releng

#13023 Some updates with unsigned builds stuck in bodhi
Closed: Fixed by kevin. Opened by decathorpe.

Closed: Fixed
Reopen Issue
  • Describe the issue

It appears that there's an issue that makes builds from submitted bodhi updates not get signed reliably. I reported this in the #releng channel on matrix a few days ago, but only noticed some rawhide updates (where this issue is more obvious) ... but there are a few more, that have now been ejected from bodhi composes:

  • https://bodhi.fedoraproject.org/updates/FEDORA-EPEL-2025-cb0c03a2ee
  • https://bodhi.fedoraproject.org/updates/FEDORA-2025-2da8bdd3cf
  • https://bodhi.fedoraproject.org/updates/FEDORA-EPEL-2025-72777ee065
  • https://bodhi.fedoraproject.org/updates/FEDORA-EPEL-2025-1e47baf611
  • https://bodhi.fedoraproject.org/updates/FEDORA-EPEL-2025-b81f71abf1
  • https://bodhi.fedoraproject.org/updates/FEDORA-EPEL-2025-f20474fc4b
  • https://bodhi.fedoraproject.org/updates/FEDORA-EPEL-2025-a71efec58d
  • https://bodhi.fedoraproject.org/updates/FEDORA-EPEL-2025-1e61345af3
  • https://bodhi.fedoraproject.org/updates/FEDORA-2025-b15ee8b005

I think this is a complete list, but I cannot guarantee that. Also looks like I was quite unlucky - 7 out of 9 affected updates are mine :(

It seems that this is associated with unreliable behaviour from koji recently, but I'm not sure how it could miss messages from the message bus - isn't the message queue architecture meant to prevent "I missed this" kind of issues?

  • When do you need this? (YYYY/MM/DD)

Not urgent but soon would be good. So let's say 2025/10/24?

  • When is this no longer needed or useful? (YYYY/MM/DD)

N/A / when all releases associated with those updates hit EOL / when new updates for those packages obsolete the stuck updates / whichever comes first.

  • If we cannot complete your request, what is the impact?

Some submitted updates not landing and not reaching buildroots and users.

Looks like this has partially resolved itself, but builds in one of the updates still contains unsigned builds.

Metadata Update from @amedvede:
- Issue tagged with: medium-gain, medium-trouble, sprint-3

It seems that this is associated with unreliable behaviour from koji recently, but I'm not sure how it could miss messages from the message bus - isn't the message queue architecture meant to prevent "I missed this" kind of issues?

If the message gets to rabbitmq yes. If the message doesn't get from koji to rabbitmq due to the timeout issue, then no.

The only one I see still having problems here is nodejs-aw-webui-0^20251017.f38b119-1.fc43. Will fix that one.

Can we close this ticket?

Metadata Update from @jnsamyak:
- Issue assigned to kevin

All these are stable or unpushed, so yeah...

Metadata Update from @kevin:
- Issue close_status updated to: Fixed
- Issue status updated to: Closed (was: Open)
Metadata

medium-gain medium-trouble sprint-3

None
None
None
Static archive. pagure.io was sunset in 2026; this is a read-only snapshot.
Red Hat Logo

Fedora is sponsored by Red Hat. Learn more about the relationship between Red Hat and Fedora »

© Red Hat, Inc. and others.