NC2 and Hybrid/Multicloud Operations
NC2 runs the Nutanix AHV stack on bare metal inside AWS or Azure, so the same Prism Central, APIs, and CLI manage cloud capacity. The exam tests where the boundary sits between Nutanix-managed and cloud-provider-managed, and how traffic returns on-prem.
What NC2 is and isn't
- NC2 is infrastructure: AHV clusters in a cloud VPC/VNet, managed through Prism Central. It is not a managed Kubernetes service; you still build clusters with NKP and CAPX.
- The cloud provider supplies bare metal, VPC networking, and identity; Nutanix supplies the hypervisor, storage fabric, and Prism.
- A Prism Central instance manages the NC2 cluster - dedicated to the cloud or reached over a private connection from on-prem.
Connectivity and multicloud
- AWS: VPC, subnets, VPN or Transit Gateway. Azure: VNet, VPN Gateway or ExpressRoute.
- Non-overlapping address space is mandatory - cluster and pod CIDRs must not collide with on-prem ranges crossing the tunnel.
- Flow Virtual Networking can build overlay VPCs inside NC2 for tenant isolation.
- Nutanix Central gives a single pane across on-prem, NC2, and other clouds.
Worked example
NKP on-prem must place a workload cluster into NC2. Reasoning: register the NC2 Prism Central as a CAPX endpoint, verify the AHV image exists in the target cluster, confirm the pod CIDR does not overlap the VPN range, then create the CAPI cluster. If creation hangs on infrastructure, it is almost always credentials, image, or IP pool.
Traps
Calling NC2 a Kubernetes service; assuming cloud-native load balancers exist inside NC2 (you still need MetalLB or an external LB); overlapping subnets across the VPN; planning a migration without a data-movement tool such as Nutanix Move.