tashi-rokt opened a new pull request, #3772:
URL: https://github.com/apache/celeborn/pull/3772
### What changes were proposed in this pull request?
Add a `ManualExcludedWorkerCount` master gauge that reports only workers
excluded through the manual exclusion mechanism.
The existing `ExcludedWorkerCount` remains unchanged and continues to report
the combined automatic and manual exclusion count. The new gauge is also
documented in the master metrics reference.
### Why are the changes needed?
`ExcludedWorkerCount` combines automatic and manual exclusions, so operators
cannot determine from metrics whether persistent manual exclusion state is
preventing otherwise healthy workers from becoming available.
The REST API exposes the two states separately, but diagnosing this
condition currently requires querying the active master. Exporting the manual
count separately makes the distinction observable through the existing metrics
system without changing exclusion behavior.
### Does this PR resolve a correctness bug?
- [ ] Yes
### Does this PR introduce _any_ user-facing change?
- [ ] Yes
### How was this patch tested?
- Added a `MasterSuite` test that verifies the emitted gauge follows manual
exclusion state from 0 to 1 and back to 0.
- Ran:
- `./build/mvn -pl master -am -Dtest=none
-DwildcardSuites=org.apache.celeborn.service.deploy.master.MasterSuite test`
- `./build/mvn spotless:check -Pgoogle-mirror,spark-3.3`
Both commands passed locally.
### Draft note
The CELEBORN Jira is still to be created. Replace `CELEBORN-XXXX` in the
title and link the Jira before marking this PR ready for review.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
