In an NDB HA (High Availability) setup, when planning an electrical maintenance window requiring the clusters hosting NDB to be powered off, the administrator must follow a controlled shutdown sequence to preserve data integrity and minimize disruption. After shutting down the load balancer server (HAProxy), the next step is to shut down the API standby node. In an NDB HA configuration, the API leader node handles active operations, while the standby node provides redundancy. Shutting down the standby node first ensures the leader remains operational, allowing ongoing management tasks to continue until the final shutdown of the leader node.
Option A (In the sudo systemctl stop patroni command) is incorrect because Patroni is used for PostgreSQL HA, not NDB HA node management.
Option B (Shutdown the API standby node) is correct as it follows the recommended HA shutdown order, preserving active services.
Option C (Run the high-availability list command) is incorrect because this is a diagnostic command, not a shutdown step.
Option D (Shutdown the API leader node) is incorrect because the leader should be the last node shut down to maintain service availability.
This sequence ensures a graceful shutdown of the NDB HA environment.
References
Nutanix Database Service (NDB) User Guide, Chapter 3: Configuring an NDB Environment, Section: Managing HA Shutdown Procedures
Nutanix Support & Insights, Knowledge Base Article: "Shutting Down NDB HA During Maintenance"
Nutanix Certified Professional - Database Automation (NCP-DB) v7.5Blueprint, Section 3: Configure an NDB Environment
