Prometheus tutorial

[a-thaler]

Description

Following the new Strategy, deprecation of the monitoring component should happen. In order to execute that, a replacement in the shape of a tutorial needs to be ready. So the goal is to have a tutorial on how to deploy a typical stack based on prometheus and grafana in a self-managed manner into a Kyma runtime, having a similar good integration experience as it is now. Hereby, it is essential to not copy anything from upstream to have a very low maintenance effort.

Criteria

• have a tutorial on how to deploy prometheus and co. via helm
• outline how to access the API but do not expose it, just describe port-forward
• tutorial artifacts should be usable in automated tests (own resources in a repository, not in-line markdown)
• do not copy the upstream chart anywhere
• add a license disclaimer on top
• brings k8s and istio setup
• supports custom metrics with mTLS
• enables generic scraping for workloads having prometheus annotation present

--- to be done

• At the moment the kubernetes/statefulset dashboard is missing, please figure out why
• have an alerting and custom metric example referenced, clear out confusion with current stack
  • remove the old tutorials from the public docu and just keep a redirect text
  • move the old monitoring-custom-alert/monitoring-custom-metrics examples as sub-instructions into the new tutorial and have it based on the new custom stack (using same placeholders)

--- will be done later

• Potentially bundle eventing and function dashboards

[chrkl]

The statefulset dashboard was removed from the upstream chart and the kubernetes-mixin project since it duplicates information of other dashboards:

• [kube-prometheus-stack] Delete statefulset Grafana dashboard prometheus-community/helm-charts#1892
• dashboards: remove statefulset dashboard kubernetes-monitoring/kubernetes-mixin#598

[chrkl]

@abbi-gaurav FYI

[chrkl]

The following changes have been performed:

• Making the monitoring-custom-alert and monitoring-custom-metrics examples subresources of the Prometheus example and aligned all commands and namespaces
• Removed monitoring tutorials from the Kyma docs and added links to the examples
• Enhanced the Prometheus example with instructions to set up alerting channels