As a Logging administrator I want to manage the Logging view plugin in OpenShift Console by managing the ObservabitlityUIPlugin resource, so that I can manage Logs Console independently from cluster-logging-operator releases.

Acceptance Criteria

1. Given as a Logging administrator I create a LoggingStack resource then I want the Cluster-Observability-Operator to install an ObservabilityUIPlugin resource that configures the Logs Console.
2. Given as a Logging administrator I disable the Logs console in the LoggingStack resource then I want the Cluster-Observability-Operator to omit creating an ObservabilityUIPlugin resource.
3. Given the Cluster-Observability-Operator reconciles an ObservabilityUIPlugin resource then I want Cluster-Observability-Operator to create and manage the logging-view-plugin deployment based on the OCP version and Logging channel hosting the Logging stack.

Developer Notes

• The original ocpConsole spec in CLO used to have a couple of extra fields to address logging-view-plugin related configuration. The following fields need to added to the COO UIPlugin spec for type logging:
  • NodeSelectors and Tolerations (https://github.com/openshift/cluster-logging-operator/blob/master/api/logging/v1/clusterlogging_types.go#L102-L133)
  • ocpConsole spec: logsLimit and timeout (See https://github.com/openshift/cluster-logging-operator/blob/master/api/logging/v1/clusterlogging_types.go#L166-L179)
  • Move selection of the LokiStack resource from logStore spec into UIPlugin spec.
• The CLO console plugin reconciler implements a compatibility matrix check for OCP version that needs to be transferred to COO UIPlugin reconciliation too (See https://github.com/openshift/cluster-logging-operator/blob/master/internal/visualization/console/feature_map.go#L21-L36)