RevisionName templating and substitution.

[muscionig]

DRAFT up for discussion

Changes

This PR adds support for replacing the revisionName in the PrometheusMetric field. I discussed this with @skonto in the slack channel and it seemed a necessary feature to separate the scaling of different revisions. This is still in draft, it is missing unit tests and complete functional testing.

Minor Changes

• 🎁 adds a templating feature that allows to substitute the revisionName under the PrometheusQuery annotation. The feature uses template and sprig. I thought of using sprig to make sure the variable substitution is solid and to potentially support additional substitutions in the future.

Usage:

apiVersion: serving.knative.dev/v1
kind: Service
metadata:
  name: metrics-test
spec:
  template:
    metadata:
      labels:
        app: metrics-test
      annotations:
        ...
        autoscaling.knative.dev/metric: "http_requests_total"
        autoscaling.knative.dev/prometheus-query: "sum(rate(http_requests_total{pod=~\"{{ .revisionName }}.*\"}[1m]))"

I tested this with Istio

Output of the controller:

### service with no substitution
autoscaler-keda 2024/11/20 06:42:59 PrometheusQuery rendered: sum(rate(http_requests_total{}[1m]))                                    
autoscaler-keda {"severity":"INFO","timestamp":"2024-11-20T06:42:59.698756553Z","logger":"hpaautoscaler","caller":"controller/controller.go:548","message":"Reconcile succeeded","commit":"6f54184-dirty","knative.dev/controller":"knative.dev.autoscaler-keda.pkg.reconciler.autoscaling.hpa.Reconciler","knative.dev/kind":"autoscaling.internal.knative.dev.PodAutoscaler","knative.dev/traceid":"12a29044-6fe7-442b-a31c-7913633203cf","knative.dev/key":"metrics-test-istio/metrics-test-istio-00001","duration":"10.55125ms"}


### service with substitution
autoscaler-keda 2024/11/20 06:43:12 PrometheusQuery rendered: sum(rate(http_requests_total{pod=~"metrics-test-00001.*"}[1m]))
autoscaler-keda {"severity":"INFO","timestamp":"2024-11-20T06:43:12.122399087Z","logger":"hpaautoscaler","caller":"hpa/keda_hpa.go:98","message":"Creating Scaled Object \"metrics-test-00001\"","commit":"6f54184-dirty","knative.dev/controller":"knative.dev.autoscaler-keda.pkg.reconciler.autoscaling.hpa.Reconciler","knative.dev/kind":"autoscaling.internal.knative.dev.PodAutoscaler","knative.dev/traceid":"0fbaf41d-ecbf-4cf5-b417-64b2543e2e7f","knative.dev/key":"default/metrics-test-00001"}

Scaling of only one service:

/kind documentation
/kind enhancement

[knative-prow]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by: muscionig
Once this PR has been reviewed and has the lgtm label, please assign dprotaso for approval. For more information see the Kubernetes Code Review Process.

The full list of commands accepted by this bot can be found here.

Needs approval from an approver in each of these files:

• OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[knative-prow]

Hi @muscionig. Thanks for your PR.

I'm waiting for a knative-extensions member to verify that this patch is reasonable to test. If it is, they should reply with /ok-to-test on its own line. Until that is done, I will not automatically test new commits in this PR, but the usual testing commands by org members will still work. Regular contributors should join the org to skip this step.

Once the patch is verified, the new status will be reflected by the ok-to-test label.

I understand the commands that are listed here.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.

[skonto]

As a side note Serving supports byo revision names instead of being auto-generated.

kind: Service
metadata:
  name: hello
spec:
  template:
    metadata:
      name: hello-1

If you want to change that service above, then you need to manually change hello-1 to something else. PA name is always the same as with revision name.

[muscionig]

Thanks! I think it would be helpful to include this, along with the other points we've discussed, in a "Usage Recommendations" document. Should we log an issue to capture these notes? That way, we can reference it and update it with new considerations as they come up.

[skonto]

@muscionig let's proceed with this and add some tests to verify that it works as expected. Next we can add some docs.

[muscionig]

@skonto Thanks for the review! I've addressed the dependencies and added a test case. I have a couple of questions:

1. Would you add logic to validate user input? For example, if the query annotation contains anything other than revisionName (e.g., sum(rate(http_requests_total{pod=~\"{{ .wrongValue }}.*\"}[1m]))), the output becomes sum(rate(http_requests_total{pod=~"<no value>.*"}[1m])).
2. Should we gate this feature through the configMap? Personally, I believe it should always be enabled.