[
https://issues.apache.org/jira/browse/FLINK-39490?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Robert Metzger reassigned FLINK-39490:
--------------------------------------
Assignee: Purushottam Sinha
> [web dashboard] Improve feedback when cluster is unreachable
> ------------------------------------------------------------
>
> Key: FLINK-39490
> URL: https://issues.apache.org/jira/browse/FLINK-39490
> Project: Flink
> Issue Type: Improvement
> Components: Runtime / Web Frontend
> Reporter: Purushottam Sinha
> Assignee: Purushottam Sinha
> Priority: Minor
> Labels: pull-request-available
>
> Right now when the Flink cluster becomes unreachable — the JobManager going
> down, a network blip, or a proxy returning 5xx — the dashboard doesn't really
> tell you what's going on. You just end up with stale data sitting there and
> individual errors popping up for whatever request happened to fail, which
> doesn't make it clear that the cluster itself is the problem.
>
> It'd be much nicer if the dashboard waited a few failed requests before
> deciding something is wrong, then showed a single clear "cluster unreachable"
> message. That one message should stay on screen for as long as the cluster is
> still down, go away on its own once things come back.
> FYI: Created as faced the issue in local development and testing.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
