A greenfield VMware Cloud Foundation (VCF) 9.1 deployment completes successfully, but the 3-node VCF Operations instance is completely inaccessible when using the Virtual IP (VIP) Fully Qualified Domain Name (FQDN) provided during the deployment workflow. This failure prevents you from connecting to or managing the cluster through a unified management endpoint.
The VCF Operations deployment finishes without returning workflow errors.
Individual primary, replica, and data nodes are fully functional and accessible via their direct IPs or FQDNs.
Network traffic directed specifically to the cluster VIP FQDN fails to resolve or connect.
No native load balancer or virtual IP is automatically provisioned during the VCF Operations 9.1 setup.
VCF Operations 9.1.x
VCF Operations 9.x does not feature a native integrated load balancer or an out-of-the-box VIP. The multi-node High Availability model requires an external HTTP load balancer to route and distribute stateless API and UI management traffic across the cluster members. This issue occurs because a VIP FQDN was defined during deployment without configuring an external load balancing solution (such as NSX Advanced Load Balancer) to front the cluster.
You must deploy and configure an external load balancing solution manually:
Deploy an external load balancer, such as the NSX Advanced Load Balancer (Avi Vantage), within your management environment.
Configure the external load balancer backend pool to include all nodes of the VCF Operations cluster (the primary, replica, and data nodes).
Set up the virtual service on the load balancer using the exact VIP FQDN and IP address designated during the VCF deployment workflow.
Verify that the load balancer correctly distributes HTTP/HTTPS traffic across the cluster nodes and that the VIP FQDN becomes accessible.
This is further detailed in the Load Balancer Models documentation