action #98243
closedsalt-states-openqa pipeline/job "test-worker" fails because zypper reports an error size:M
0%
Description
Observation¶
It seems that recently our pipeline started to fail. Example: https://gitlab.suse.de/openqa/salt-states-openqa/-/jobs/557702#L8526
Steps to reproduce¶
I was not able to reproduce this manually in a podman-container running on my workstation so I assume this has something to do with a change with our gitlab runners.
Problem¶
The failed testcases all include one of the following 2 comments:
One or more requisite failed: openqa.worker.worker-openqa.packages
User openvswitch is not available Group openvswitch is not available
/etc/logrotate.d/openvswitch: file not found
An error was encountered while installing package(s): Zypper command failure: Check Zypper's logs.
Judging by the last one, it may be the cause of the previous ones. It is the first error that is encountered.
Suggestion¶
- Check if the only problem is the zypper failure in the begining
- Can we retry the package installation in salt? (check salt docs)
- Can we retry the pipeline conditionally? (last resort)
Updated by livdywan over 3 years ago
These looks like different issues to me:
- https://gitlab.suse.de/openqa/salt-states-openqa/-/jobs/557702#L8526
User openvswitch is not available Group openvswitch is not available
https://gitlab.suse.de/openqa/salt-states-openqa/-/jobs/557704see #90275#note-35Unable to manage file: Jinja variable 'dict object' has no attribute 'openqaworker-arm-5'
Updated by livdywan over 3 years ago
- Subject changed from salt-states-openqa pipeline/job "test-worker" fails because zypper reports an error to salt-states-openqa pipeline/job "test-worker" fails because zypper reports an error size:M
- Description updated (diff)
- Status changed from New to Workable
- Target version set to Ready
Updated by nicksinger over 3 years ago
Since I couldn't reproduce locally I restarted the job of the pipeline for my MR at https://gitlab.suse.de/nicksinger/salt-states-openqa/-/pipelines/197290 and realized that it succeeded this time. I just restarted the same job for a different pipeline (https://gitlab.suse.de/nicksinger/salt-states-openqa/-/pipelines/197287) - if this also completes now I think we can reject this bug and assume we had some issue in our gitlab runner infrastructure.
Updated by nicksinger over 3 years ago
- Status changed from Workable to Rejected
Apparently the underlying issue got fixed as https://gitlab.suse.de/nicksinger/salt-states-openqa/-/pipelines/197287 also completed successfully. I couldn't find any hint what was changed which is kind of frustrating to just close a ticket as "now it works again". But well, at least MRs can be merged again with proper testing :)