action #168775
closed[containers] Prepare 15-SP7 Container Host
0%
Description
With Beta1 in the Pipeline we should slowly get the 15-SP7 Container host in https://openqa.suse.de/group_overview/377 going.
Necessary parts to achieve this:
- Add job group definitions in https://gitlab.suse.de/qac/qac-openqa-yaml/-/blob/master/containers/create_stable_hosts.yaml?ref_type=heads (Might require adding Medium types to OSD)
- Add a trigger to https://gitlab.suse.de/qac/aceto-magico
Acceptance criteria¶
- AC1: Weekly asset generation job for 15-SP7 is being scheduled and passes
Files
Updated by mdati 2 months ago · Edited
MR https://gitlab.suse.de/qac/qac-openqa-yaml/-/merge_requests/1904
Asset generation job for 15-SP7 GM HDD_1
Products/Medium created in OSD
Updated by mdati 2 months ago · Edited
Merged MR 1904. Being SLE 15-sP7 GM not ready yet, ISO from group 313 used.
MR https://gitlab.suse.de/qac/aceto-magico/-/merge_requests/22 to add new HDD generation for SLE 15-sP7 too.
CREATED and MERGED.
Steps n.1 and n.2 in header completed.
Updated by mdati 2 months ago · Edited
- Status changed from In Progress to Feedback
Weekly asset generation job next run should pass and create HDD_1 qcow2 for SLE 15-SP7:
https://openqa.suse.de/tests/overview?distri=sle&version=15-SP7&build=GM&groupid=377
Updated by mdati 2 months ago · Edited
- Status changed from Feedback to In Progress
The test SLE 15-SP7 in group 377 failed, because of 2 new needles needed and a new case occurred in sles checks:
- FAIL in autoyast/installation, created new needles: installation-autoyast-boot-20241105, installation-autoyast-boot-text-20241105.
- FAIL in autoyast/login, created new needle: login-autoyast-system-login-console-SLE-installation-20241105.
- Fail in containers/install_updates: created new PR https://github.com/os-autoinst/os-autoinst-distri-opensuse/pull/20560
Updated by mdati 2 months ago · Edited
Issue in containers/install_updates, PR 20560, resulted a product bug, as discussed in slack: open bsc#1232902 and PR updated to softfail until issue present.
A VR on that branch passed the point, but then continuing testing the image on zypper:command not found. It resulted installed outside the PATH, in /usr/lib/zypper.
Updated by mdati about 2 months ago · Edited
Updated by mdati about 2 months ago · Edited
All scheduled tests for ARCH=ppc64le don't start.
Noted that those machines result in an area NOT zone-cc (recentily introduced).
To make it run, hmc type can be used, ppc64le+hmc, applying proper changes in settings.
Updated by mdati about 2 months ago · Edited
- File install-15sp7.png install-15sp7.png added
NOTE about beta install needles notes 7,8:
Finally, adding in test settings BETA=1, the correct needles and tags were normally selected and the expected login console appeared and also os-release with all the correct data. The irregular os-release configuration never appeared anymore in tests run and tests pass ok.
The PR 20560 has been closed.
Consider that, new temporary needles were introduced to manage beta-release-boot and after beta warning pop up, the installation resulted skipping or interrupting packages loading.
As result, an ISO installation console shell appeared tty6 :/, without a login and this shell contained the irregular os-release of bsc#1232902.
Noted that, stopping that test, debugging on vnc, the wrong console did't appear spontaneously (probably because the paused shell had all the time to proceeded on package installation, setting the right configuration), but manually working,even the expected login with user/password appeared.
The only way to have that shell was to pause the test in vnc and in install menu select: install_Abort/ Expert/ Start shell.
See 15sp7 install shell screenshot of SLE-15-SP7-Online-x86_64-Build36.1-Media1.iso
Updated by mdati about 2 months ago · Edited
The ppc64le waiting tests discussion is ongoing: tests not able to start being qemu_ppc64le external to zone-cc. hmc seems not the right way.
See post.
Updated by mdati about 2 months ago
- Status changed from In Progress to Feedback
Actual ticket status is all tests in https://openqa.suse.de/tests/overview?distri=sle&version=15-SP7&build=GM&groupid=377 running ok,
but only ppc64le cannot start yet, due to recent zone-cc limitations; that last issue should be managed in another dedicated ticket.
Updated by ph03nix about 1 month ago
- Status changed from Feedback to In Progress
@mdati we can use emulated powerpc workers, see e.g. https://openqa.suse.de/tests/16007929 for the PowerPC test runs for now.
Updated by pherranz about 1 month ago
- Related to action #170248: [Containers] Fix asset generation jobs for PPC64 added
Updated by mdati about 1 month ago
Indeed, if we should emulated powerpc workers, we can refer to previous work in https://progress.opensuse.org/issues/163451#note-9.
Updated by mdati about 1 month ago · Edited
Today status in group 377 for SLE 15-SP7, All tests pass for all the architectures, including ppc64le , without emaulation, being issues related to the CC zone, a.t.m. not present anymore.
The tests triggering is already included in aceto magico, since the MR https://gitlab.suse.de/qac/aceto-magico/-/merge_requests/22 merged.
Updated by mdati about 1 month ago · Edited
- Status changed from Feedback to Resolved
Tests in group 377 for SLE 15-SP7, are all set: a single failure noted on s390x-FIPS part will be managed in https://progress.opensuse.org/issues/173557 related to bsc https://bugzilla.suse.com/show_bug.cgi?id=1233233