-
Story
-
Resolution: Done
-
Major
-
None
-
None
-
3
-
False
-
False
-
NEW
-
NEW
-
-
Logging (Core) - Sprint 211, Logging (Core) - Sprint 212, Logging (Core) - Sprint 213
Story
As a member of the OpenShift Logging Team,
I want telemetry metrics of the CLO
so that we have a better understanding of how CLF is being utilized by customers (e.g. which outputs)
Acceptance Criteria
- Verifying telemetry metrics are available from the CLO metrics endpoint
- PR against cluster-monitoring-operator to whitelist metrics
Logging telemetry metrics are visible in telemetry console.
Notes
- Push telemetry metrics to Red Hat as described by: https://docs.google.com/document/d/1s7odpnziZSaAsnFFHgsZs12dFzgXcy9eNJOSO4l1qCI
The metrics may undergo some changes if issues are discovered during implementation.
-------------------------------------------------------
TASKS BREAKUP
Breaking up into smaller Tasks/Milestones by following the Process for enabling sending metrics via telemetry as stated by https://docs.google.com/document/d/1a6n5iBGM2QaIQRg9Lw4-Npj6QY9--Hpx3XYut-BrUSY/edit#heading=h.b45lllaborcs
Step 1
- Raise Consultancy Requests to Monitoring team for In-depth questions/consultations. Architectural review. Troubleshooting. Investigations. Internal enablement. Documentation. Pull-Request reviews
Step 2
- Get approval from Clayton Coleman to verify that you are within the metrics budget and to ensure that the new metrics do not have unbound cardinality
Step 3 ensuring underlying metrics can be scrapped by prometheus via service monitor, follow the below steps
- For 4.4.x and greater open PR in the cluster-monitoring-operator to allowlist the metrics (4.1.z, 4.2.z, 4.3.z see the “Process for backporting telemetry metrics”)
- Modify manifests/0000_50_cluster_monitoring_operator_04-config.yaml.
- Execute `make docs`
- Open a PR against cluster-monitoring-operator with the above changes.
- At the next deployment of the telemetry backend, the newly added metrics will be allow-listed.
Step 4
- In order to allowlist those metrics also on Red Hat premises (server-side) for Telemeter, bump the rhobs configuration repository (where the Telemeter configuration lives):
- In the rhobs/configuration repo, regenerate the manifests by running `make whitelisted_metrics && make` and commit the changes.
- See this PR as an example. For any questions ask on #forum-observatorium on coreos slack channel.