArchivedExecutionGraph}}
which need job state to be a globally terminal state. We can generate an
instance of {{AttemptedExecutionGraph}} when calling
{{ExecutionGraph#tryRestartOrFail}} method.
In addition, we may need a similar class named {{AttemptedExecutionGraphStore}}
to store the instances of {{AttemptedExecutionGraph}} and give an
implementation based on file and memory. And we may give a new entry in
{{ExecutionGraphCache}}.
This a coarse-grained idea, WDYT?
> show jobs failover in history server as well
> --------------------------------------------
>
> Key: FLINK-12662
> URL: https://issues.apache.org/jira/browse/FLINK-12662
> Project: Flink
> Issue Type: Improvement
> Components: Runtime / REST
> Reporter: Su Ralph
> Assignee: vinoyang
> Priority: Major
>
> Currently
> [https://ci.apache.org/projects/flink/flink-docs-release-1.8/monitoring/historyserver.html]
> only show the completed jobs (completd, cancel, failed). Not showing any
> intermediate failover.
> Which make the cluster administrator/developer hard to find first place if
> there is two failover happens. Feature ask is to
> - make a failover as a record in history server as well.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
