There's a flashing cursor, so we never actually succeed on any of
these, they all hit the 30 second timeout. Similarity when the
cursor flashes seems to be just over 38, so let's go with that.
Signed-off-by: Adam Williamson <adamwill@fedoraproject.org>
The prerelease note check only kicks in on candidate composes, so
we get to update the needles now. Also we get slightly different
KDE notification appearances for some reason.
Signed-off-by: Adam Williamson <adamwill@fedoraproject.org>
This test is now extremely unreliable on x86_64 on f43 also, not
just aarch64 as before. I've no idea why. Let's just make it a
soft failure as we did for aarch64; I don't have time to look
into it in detail.
Signed-off-by: Adam Williamson <adamwill@fedoraproject.org>
The updates-archive repo was fixed a while back, but now openh264
repo for Rawhide is broken again. So swap this out to disable the
openh264 repo for Rawhide tests.
Signed-off-by: Adam Williamson <adamwill@fedoraproject.org>
Seems it can take more than 30 seconds after clicking the link
for gnome-control-center to run and switch to the desired tab.
Signed-off-by: Adam Williamson <adamwill@fedoraproject.org>
We seem to be exceeding it at least sometimes, e.g.
https://openqa.fedoraproject.org/tests/5197417 has a "No space
left on device" error when copying to /var/tmp , which is the
mounted scratch disk.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
While https://koji.fedoraproject.org/koji/taskinfo?taskID=149431017
is building, all Rawhide tests are failing because it cannot be
installed as it'd break libreoffice-core's deps. Drop --best for
now to work around this. Remove this when that build is stable.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
The needle @lruzicka added effectively makes this "click the
box and make sure it's active before we start typing" workaround
I added recently a no-op, so it's wrong. The problem seems to be
that in at least one test, the box *never* went active:
https://openqa.fedoraproject.org/tests/5189380#step/_boot_to_anaconda/9
so instead, let's try clicking another time if the box doesn't go
active the first time.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
ac97 doesn't work on aarch64. es1370 seems to work on both so
we'll go with that. We also need a new needle on aarch64 for
some reason, the icon is slightly different.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
We no longer need to go into the kebab menu, the link to custom
storage is right there on the page now.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
rpm-ostree is fixed now. Yay. We need a small workaround in the
overlay test for F45 updates-archive repo not existing.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
This works around the weird
https://bugzilla.redhat.com/show_bug.cgi?id=2456660 /
https://gitlab.freedesktop.org/pipewire/wireplumber/-/work_items/946
which turns out to be somehow associated with the specific qemu
audio device os-autoinst defaults to - intel-hda. I played around
and could not reproduce the issue on bare metal or in a standard
virt-manager VM (which uses an ich9-intel-hda device and some
other different options to what os-autoinst does). So I tried a
few options on staging, and found that using ac97 avoids the
problem.
There's probably still some kind of real bug here, but I'm not
sure I care enough to figure out what it is. It seems pretty
clear that, on the whole, the panel volume indicator works.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
We're commonly hitting test failures on webui on the language
select screen because the system is still heavily loaded when
the language select dialog is first visible. Let's try asserting
its active state and then waiting for the screen to be still for
ten seconds (this is in case an 'application is unresponsive'
popup appears) before we start typing.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
FreeIPA in F43 got ahead of later releases, which breaks FreeIPA
upgrade tests. I've done bump builds for all releases, they need
to go here till they go stable.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
They're known broken ATM, the failures are cluttering up the
overview, and I can't bodge together a fix.
See: https://github.com/coreos/rpm-ostree/issues/5623
Signed-off-by: Adam Williamson <awilliam@redhat.com>
