Things Fedora QA Robots Do
summary
Fedora QA runs a set of small fedmsg consumers that detect compose and update events and trigger concrete actions: openQA scheduler hits openQA to run tests, reporters forward results to resultsdb and edit release-validation wiki pages via wiki TCMS, check_compose generates per-compose reports and relval creates validation-event pages. Robots deploy via Ansible on openQA hosts and mostly work but occasionally fail when fedmsg, fedfind or auth break. Several manual tasks remain — update trackers and blocker-milestone updates — because no reliable programmatic “new release” event exists.
AI-written summary · about
chapters (8)
- 0:00Start
- 1:32Fedmsg
- 3:42Fedfind etc.
- 14:58relvalconsumer
- 17:37autocloudreporter
- 20:28Cl pipeline
- 26:14blockerbugs
- 26:49Summary
related talks
29mTowards stronger checks on dependencies and compose inputs in Fedora
15mFedora Zuul CI - an update
01h 05mFedora CI Meetup
14mUsing packit and testing farm for CI/CD and automation in an upstream project
13mJourney of Automation - Github, Galaxy, Fedora
20mAutomating maintainers work: The Prototype
25mCI/CD for Fedora packaging with Zuul
47m35 Fedora Releases in 30 Minutes (plus Q&A!)