So, er, yeah....sorry :/ I was updating openQA to the latest tests and stuff today and accidentally updated the system completely. That seemed to stop openQA working properly - I don't think the openSUSE folks are testing the 'stable' release against current openSUSE any more, and I think some stuff changed in the more recent Mojolicious versions they've sent out which breaks the old 'stable' release.
So I tried to fix it for a bit and couldn't, and can't find a way to downgrade stuff back again. So I decided to just bump up to the latest 'unstable' openQA - it's working fine on happyassassin and upstream is actually running it on openqa.opensuse.org if you look (it's got the newer web UI and stuff). So I did that, and then went to reboot the machine to get a clean start...and it never seems to have woken up again. Sooo...yeah, I busted it. Sorry :( I tried poking around in the RH provisioning system where the box seems to live, but couldn't get it to wake up again (hope I didn't make it worse).
Can anyone who has access to the management console or whatever please get the machine up and running again? Sorry again!
On Tue, 2015-09-29 at 18:12 -0700, Adam Williamson wrote:
So, er, yeah....sorry :/ I was updating openQA to the latest tests and stuff today and accidentally updated the system completely. That seemed to stop openQA working properly - I don't think the openSUSE folks are testing the 'stable' release against current openSUSE any more, and I think some stuff changed in the more recent Mojolicious versions they've sent out which breaks the old 'stable' release.
So I tried to fix it for a bit and couldn't, and can't find a way to downgrade stuff back again. So I decided to just bump up to the latest 'unstable' openQA - it's working fine on happyassassin and upstream is actually running it on openqa.opensuse.org if you look (it's got the newer web UI and stuff). So I did that, and then went to reboot the machine to get a clean start...and it never seems to have woken up again. Sooo...yeah, I busted it. Sorry :( I tried poking around in the RH provisioning system where the box seems to live, but couldn't get it to wake up again (hope I didn't make it worse).
Can anyone who has access to the management console or whatever please get the machine up and running again? Sorry again!
OK, I managed to get it back up, and openQA is back up and running properly, I think. It's chewing through Final TC1 tests now.
Turns out the problem was that we had it using a repo called "devel:languages:perl" , which upstream no longer recommends - I guess it used to be needed to get some perl modules, but now it's too new, and the missing perl modules are in a special repo in the openQA OBS area, the one that https://github.com/os-autoinst/openQA/blob/master/docs/Installing.asciidoc calls "openQA-perl-modules". I ditched the other repo, added that one, did a 'zypper dup' (which is like distro-sync) and openQA stopped acting weird.
So we got a bit of an unplanned upgrade, but it does seem like everything is working, and the good news is now we're on a newer openQA which has the feature we need to use different disk images for the UEFI tests, and also has a patch I just got merged to allow batch job querying by ID so I can maybe tweak the 'wait for jobs to complete' stuff again.
sorry for the mess!
qa-devel@lists.fedoraproject.org
-
Adam Williamson