-
Bug
-
Resolution: Unresolved
-
Major
-
None
-
1.2.2
-
Quality / Stability / Reliability
-
False
-
-
None
-
None
-
None
-
None
-
None
-
None
-
None
-
None
Title: AlertManagerConfig is not taking effect in alertmanager pods deployed with coo
When creating alertmanagerConfig resource, and checking the configuration in the alertmanager pod. It shows the default config and not the desired one (alertmanagerConfig):
oc exec -n <namespace-name> <alertmanager-pod-name> -- amtool config show --alertmanager.url http://localhost:9093
The obo-prometheus-operator pod is trying to fetch the wrong secret for the default Alertmanager, leading to the following errors in its logs:
2025-10-16T16:35:38.102157285Z ts=2025-10-16T16:35:38.102113176Z level=info caller=/workspace/pkg/alertmanager/operator.go:787 msg="config secret not found, using default Alertmanager configuration" component=alertmanager-controller alertmanager=example-coo-monitoring-stack namespace=ns1-coo secret=alertmanager-example-coo-monitoring-stack
The secret that holds the alertmanager configuration is called "alertmanager-example-coo-monitoring-stack-generated"
For a workaround:
- Create a secret manually named as per the error in the alertmanager pod "alertmanager-example-coo-monitoring-stack" with the required alertmanager configuration.
- It resolves the error and reflect the correct config in the pod:
oc exec -n <namespace-name> <alertmanager-pod-name> -- amtool config show --alertmanager.url http://localhost:9093
- NOTE: The workaround fixes the issue as manually created secret reflects the required alertmanagerconfig in the pod but it will not create the alertmanagerconfig CRD instance.