-
Bug
-
Resolution: Obsolete
-
Critical
-
Logging 6.0.0
Description of problem:
when cluterlogging/<instance_name> managementState=false. It is the customer's response to migrate the Logging. CLO shouldn’t upgrade it
How reproducible:
Always
Steps to Reproduce:
1. Logging 5.9 forward logs to azureMonitor
cat <<EOF|oc create -f - apiVersion: "logging.openshift.io/v1" kind: "ClusterLogging" metadata: name: "clf-to-azure" spec: managementState: "Managed" collection: type: "vector" nodeSelector: node-role.kubernetes.io/worker: "" -- apiVersion: "logging.openshift.io/v1" kind: "ClusterLogForwarder" metadata: name: clf-to-azure namespace: project-to-azure spec: outputs: - name: azure-out1 type: azureMonitor azureMonitor: customerId: <spaceID> logType: ocp_logs1 azureResourceId: "<resourceID>" host: "ods.opinsights.azure.com" secret: name: azuremonitor-secret tuning: delivery: AtLeastOnce maxWrite: 10M pipelines: - name: pipe1 inputRefs: - application outputRefs: - azure-out1 serviceAccountName: logs-collector EOF
2. upgrade managementState to Unmanaged in ClusterLogging/clf-to-azure
3. upgrade CLO from 5.9 to 6.0
Actual results:
clusterlogforwarders.logging.openshift.io is upgraded to clusterlogforwarders.observability.openshift.io
Expected results:
clusterlogforwarders.logging.openshift.io isn’t migrated. user should migrate it manually