VMware Telco Cloud Automation 3.1.0 introduces compatibility with TKG version 2.5.0, which includes a new diagnostic plugin for gathering crashd logs from management clusters. This plugin is in its initial release and does not yet support crashed log collection for v2 workload clusters. During TCA installation, the diagnostic plugin is not installed by default and is only added when creating a management cluster with TKG version 2.5.0 (using Kubernetes version 1.28.4) or upgrading an existing management cluster to TKG 2.5.0.

Follow the steps below to utilize this plugin for collecting the management cluster's crash log bundle.

Procedure

  1. Login to VMware Telco Cloud Automation platform with SSO credentials.
  2. Navigate to Virtual Infrastructure.
  3. Click on the menu option on the left of TCA-CP on which the management cluster is deployed and select Open Cluster Terminal.
  4. On TCA-CP terminal, use the command below to query for kbs-tkg250 pod where the diagnostic plugin is installed. First column value is the namespace in which kbs pod is running and second column below is the pod name.
    getPod 
    debug-pod [ ~ ]$ kubectl get po -A | grep kbs-tkg250
tca-cp-cn                  kbs-tkg250-6bf6599b8f-s6zd8                         1/1     Running     0              2d
  5. Exec to kbs-tkg250 pod using the following command:
    execPod 
    kubectl exec -it -n namespace                  pod_name -- bash
  6. Use the diagnostic plugin guide, Troubleshooting Clusters with the Diagnostics Plugin, to collect management cluster log bundle once inside the KBS pod.
  7. In certain cases, the KBS pod may not have the diagnostic plugin installed. In such a scenario, use the following command before proceeding with collection of management cluster logs. This command must be executed from within the KBS pod.
    installTanzuDiagnostic 
    root [ / ]# tanzu plugin install diagnostics