If a cluster node is no longer required, it should be decommissioned in a controlled manner to prevent disruption to any VODs hosted on that node.
Before You Begin
Before decommissioning a node:
Ensure all VODs hosted on the node have been deleted from the web interface.
Recreate any required VODs on alternative nodes.
Verify that no users or client systems are actively relying on services provided by the node.
Because VMS Gateway 9.2 does not support automatic workload migration, all VODs must be manually recreated on other nodes before the node is removed.
Step 1 - Prevent New Workloads Being Assigned
On the primary node, cordon the node to prevent any new workloads from being scheduled:
View all cluster nodes:
sudo /opt/vms-gateway/bin/kubectl get nodes
Cordon the node:
sudo /opt/vms-gateway/bin/kubectl cordon <node-name>
Confirm the node status:
sudo /opt/vms-gateway/bin/kubectl get nodes
Step 2 - Remove All VODs
Using the VMS Gateway web interface:
Navigate to the node being decommissioned.
Delete all VODs hosted on the node.
Recreate any required VODs on alternative nodes.
Verify that the node shows no remaining workloads before proceeding.
Step 3 - Uninstall VMS Gateway
Log onto the node being decommissioned and uninstall VMS Gateway using the appropriate uninstall procedure for the operating system.
Step 4 - Remove the Node from the Cluster
From the primary node, remove the node from Kubernetes:
View all cluster nodes:
sudo /opt/vms-gateway/bin/kubectl get nodes
Remove the node:
sudo /opt/vms-gateway/bin/kubectl delete node <node-name>
Step 5 - Verify Removal
Verify the node is no longer present in the cluster:
sudo /opt/vms-gateway/bin/kubectl get nodes
The node can now be removed from the VMS Gateway web interface.
Note: Decommissioning a worker node does not affect VODs running on other nodes. However, decommissioning the primary node is effectively a cluster rebuild operation and should only be performed as part of a planned migration or recovery procedure.