action #160089
closed
coordination #157969: [epic] Upgrade all our infrastructure, e.g. o3+osd workers+webui and production workloads, to openSUSE Leap 15.6
Handle uncommented package lock on "kernel-default" and "kernel-default-base" on openqa-piworker
Added by okurz 7 months ago.
Updated 7 months ago.
Category:
Regressions/Crashes
Description
Motivation¶
While upgrading machines from Leap 15.5->15.6 I encountered package locks for "kernel-default" and "kernel-default-base" on openqa-piworker which I left in place while conducting the upgrade. Both package locks don't have a description so it's not clear why we have those.
# zypper ll
# | Name | Type | Repository | Comment
--+---------------------+---------+------------+--------
1 | kernel-default | package | (any) |
2 | kernel-default-base | package | (any) |
Acceptance criteria¶
- AC1: There are no uncommented package locks on openqa-piworker
Suggestions¶
- Ask around and look in old tickets why we have those package locks
- As necessary remove the package locks and ensure that we run an upgraded kernel without problems or have an according comment with bug or ticket reference explaining why we have the package lock in place and what would be necessary to remove the lock again
- Copied from action #157996: Upgrade all other LSG QE salt controlled machines to openSUSE Leap 15.6 added
- Assignee set to jbaier_cz
- Related to action #132860: openqa-piworker is unstable and needs regular power-cycles size:M added
- Status changed from New to Resolved
- Due date set to 2024-05-27
- Status changed from Resolved to Feedback
Let's please be explicit about the kernel upgrade then. https://gitlab.suse.de/openqa/salt-pillars-openqa/-/merge_requests/806 was removed but so far openqa-piworker still runs the old kernel. I would like to have an explicit open ticket in case there are problems with the new kernel or if the system does not survive the next reboot. If openQA jobs are executed fine after rebooting into the new kernel then everything should be good and you can resolve.
$ zypper ll | grep poo
1 | kernel-default | package | (any) | poo#132860
2 | kernel-default-base | package | (any) | poo#132860
$ systemctl cat auto-update | grep Exec
ExecStart=/bin/sh -c 'zypper -n --non-interactive-include-reboot-patches patch --r
eplacefiles --auto-agree-with-licenses --download-in-advance && needs-restarting -
-reboothint >/dev/null || (command -v rebootmgrctl >/dev/null && rebootmgrctl rebo
ot ||:)'
I took a quick look as I was surprised the kernel wasn't replaced. It looks like it's not running auto-update.sh and the update-update timer is inactive. Some inconsistency?
because openqa-piworker is not an openQA worker the service that should be running is auto-upgrade, not auto-update. And I checked that auto-upgrade.timer is enabled and last ran successfully today in the morning.
okurz wrote in #note-7:
because openqa-piworker is not an openQA worker the service that should be running is auto-upgrade, not auto-update. And I checked that auto-upgrade.timer is enabled and last ran successfully today in the morning.
Right. I guess I was wrong to assume we manage locks on that machine, so all (not) working as expected.
So openqa-piworker has the new kernel installed but hasn't rebooted yet. @jbaier_cz please either explicitly reboot or just monitor what happens after the next automatic reboot. sudo rebootmgrctl status says
Status: Reboot requested, waiting for maintenance window
which would be next Sunday.
Yeah, I did as well checked the machine this morning to see the kernel is successfully installed after me removing the locks manually yesterday. As the reboot window is nearby, I will keep it alone for the weekend and we will see on Monday.
- Status changed from Feedback to Resolved
I guess all good here.
openqa-piworker:~> uname -a
Linux openqa-piworker 6.4.0-150600.20-default #1 SMP PREEMPT_DYNAMIC Thu May 9 20:34:03 UTC 2024 (ed3fdbb) aarch64 aarch64 aarch64 GNU/Linux
- Due date deleted (
2024-05-27)
Also available in: Atom
PDF