Troubleshooting upgrade issues

Unified Compute Platform (UCP) Advisor Software Installation Guide

Part Number

Use the following guidelines and suggested steps to help resolve UCP Advisor upgrade issues.

Upgrading UCP Advisor to the current version fails.
What it Means:
UCP Advisor could not be upgraded to the current version.
Corrective Action:
  1. Log on to the vCenter and revert the snapshots of all the UCP Advisor cluster nodes.
  2. (Optional) Update the command devices as required.
  3. Power on the Master Node VM and the Worker Node VMs.
  4. Verify that the Kubernetes services are running. Enter:
    kubectl get pods -n ucp
    Note: If any pod is in the Warning, Error, or init(0) state, then reset the pod. Enter:
    kubectl rollout restart deployment ucpadvisor-kong -n ucp
  5. Verify that the UCP system devices are reconnected and that they are in the Ready state. Enter:
    kubectl get nodes -n ucp
  6. Download the UCP Advisor 4.5.0 upgrade package ucpadvisor-air-gapped-v4.5.0-88.tar.gz from the Hitachi Support Connect Portal at: (login required), and then copy it to the UCP Advisor Master Node VM.
  7. Extract the UCP Advisor upgrade file. Enter:
    sudo tar -xvzf ucpadvisor-air-gapped-v4.5.0-88.tar.gz
  8. Navigate to the installer/scripts/ folder, and then run the following commands:
    chmod 755
  9. Log on to UCP Advisor and verify all the UCP systems and devices.
  10. Clear the host key configurations for the network switches.
    • For Ethernet switches: Enter
      clear ssh hosts
    • For Fibre Channel switches: Enter
      sshutil delknownhost <IP address of the Master Node VM>