[
https://issues.apache.org/jira/browse/FLINK-40530?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
ASF GitHub Bot updated FLINK-40530:
-----------------------------------
Labels: pull-request-available (was: )
> HistoryServerArchiveFetcher.updateJobOverview() aggregates all archived job
> overviews without per-entry error isolation, causing a single malformed
> archive to blank out the entire /jobs/overview response
> ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
>
> Key: FLINK-40530
> URL: https://issues.apache.org/jira/browse/FLINK-40530
> Project: Flink
> Issue Type: Bug
> Reporter: Archit Goyal
> Priority: Critical
> Labels: pull-request-available
>
> A single malformed/incompatible archive's overview JSON throws inside the
> aggregation loop, which has no per-entry isolation, so the whole
> /jobs/overview update aborts and returns empty/stale data for every job, not
> just the bad one.
> Fix: catch/skip per-entry failures and guard against wiping a good overview
> if all entries fail.
> Discovered during 1.x -> 2.x migration
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
