Closed gparvin closed 2 years ago
Just updating the apiserver didn't seem to fix it. Trying oc edit kubecontrollermanager cluster
and making the same update there now. After making the update in both I'm not seeing this problem anymore.
Is there a plan to fix this issue? It is really annoyed because running the automation repeatedly may result in problems.
I am working on testing a patch now.
Information from my initial investigation of this issue: https://gist.github.com/rooftopcellist/f521164446ab554a1d13410138986549
In OCP 4.8 https://kubernetes.io/docs/concepts/workloads/controllers/ttlafterfinished/ is enabled by default. This means jobs get deleted after an hour. When the job gets deleted, it seems the AnsibleJob wants to re-run the automation again. This causes the automation to run hourly.
The feature gate for
TTL Controller for Finished Resources
can be disabled which I think will prevent this from happening. I'm trying that out now. To disable you do this:Run:
oc edit kubeapiservers.operator.openshift.io cluster
and then update theunsupportedConfigOverrides
It would be nice if the AnsibleJob status could be used to prevent the re-run to allow job cleanup to work.