AHV Virtualization and VM Lifecycle Management
What this domain tests
AHV lifecycle questions cover VM creation, cloning, snapshots, live migration, HA, maintenance mode, and guest integration.
Core rules and concepts
- AHV is KVM-based and managed through Prism. Acropolis is the management layer.
- A VM has vCPU, memory, disks, NICs, boot order, and categories. Disk buses include SCSI, IDE, and PCI; virtIO drivers improve performance.
- Nutanix Guest Tools (NGT) enables VSS-consistent snapshots, file-level restore, and self-service restore. Crash-consistent snapshots do not require NGT.
- Live migration moves a running VM between hosts with no downtime when NDFS-backed storage and compatible CPU features are present.
- HA restarts VMs on surviving hosts after a host failure. It requires cluster HA enabled and enough reserved capacity.
- Maintenance mode evacuates VMs before host service.
- Affinity and anti-affinity rules spread or co-locate VMs; Acropolis Dynamic Scheduling balances load.
Worked example
A VM on host A must be patched. Put host A into maintenance mode: AHV live-migrates the VM to host B, then patches. If host A fails instead and HA is enabled, the VM restarts elsewhere, but the guest sees a reboot. HA is not live migration. If the VM uses PCI passthrough for a GPU, live migration may be blocked.
Traps
Do not say HA provides zero-downtime failover. Do not require NGT for every snapshot. Do not confuse cloning with snapshots: clones create new VMs, snapshots are restore points. Do not assume live migration works across incompatible CPU generations without CPU masking.