action #151286
closed
[Tools] Investigate s390x-zVM-vswitch-l2 worker issue: Could not retrieve required variable ZVM_GUEST_SUBNETMASK size:S
Added by zoecao about 1 year ago.
Updated about 1 year ago.
Description
Motivation¶
Recently the s390x zvm workers (WORKER_CLASS: s390x-zVM-vswitch-l2) don't work well, need to investigate the issue and help to fix it:
[2023-11-23T04:03:03.239974+01:00] [warn] [pid:111671] !!! main: Could not retrieve required variable ZVM_GUEST_SUBNETMASK at /usr/lib/os-autoinst/testapi.pm line 707.
Openqa failures on this issue:
https://openqa.suse.de/tests/12873356
https://openqa.suse.de/tests/12837164
https://openqa.suse.de/tests/12837748
Acceptance criteria¶
- AC1: All stakeholders are aware why these variables are not present anymore
Suggestions¶
- Ensure all scheduled product configs can be adapted
- Provide help as needed to make adjustments
- Maybe the error could be clearer? But, most likely this will go away soon so don't spend too much time (don't add new code)
- Tags set to infra
- Assignee set to nicksinger
- Priority changed from Normal to Urgent
- Target version set to Ready
the mentioned jobs have been manually restarted or were already scheduled, hence not getting updated variables from MACHINE.
I suggest to either retrigger the whole build for s390x or cancel currently scheduled jobs and wait for next build
- Subject changed from [Tools] Investigate s390x-zVM-vswitch-l2 worker issue: Could not retrieve required variable ZVM_GUEST_SUBNETMASK to [Tools] Investigate s390x-zVM-vswitch-l2 worker issue: Could not retrieve required variable ZVM_GUEST_SUBNETMASK size:S
- Description updated (diff)
- Status changed from New to In Progress
- Due date set to 2023-12-08
Setting due date based on mean cycle time of SUSE QE Tools
- Description updated (diff)
- Status changed from In Progress to Feedback
@zoecao did you observe any issues with jobs posted recently? As @mgriessmeier already mentioned these jobs where all scheduled before our changes and therefore cannot work and there is not much we can do about except rescheduling the whole product which is a bit much to get just 3 jobs working.
- Priority changed from Urgent to High
As @nicksinger provided a good explanation and explained that new production jobs are not impacted and only old retriggered ones reducing prio to "High"
- Description updated (diff)
- Due date deleted (
2023-12-08)
- Status changed from Feedback to Resolved
recovered nicksinger's unintentional description revert. There was no further feedback. Checking the original jobs we found now that jobs in the original scenarios have progressed further, e.g. two passed
good enough
Also available in: Atom
PDF