In a multi-cluster NDB instance with HA enabled, the NDB architecture includes multiple API server VMs to ensure availability. If more than one API server VM is down after a failure, the administrator should ensure at least one API server VM is online. NDB’s HA design tolerates multiple API server failures as long as one remains operational, and the remaining VMs will recover automatically through the HA failover mechanism. This leverages Nutanix’s distributed architecture and self-healing capabilities to restore full functionality without manual intervention beyond ensuring one VM is running.

Other options are incorrect:

A. Restart all database server VMs as NDB HA only tolerates single API server failure: NDB HA can handle multiple API server failures, and restarting database VMs is unnecessary.

B. Manually distribute connections across remaining clusters: This is not required, as HA manages connections automatically.

C. Call Nutanix support to reset all API server connections: Support is not needed for this scenario; recovery is automated.

Thus, the verified answer is D, reflecting NDB’s HA resilience.

Official Nutanix Database Automation References

Nutanix Database Management & Automation (NDMA) course, Module 4: High Availability and Disaster Recovery, Lesson 4.3: Recovering from Failures.

Nutanix Certified Professional - Database Automation (NCP-DB) v7.5Knowledge Objectives, Section 4: Troubleshoot NDB Solutions, Objective 4.4: Recover NDB HA (applicable to v7.5).

Nutanix NDB Administration Guide: "High Availability Recovery Procedures" section.