Skip to content

Conversation

@AlanCoding
Copy link
Member

@AlanCoding AlanCoding commented Dec 9, 2025

SUMMARY

This is a mop-up of any remaining issues I saw while trying to do performance testing related to dispatcherd.

We had a case where scale-ups could result in missed messages, making these messages really important. Although the lost messages were recovered, it was still a pretty bad bug. It should be fixed now, but the logs are also valuable to keep.

This also adds some re-scheduling logic, which is going off of a gut intuition more than anything else. This did not ever cause a delay in job starting, and in practice it was always quite fast.

ISSUE TYPE
  • Bug, Docs Fix or other nominal change
COMPONENT NAME
  • API

Note

Batch-dispatch up to 25 waiting jobs with timing/id logging and self-reschedule; update bulk job launch test to override max launch setting.

  • Backend (tasks):
    • dispatch_waiting_jobs in awx/main/tasks/jobs.py:
      • Limit to dispatching up to 25 waiting jobs per run.
      • Collect dispatched job IDs and log IDs with elapsed time.
      • Self-reschedule when multiple jobs are dispatched; debug-log when none.
  • Tests:
    • awx/main/tests/functional/test_bulk.py:
      • test_bulk_job_launch_queries uses override_settings(BULK_JOB_MAX_LAUNCH=num_jobs + 1) to allow requested job count.

Written by Cursor Bugbot for commit f903aad. This will update automatically on new commits. Configure here.

@sonarqubecloud
Copy link

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant