Flatpak builds have stopped working again. This time it's the container build failing to connect to https://odcs.fedoraproject.org/api/1/composes/ and timing out after 3600 seconds:
$ fedpkg flatpak-build Created task: 50937779 Task info: https://koji.fedoraproject.org/koji/taskinfo?taskID=50937779 Watching tasks (this may be safely interrupted)... 50937779 buildContainer (noarch): free 50937779 buildContainer (noarch): free -> open (buildvm-x86-03.iad2.fedoraproject.org) 50937779 buildContainer (noarch): open (buildvm-x86-03.iad2.fedoraproject.org) -> FAILED: Fault: <Fault 2001: 'Image build failed. Error in plugin resolve_module_compose: Retrieving https://odcs.fedoraproject.org/api/1/composes/4549 timed out after 3600 seconds. OSBS build id: flatpak-runtime-f32-8171a-5'> 0 free 0 open 0 done 1 failed 50937779 buildContainer (noarch) failed
Flatpak builds last worked fine on Friday.
Metadata Update from @siddharthvipul1: - Issue tagged with: medium-gain, medium-trouble, ops
Quite odd. Nothing I know of changed... and I can curl that url just fine from the builder.
@cverna any ideas here?
Metadata Update from @mohanboddu: - Issue priority set to: Waiting on Assignee (was: Needs Review)
This usually means that odcs did not manage to finish the compose. In orchestrator.log there is a bit more info.
2020-09-07 11:43:39,124 platform:- - atomic_reactor.odcs_util - DEBUG - Getting compose information for information for compose_id=4549 2020-09-07 11:43:39,130 platform:- - atomic_reactor.odcs_util - DEBUG - Retrying request compose_id=4549, elapsed_time=0.00635981559753418
You can use the compose_id to check on ODCS https://odcs.fedoraproject.org/api/1/composes/4549 and it seems that the compose state is still "wait". I am not sure who could help us with ODCS :-)
@jkaluza ^
odcs-celery-backend.service on odcs-backend01.iad2.fedoraproject.org was down for some reason. I've started it using systemctl start odcs-celery-backend.service.
systemctl start odcs-celery-backend.service
Nice, thanks @jkaluza!
Just tested a build and this looks fixed. Thanks again everybody involved!
Metadata Update from @kalev: - Issue close_status updated to: Fixed - Issue status updated to: Closed (was: Open)
Log in to comment on this ticket.