Create hdd for libyui tests

[sofiasyria]

• Related gitlab PR: https://gitlab.suse.de/qsf-y/qa-sle-functional-y/-/merge_requests/352
• Related ticket: https://progress.opensuse.org/issues/90368
• Verification run:
  SLE: https://openqa.suse.de/tests/overview?distri=sle&version=15-SP3&build=_ssyrianidou_2&groupid=96
  Tumbleweed: https://openqa.opensuse.org/tests/1731761
  https://openqa.opensuse.org/tests/1731762
  https://openqa.opensuse.org/tests/1731763

[jknphy]

Code LGTM. Some things I spot:

• Wondering if we should try with another arch in SLE, but we just might configure it in prod, the creation process is the same for the rest of archs.
• Check typo in commit "Remove scc_cneanup from schedule yast2_gui"
• Does it work for firstboot, I don't see it clearly in latest verification, perhaps is outdated.
• Similar for tw, could you share verification for https://openqa.opensuse.org/tests/1729867 so we don't miss anything. Did you find any difference for Leap https://openqa.opensuse.org/tests/1724090 ?

[sofiasyria]

* Does it work for firstboot, I don't see it clearly in latest verification, perhaps is outdated.

* Similar for tw, could you share verification for https://openqa.opensuse.org/tests/1729867 so we don't miss anything. Did you find any difference for Leap https://openqa.opensuse.org/tests/1724090 ?

I had some issues yesterday with my workstation. It doesn't see any needles and tests fail. I will move all teh vrs to development group today

[jknphy]

I noticed in the VR that for s390x the name that is publish and the name that is picked in the child is different, therefore fails.
For the other ones failing I noticed you scheduled again, but check the WORKER_CLASS, it has two values, it should be fine removing one. If you can adjust that would be great, to be safe.
Other than that LGTM.

[sofiasyria]

I noticed in the VR that for s390x the name that is publish and the name that is picked in the child is different, therefore fails.

The test doesn't publish any image, for some reason I don't understand. I will try to add svirt_upload_assets in the schedule.

For the other ones failing I noticed you scheduled again, but check the WORKER_CLASS, it has two values, it should be fine removing one. If you can adjust that would be great, to be safe.

I changed that yesterday. Now it fails because of some scc problem. The test cannot install yast2-firstboot.

[jknphy]

you have custom SCC_URL, instead of http://all-187.1.proxy.scc.suse.de which probably is affecting registration.
for the images in s390x looks like isos post was not triggered right, because there is not dependency tab in https://openqa.suse.de/tests/6007410

[sofiasyria]

you have custom SCC_URL, instead of http://all-187.1.proxy.scc.suse.de which probably is affecting registration.

The custom scc url is not the problem here. I have the same url for x86_64 and it works. The firstboot tests don't run for other arch than x86_64 in productions, so there is a change that our scc never had those packages for ppc and aarch64. I will try to ask someone from scc team. But these failed tests shouldn't worry us for the merging of the PR.

for the images in s390x looks like isos post was not triggered right, because there is not dependency tab in https://openqa.suse.de/tests/6007410

What dependency is missing? How would a child test affect the image of the parent test from publishing?

[jknphy]

LGTM