fedora-infrastructure

#11247 fedora-bodhi-critpathcron job failures
Closed: Fixed by kevin. Opened by mattia.

Closed: Fixed
Reopen Issue

Describe what you would like us to do:

Since a week or so I'm getting alerts about fedora-bodhi-critpathcron job failures.
However, I'm unable to view alerts details because it seems I have no access to them.
Can someone take a look and check what's going on and what's the impact of these alerts?

[FIRING:1] CronjobFailed bodhi (kube-rbac-proxy-main fedora-bodhi-critpathcron https-main kube-state-metrics openshift-monitoring/k8s kube-state-metrics warning)
Labels
alertname = CronjobFailed
container = kube-rbac-proxy-main
cronjob = fedora-bodhi-critpathcron
endpoint = https-main
job = kube-state-metrics
namespace = bodhi
prometheus = openshift-monitoring/k8s
service = kube-state-metrics
severity = warning
Annotations
description = Latest execution of CronJob bodhi/fedora-bodhi-critpathcron failed to complete.
summary = Latest job fedora-bodhi-critpathcron has failed.

https://console-openshift-console.apps.ocp.fedoraproject.org/monitoring/#/alerts?receiver=bodhi%2Fappowners-alerts%2Fdefault

When do you need this to be done by? (YYYY/MM/DD)

Metadata Update from @jnsamyak:
- Issue tagged with: Needs investigation

Metadata Update from @jnsamyak:
- Issue assigned to jnsamyak

here is the log: https://humaton.fedorapeople.org/fedora-bodhi-critpathcron-28029790-ggmc8-critpath.log

Yes, I already looked at this, should have tried to communicate to folks that also got the alert.

I'm pretty sure it's due to the pre-bitflip f38 directory. Its trying to download direct from the master mirrors, but f38 dir wasn't open for it.

I just fired off a manual run, but I think it will just finish.

Yep. It finished fine now.

So, if we want to fix this to not happen during the next cycle we need to not query bodhi for this info or try both releases/N and development/N or something.

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

sigh. we know about this little sync problem between the mirrors and the bodhi data, but I didn't foresee that it'd be a problem for the critpath script.

it will fix itself once f38 is really released (i.e. now), but I can probably do...something...so this doesn't happen next time. Let me poke at it.

https://pagure.io/releng/pull-request/11392 should fix this for future cycles, I hope.
Metadata

Needs investigation

None
None
Lowest
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.