metrics for workloads pods and http requests missing

[csantanapr]

metrics for workloads pods and http requests missing

In what area(s)?

/area monitoring
/kind bug

What version of Knative?

0.6.0

Expected Behavior

When installing monitoring.yaml from the 0.6.0 release I should be able to open grafana from the knative-monitoring namespace and the dashboards should have data.

Actual Behavior

Both bashboards are empty and the drop downs don't populate with configurations or revisions instances

Steps to Reproduce the Problem

I tried in 2 environments:

1. IKS (IBM public Cloud)

2. Local using kind

3. Install knative serving including monitoring.yaml from the release page

4. Port forward grafana from knative-monitoring

kubectl port-forward --namespace knative-monitoring \
$(kubectl get pods --namespace knative-monitoring \
--selector=app=grafana --output=jsonpath="{.items..metadata.name}") \
3000

3. Open dashboards

• "Knative Serving - Scaling Debugging"
• "Knative Serving - Revision HTTP Requests"

4. Results: No data found
   

Other dashboards have data about knative control plane.

[JRBANCEL]

"Knative Serving - Scaling Debugging" has been fixed: #4096

[dgerd]

@csantanapr Can you confirm @JRBANCEL's change fixes this issue?

@JRBANCEL Is this something that should be cherry-picked into the Serving 0.6 release?

[csantanapr]

Pod count dashboard is fix via the change mentioned.

The http request dashboard still missing data.
Does this dashboard depend on having istio sidecars injected into ksvc?
I assumed that the metrics came from Knative side card and no istio side card was required.

[dgerd]

@csantanapr It used Istio metrics up through Knative 0.6. This has been changed in 0.7 to enable metrics without sidecar.

See 4e5cf96

[csantanapr]

I will try with 0.7 and report back

[knative-housekeeping-robot]

Issues go stale after 90 days of inactivity.
Mark the issue as fresh by adding the comment /remove-lifecycle stale.
Stale issues rot after an additional 30 days of inactivity and eventually close.
If this issue is safe to close now please do so by adding the comment /close.

Send feedback to Knative Productivity Slack channel or file an issue in knative/test-infra.

/lifecycle stale

[dgerd]

This issue was originally for Knative 0.6. It looks like we may have similar issues in HEAD, but that is being tracked in #6255.

Resolving as duplicate.