Hi folks! Just wanted to let you know that Bodhi's "Re-Trigger Tests" mechanism should now also work for tests run by openQA. Previously, it only worked for tests run by Fedora CI.
From now on, if you hit "Re-Trigger Tests" or run `bodhi updates trigger-tests` on an update that openQA has run some tests on, it will re-schedule the tests.
Technically speaking, it finds all 'flavors' (these are "updates- container", "updates-everything-boot-iso" etc.) for which tests have previously been run on the update, and schedules a new run of the tests for those flavors. This means that in a corner case (where we changed the test definitions themselves since the previous run(s)) the exact same set of tests may not be run again. It also means that *all* tests are always restarted, not just failed ones.
Please let me know, and/or file an issue at https://pagure.io/fedora-qa/fedora_openqa/issues , if you have any problems with this.
Thanks!