QA (public) » openQA Project (public)

Observation¶

When there are investigation jobs that are cancelled or obsoleted then the investigate:retry job is retrying 1440 times with a delay of one minute before giving up:
https://openqa.suse.de/minion/jobs?id=15293289

If many of those jobs exist, Minion will spend a lot of time retrying those instead of working on other jobs that could run and do useful stuff.

Maybe openqa-investigate can detect that jobs are cancelled.

• AC1: If investigate jobs are cancelled, openqa-investigate detects that and creates the final comment without waiting any longer

---
args:
- env ...
  grep_timeout=60 nice ionice -c idle /opt/os-autoinst-scripts/openqa-label-known-issues-and-investigate-hook
- 17444939
- delay: 60
  kill_timeout: 10s
  retries: 1440
  skip_rc: 142
  timeout: 10m
attempts: 1
children: []
created: 2025-04-24T15:48:34.951746Z
delayed: 2025-04-25T10:20:42.619188Z
expires: ~
finished: ~
id: 15293289
lax: 0
notes:
  hook_cmd: env ...
    grep_timeout=60 nice ionice -c idle /opt/os-autoinst-scripts/openqa-label-known-issues-and-investigate-hook
  hook_rc: 142
  hook_result: ''
parents: []
priority: 0
queue: default
result: ~
retried: 2025-04-25T10:19:42.619188Z
retries: 921
started: 2025-04-25T10:19:35.150145Z
state: inactive
task: hook_script
time: 2025-04-25T10:20:24.275635Z
worker: 1989

Suggestions¶

• Note that the main investigate:retry job might be cancelled itself, or one of the other jobs. Handling both cases might require different changes to the code