VIO Duplicate IP allocated to new worker nodes after deleting some worker nodes

Products

• VMware
• VMware Integrated OpenStack

Issue/Introduction

Symptoms:

• Scale down worker nodes and then scale back up
• New worker nodes will have IPs of the existing worker nodes.

Environment

VMware Integrated OpenStack 6.x
VMware Integrated Openstack 7.x

Resolution

• This is a know issue affecting VMware Integrated Opentack 6.x.
• This is a know issue affecting VMware Integrated Opentack 7.x.

Workaround:
Note:  This will need to be done for each controller with the duplicate IP

1. Find controllers with duplicate IP's

#viocli get deployment

root@oms [ ~ ]# viocli get deployment
PUBLIC VIP    PRIVATE VIP   HIGH AVAILABILITY
172.22.0.70   172.20.1.50   Enabled
NODE NAME               ROLE         VALIDATION   STATUS    IP
controller-4szrk8glhl   Controller   Success      Running   172.20.1.2
controller-7d522qjbk8   Controller   Success      Running   172.20.1.3
controller-msbtr5lwmr   Controller   Success      Running   172.20.1.1
controller-tnmm4x54t2  Controller                           172.20.1.1
oms.vio.local           Manager      Success      Running   172.20.0.45

2. kubectl cordon controller-tnmm4x54t2
3. kubectl drain controller-tnmm4x54t2 --ignore-daemonsets --delete-local-data
4. Go to vCenter client, check controller-tnmm4x54t2 VM, make sure it only has one disk
5. osctl edit machine controller-tnmm4x54t2
6. Change the IP to the next free IP

ie change 172.20.1.1 to  172.20.1.4

6. kubectl delete node controller-tnmm4x54t2
7. Go to vCenter client, right click controller-tnmm4x54t2 VM and select Power > Power Off
8. After the VM is powered off, right click the VM and select Remove from Disk

Note:  vio-manager will try to power on the VM automatically. Will need to delete the VM quickly after powered off

Now cluster-api should automatically detect the VM is missing and recreate one with the new IP assigned