-
Task
-
Resolution: Done
-
Undefined
-
None
-
None
-
None
-
False
-
-
False
-
-
When porting the swatch-metrics from a spring boot service to quarkus service, the metrics path changed from "/metrics" to "/q/metrics" and because of this, the metrics from this service are no longer visible in grafana/prometheus.
For other Quarkus services, we have a ServiceMonitor resource that configures the path to use for the Quarkus services depending on the presence of the label "prometheus: quarkus".
We tried to update the swatch metrics prometheus label from "rhsm" to "quarkus" here: https://github.com/RedHatInsights/rhsm-subscriptions/pull/2939, but we had to revert it because the deployment labels are inmutable (this might be an issue with the clowd deployments).
Therefore, we need to quickly fix this issue since at the moment this service are not contributing to exposing the metrics into grafana.
For the long term, this will be fixed at the clowder config source side: SWATCH-2091
Acceptance Criteria
- We should see the Swatch Metrics metrics into grafana. For example, the following metrics should be available:
metrics_collection_timer_seconds_count{productTag="rhel-for-x86-els-payg",status="success"} 1.0 metrics_collection_timer_seconds_sum{productTag="rhel-for-x86-els-payg",status="success"} 0.098850909 metrics_collection_timer_seconds_max{productTag="rhel-for-x86-els-payg",status="success"} 0.098850909
- is related to
-
RHCLOUD-30608 Support defining podSpec.metadata.labels in ClowdApp definitions
- New
- relates to
-
SWATCH-2091 Update Quarkus services to expose metrics in localhost:9000/metrics
- Closed