Add grafana kubebuilder plugin

[david-martin]

The changes in this PR are the result of running kubebuilder edit --plugins grafana.kubebuilder.io/v1-alpha

Verified locally via the kuadrant-operator:

• follow https://docs.kuadrant.io/0.7.0/getting-started-single-cluster/ to bring up a local cluster (using KUADRANT_REF=main
• create the dns operator servicemonitor (had to replace the namespace and matchLabel

kustomize build config/prometheus/ | yq e '
  .metadata.namespace = "kuadrant-system" |
  .spec.selector.matchLabels."control-plane" = "dns-operator-controller-manager"
' - | kubectl --context $KUBECTL_CONTEXT apply -f -

• port forward to grafana with kubectl --context $KUBECTL_CONTEXT -n monitoring port-forward svc/grafana 3000:3000
• import the 2 generated grafana dashboard json files (in this pr)

looks like this:

runtime metrics

resource metrics

[mikenairn]

Looks good to me.

Couple of questions:

1. Do we intend to define dashboards like this in each sub component and pull them into the kuadrant operator grafana setup?
2. Is there any way we can pull in grafana and prometheus setup from somewhere (kuadrant operator?) and optionally add it to the local-setup kustomization so we can test this locally without needing to set that up manually or use quickstart?
3. Any recommended docs on how to work with any of this?

[maleck13]

Looks good to me.

+1

Couple of questions:

1. Do we intend to define dashboards like this in each sub component and pull them into the kuadrant operator grafana setup?

I think we should have these in each kube-builder based component

2. Is there any way we can pull in grafana and prometheus setup from somewhere (kuadrant operator?) and optionally add it to the local-setup kustomization so we can test this locally without needing to set that up manually or use quickstart?
3. Any recommended docs on how to work with any of this?

[mikenairn]

3. Any recommended docs on how to work with any of this?

From the slack chat https://book.kubebuilder.io/plugins/grafana-v1-alpha.html?highlight=grafana#grafana-plugin-grafanav1-alpha

[david-martin]

• Is there any way we can pull in grafana and prometheus setup from somewhere (kuadrant operator?) and optionally add it to the local-setup kustomization so we can test this locally without needing to set that up manually or use quickstart?

Easiest method I've found is https://github.com/prometheus-operator/kube-prometheus/blob/main/kustomization.yaml

1 issue is the timing of CRDs being ready, and CRs being created.
Filtering to applying CRDs first, then the rest of the resources like this works well

[maleck13]

@david-martin one minor change I would suggest is the name of the service monitor. It is very generic right now and would prob clash with others. Ok to set is as part of this PR before merge? IE dns-operator-metrics-monitor ?

[maleck13]

Once this merges I want to pull it into the eng environment also

[david-martin]

name of the service monitor

@maleck13 changed.
Also changed the matchLabels selector to match the Service correctly.