Section 1 Recognize Nutanix Solutions and Tools

NCI (Nutanix Cloud Infrastructure) runs on AHV, Nutanix's own KVM-based hypervisor that is included at no extra cost, and on VMware ESXi. Check the Supported Hypervisors and Supportability of Features (by Hypervisor) pages for your AOS release, because some features exist only on AHV.

The core pieces: AOS (the storage and cluster OS), whose Distributed Storage Fabric (DSF) pools every node's local disks into one shared storage layer; AHV; Prism Element (manages one cluster, served by the CVMs); Prism Central (a separate VM that manages many clusters); and LCM for upgrades.

Every node runs a Controller VM (CVM). The CVM owns the node's disks and serves storage I/O. Its main services are Stargate (data I/O), Cassandra (distributed metadata), Curator (background cleanup and rebalancing), Zookeeper (cluster configuration) and Prism.

Storage is layered as storage pool → storage container → vDisks. Policies such as replication factor, compression, deduplication and erasure coding are set on the container, not on the VM or the pool.

Data is written RF2 (two copies) or RF3 (three copies) across different nodes. Data locality keeps a VM's data on the node it runs on, so reads stay local.

Common shapes: single-node (ROBO or edge, no node redundancy), two-node (ROBO; needs a Witness VM to avoid split-brain), three nodes or more (the standard FT1/RF2 cluster), and five or more for RF3/FT2. Clusters can also run in the public cloud as NC2 (Nutanix Cloud Clusters on AWS or Azure).

Match the product to the job. Flow Network Security handles VM microsegmentation by category. Nutanix Disaster Recovery handles protection policies, recovery plans and replication between sites. NC2 extends a cluster to public cloud. LCM handles firmware and software lifecycle. Storage-only nodes add capacity without adding hypervisor compute.

When a CVM's services fail (for example Stargate or Cassandra), the cluster redirects that node's I/O path to another CVM (autopathing). Guest VMs keep running while that happens.

Nutanix Unified Storage covers three services. Files gives SMB and NFS file shares. Objects is S3-compatible object storage that scales to petabytes of unstructured data. Volumes presents block storage over iSCSI to VMs or physical hosts, for example shared disks for a Microsoft SQL or Windows failover cluster.

Nutanix Database Service (NDB), the blueprint's "database automation", automates database lifecycle: provisioning, cloning, copy-data management, patching and time-machine style point-in-time recovery for SQL Server, Oracle, PostgreSQL, MySQL and others.

Nutanix Cloud Manager (NCM) is the management layer in Prism Central:
- Intelligent Operations (formerly Prism Pro): capacity runway, anomaly detection, behavioural learning and X-Play automation.
- Cost Management: costs for on-prem and public cloud.
- Security Central: compliance and security posture.
- Self-Service (formerly Calm): app blueprints and a Marketplace.

Nutanix Kubernetes Platform (NKP) deploys and manages Kubernetes clusters on Nutanix and in other clouds. Nutanix Data Services for Kubernetes (NDK) adds data services such as snapshots, replication and DR for stateful containers.

Nutanix Enterprise AI (NAI) is for running and managing large-language-model inference endpoints (model catalogue, endpoints, API keys) on your own infrastructure.

End-User Computing: VDI on Nutanix with Citrix or Omnissa Horizon brokers on AHV or ESXi, plus Nutanix's own desktop-as-a-service offering. Deduplication and data locality matter most for this workload.

Nutanix Move migrates VMs to Nutanix, mainly from ESXi or Hyper-V, and from AWS or Azure, onto AHV. It seeds the data, keeps it in sync incrementally, and then performs a short cutover. Move runs as an appliance VM.

Nutanix Guest Tools (NGT) is an in-guest agent. It enables application-consistent (VSS) snapshots, Self-Service Restore, VM mobility drivers and guest customization. It is installed from a mounted ISO, and NGT can be upgraded in bulk from Prism Central.

Foundation images nodes with a hypervisor and AOS and creates the cluster. It runs from a laptop, a Foundation VM, or the CVM on factory nodes. Foundation Central does this remotely and at scale: nodes call home to it, which allows zero-touch deployment to many sites.

The Witness VM sits at a third site. It arbitrates for two-node clusters and Metro Availability, so a failure doesn't cause a split-brain and the surviving node can keep serving.

Self-Service Restore (SSR) lets a guest user browse a VM's snapshots and restore their own files without the admin. It requires NGT in the guest.

X-Ray is a benchmarking and test-scenario tool. It runs repeatable workloads, plus failure and rolling-upgrade scenarios, to compare platforms during a proof of concept.

Nutanix Collector gathers configuration and utilisation data from an existing environment (vCenter, Hyper-V, Prism, AWS and others). That data feeds sizing for a Nutanix design.