Pipelines telemetry

Track OpenShift Pipelines operator metrics via Telemetry in order to increase the overall product quality.

Pipelines operator and controllers should expose the required metrics in order to be able to deduce the following

• Current status of the pipelines operator, similar to CVO operator status concention:
  https://godoc.org/github.com/openshift/api/config/v1#ClusterStatusConditionType

• Number of pipelineruns created per cluster
• Number of taskruns created per cluster
• Number of eventlistener calls from webhooks
• Number of taskruns created per clustertasks X (task name, count)
• Average TaskRun pod latency
• Number of taskruns that did not create a pod (failure)
• Number of eventlistener calls that did not create any pipelinerun or taskrun (failure)

Note:

This conflicts a little bit with SRVKP-443 or more accurately, SRVKP-443 has an impact on this.

Out of scope

Introduction of new metrics in Tekton is out of the scope for this epic. New epics should be created to capture the missing metrics

Estimation

M