I noticed some recent failed tasks on CBS:
https://cbs.centos.org/koji/taskinfo?taskID=4200875 https://cbs.centos.org/koji/taskinfo?taskID=4200876 https://cbs.centos.org/koji/taskinfo?taskID=4200877
These are all newRepo for hyperscale8s tags. They're failing, which makes sense as c8s is EOL. This isn't causing any real issue for us, just pointing it out in case there's something that needs to be fixed on the CBS side.
Looks like this isn't just Hyperscale, looks like other SIGs with c8s tags also show the same: https://cbs.centos.org/koji/tasks?state=failed&view=tree&method=all&order=-id
Metadata Update from @arrfab: - Issue assigned to arrfab
Metadata Update from @arrfab: - Issue tagged with: cbs, low-gain, low-trouble
yes, but as you mentioned it's not impacting any SIG. The reason is that while we have excluded *el8s* in the kojira ignored_tags (see https://docs.pagure.org/koji/utils/#kojira) it's still taking into account *el8* because of SIGs still building for RHEL8 itself. So only the build tags that were pointing to CentOS Stream 8 are failing when kojira (you see that it's the service account cbs-kojira that is the task owner) is trying to process the build tags.
*el8s*
*el8*
cbs-kojira
Let me just identify these old c8s tags that were created with el8-build as suffix instead of el8s-build and rename the tags (even if all locked) Other option would be to just delete the tags completely but still having these displayed (archived) is better (imho) than just deleting everything.
el8-build
el8s-build
renamed the impacted old build tags with : cbs list-tags|grep 8s|grep el8-build| while read build_tag ; do cbs edit-tag --rename=${build_tag/el8-build/el8s-build} ${build_tag} ; done
cbs list-tags|grep 8s|grep el8-build| while read build_tag ; do cbs edit-tag --rename=${build_tag/el8-build/el8s-build} ${build_tag} ; done
That should then match the *el8s* glob for ignored_tags in kojira and it should then completely ignored these old build tags ..
ignored_tags
Closing but let me know if you that would work for you
Metadata Update from @arrfab: - Issue close_status updated to: Fixed with Explanation - Issue status updated to: Closed (was: Open)
Yep that looks good, thanks!
Log in to comment on this ticket.