Automatic Fencing of a Failed Primary Cell

If a new primary cell is promoted after a primary cell failure, VMware Cloud Director automatically fences out the old primary to prevent it from restarting.

In case of a failover, if a failed primary database restarts after a new primary cell is promoted, VMware Cloud Director automatically fences out the old primary. This automation prevents the split-brain syndrome where two active databases can diverge from each other. The fencing automation stops and deactivates the vpostgres service on the old primary node. After that, you can redeploy the failed primary as a standby cell to restore the cluster health to Healthy.

For more information about viewing the cluster health status and failover mode, see View the VMware Cloud Director Appliance Cluster Health and Failover Mode.