NSX VTEP IP has been changed with creation of new Cluster and moved ESXi one by one from old cluster to new cluster with VMotioning VMs as per below steps. But we found significant performance degradation in VMs and intermittent connectivity issue and finally rolled back. Even after rolled back. Even after rolled back to old cluster within old vTEP network there was intermittent connectivity, that resolved to reboot all VMs.
Please let me know if someone face any situation like this or any steps to avoid intermittent connectivity issue. We need to perform this change again because current vTEP IP pool is full.
Is each VM required to reboot after vmotioning to new ESXi within new vTEP pool?
- Create new vCenter Cluster
- Add newly build ESXi in new cluster
- Prepare new cluster in NSX with new VTEP subnet
- Evacuate existing ESXi by migrating VMs from old cluster to new cluster one by one
- Move existing ESXi from old cluster to new cluster one by one, make sure old nsx VTEP is removed first.
- Login ESXi and vmkping for VTEP iP for other ESX VTEP IP
-