Exam NCA-7.5 Topic 1 Question 2 Discussion

Pinned VMs and Replication Factor 1-enabled VMs with node affinity cannot be safely live migrated during host maintenance. Prism therefore identifies these VMs as non-migratable and requires them to be powered off before the node can complete entry into maintenance mode.
A pinned VM has a placement constraint binding it to a particular host. Moving it automatically would violate that constraint. An RF1-enabled VM has only one copy of its applicable vDisk data. When that data is associated with the node entering maintenance, another host cannot provide the same resilient storage access expected for an ordinary RF2 or RF3 VM. Powering off prevents the workload from continuing in an unsupported or inaccessible state.
Normal live-migratable VMs are moved to other nodes, subject to available CPU and memory capacity. HA- protected workloads can also be relocated according to the maintenance workflow. Pinned and RF1-affined VMs are handled separately, and Prism prompts the administrator to confirm their shutdown.
Before proceeding, the administrator should identify business impact, obtain an approved outage window, confirm cluster resiliency, and verify that affected applications can be restarted after the node exits maintenance mode.
NHCF/NCA reference: Node maintenance, VM migration restrictions, affinity, and RF1 behavior. Official references: Node Maintenance Mode and Putting a Host into Maintenance Mode .