Use the following guidelines and suggested steps to help resolve UCP Advisor upgrade issues.
- Condition 1:
- UCP Advisor upgrade did not complete.
- What it Means:
- UCP Advisor upgrade fails with the following error:
Root directory free space of node <Node IP Address> does not meet the specifications for the selected deployment mode.
- Corrective Action:
-
- Log on to the UCP Advisor VM as a root user.
- Run the following command:
crictl rmi --prune
- Retry upgrading UCP Advisor.
- Condition 2:
- After upgrading UCP Advisor or the remote gateway to 4.7, the /home/ucpadmin folder may get deleted.
- Corrective Action:
-
Perform a soft reboot on the UCP Advisor VM or the remote gateway. Run the following command:
sudo reboot
- Condition 3:
- Upgrading UCP Advisor from 4.6 to 4.7, and then from 4.7 to 4.8 fails.
- Corrective Action:
-
- Revert to the UCP Advisor 4.7 snapshot.
- Using SSH, log on to the UCP Advisor VM as a ucpadmin user.
- Run the following commands:
MASTER_NODE=(kubectl get nodes -l node-role.kubernetes.io/master -o jsonpath='{.items[0].metadata.name}') sudo hostnamectl set-hostname "$MASTER_NODE" reboot - Wait for 10 minutes for the UCP Advisor VM to reboot and then retry upgrading UCP Advisor.