Nutanix Cloud Platform Fundamentals
This domain tests the product map, the deployment models, and the licensing boundaries that decide which capabilities a customer actually has. Questions usually ask which component owns a job and what a given form factor can do.
The stack
- AOS - the distributed platform: storage, cluster services, and Prism Element, running on every node.
- AHV - the native KVM-based hypervisor. ESXi and Hyper-V can run on the same AOS, which is why AOS and AHV must be reasoned about separately.
- Prism Element - per-cluster UI and API. Prism Central - multi-cluster management, categories, projects, IAM, and the launch point for many services.
- Services - Files (SMB/NFS), Volumes (iSCSI block), Objects (S3-compatible), NDB (databases), NKP/Karbon (Kubernetes), Calm (self-service/IaC), Flow (security), Leap (DR orchestration), Move (migration), Foundation (imaging), LCM (upgrades), Insights and Data Lens (operations and analytics).
Deployment and licensing
Form factors include appliance clusters (NX, NXi), software-only on validated hardware, single-node and small ROBO clusters at the edge, and NC2 clusters running in public clouds such as AWS and Azure. AOS is licensed per core under feature tiers (for example NCI Pro and NCI Ultimate); add-on products and Prism Central-centric services are licensed separately, so a feature visible in the UI is not proof of entitlement.
Worked example
A 5-node cluster with 8 TB drives holds 40 TB raw. RF2 yields roughly 20 TB usable before Erasure Coding. If 12 TB is genuinely stored, is a full rebuild after one node loss possible? Remaining raw is 32 TB; 12 TB at RF2 needs 24 TB, and 24 < 32, so yes - the cluster can re-protect.
Traps
Assuming RF2 always means exactly 50% usable; treating AOS and AHV as one product; expecting Prism Element to manage multiple clusters; assuming every UI feature is licensed.