#41 RFE: rpm-scratch-build job could use some retrying
Closed 3 years ago by fbo. Opened 4 years ago by churchyard.

https://fedora.softwarefactory-project.io/zuul/build/9253d8fde8ed47b0adf6b5f0152dc5aa

[                                    ]  00% 00:00:00     0.00 B        - B/sec
[====================================] 100% 00:00:00 835.96 KiB  13.78 MiB/sec
Building python-setuptools-45.2.0-1.fc33.src.rpm for rawhide
Created task: 41970619
Task info: https://koji.fedoraproject.org/koji/taskinfo?taskID=41970619
Watching tasks (this may be safely interrupted)...
41970619 build (rawhide, python-setuptools-45.2.0-1.fc33.src.rpm): free
41970619 build (rawhide, python-setuptools-45.2.0-1.fc33.src.rpm): free -> open (buildvm-17.phx2.fedoraproject.org)
  41970620 rebuildSRPM (noarch): closed
  41970658 buildArch (python-setuptools-45.2.0-1.fc33.src.rpm, x86_64): open (buildhw-07.phx2.fedoraproject.org)
Could not execute build: 502 Server Error: Proxy Error for url: https://koji.fedoraproject.org/kojihub

This happens fairly often with Koji. I suppose there could be some try-sleep-retry loop to mitigate that. See also: https://pagure.io/fedora-ci/general/issue/76 for some more complex issues with koji watch task.


Hi,

I've merged that change (can't be tested speculatively as rpm-scratch-build is defined in a config project) https://pagure.io/zuul-distro-jobs/pull-request/49 That should make the task more resilient. Let me know if you still see any issues.

To understand this better, that fix basically starts a new scratch build on such occasion, right?

Yes, based on the command output Ansible will decide to re-run (Server Offline, ServerError) or not (legitimate build failure) the koji scratch build task.

Please re-open if the issue is still there.

Metadata Update from @fbo:
- Issue status updated to: Closed (was: Open)

3 years ago

Login to comment on this ticket.

Metadata