NDB HA is a feature that provides high availability and resilience for the NDB service and its components. NDB HA creates a cluster of API server VMs, repository VMs, and HAProxy VMs across multiple Nutanix clusters. Each component has a leader and a set of followers that are synchronized and can take over the leader role in case of a failure. NDB HA also monitors the health and status of the NDB components and performs automatic failover and recovery actions when possible. However, some failure scenarios may require manual intervention from the administrator to restore the NDB service. One such scenario is when more than one API server VM is down. The API server VMs are responsible for handling the requests from the NDB UI and CLI, as well as communicating with the repository VMs and the HAProxy VMs. If more than one API server VM is down, the NDB service may become unavailable or unstable, and the administrator will need to manually restart or redeploy the affected API server VMs. The other options do not require manual interaction for the recovery, as NDB HA can handle them automatically. If one repository VM is down, NDB HA will fail over to another repository VM that has the latest data. If more than one HAProxy VM is down, NDB HA will route the traffic to the remaining HAProxy VMs. If the NDB server process is down, NDB HA will restart the process or fail over to another API server VM. References:
Nutanix Certified Professional - Database Automation (NCP-DB) v6.5, Section 2 - Deploy and Configure an NDB Solution, Objective 2.3: Configure NDB High Availability
Nutanix Database Management & Automation (NDMA) Course, Module 3: Nutanix Database Service (NDB) Installation and Configuration, Lesson 3.3: Configuring NDB High Availability, Topic: NDB HA Architecture and Components
Nutanix Support & Insights, RA-2035: Nutanix Database Service High Availability, Section: NDB HA Failure Scenarios and Recovery Actions
