Fedora 24 delay, and more openQA work

You probably read a more sensationalized version of this on Phoronix already, but we had the Fedora 24 go/no-go meeting today and decided to slip for a week. The sole reason for this was this blocker bug, which is to do with booting Windows from the Fedora boot menu on a UEFI system, wasn’t fixed in time. If you don’t need to do that, you could frankly go ahead and grab a Fedora 24 nightly and use it. It’ll be fine. Go ahead, have a ball.

Aside from helping our awesome new interns Sumantro and company with validation testing and running test days, I’ve been spending most of my time lately teaching openQA new stuff. The robot’s getting pretty good now! We’re testing install in Russian, running a terminal in KDE and GNOME, the Server default firewall configuration, Cockpit, FreeIPA enrolment via Cockpit, installing to an iSCSI target, and – well, I just finished this today – several NFS install tests.

Getting inter-dependent tests working in openQA has been a big help in letting us extend coverage, it’s been fun trying to knock out as many of the validation tests as possible, and we’re doing pretty well now. It’s definitely very nice not to have to painfully re-do all the setup for stuff like the FreeIPA and iSCSI tests over and over again.

I also reported and helped fix quite a few bugs that were found by these tests, including this fun keyboard layout bug which took a couple of days of research to figure out (well, I say ‘research’, one whole afternoon of it was manually bisecting the offending systemd commit with Fedora 21 RPM builds…)

Leave a Reply

Your email address will not be published. Required fields are marked *