Nutanix NCA key terms

An IT architecture that delivers compute, storage, and networking as a single software-defined system on commodity x86 servers, managed as one pooled resource instead of separate infrastructure silos.

Nutanix's core software platform. It runs on every node and provides the distributed storage fabric, cluster services, hypervisor integration, and the APIs that Prism and AHV rely on.

A multi-cluster management appliance that gives one console across many Nutanix clusters and adds cross-cluster features such as categories, projects, role-based access control, and advanced analytics.

The web console and API bound to a single Nutanix cluster. It handles VM operations, storage configuration, health monitoring, and alerting for that cluster only.

Nutanix's native, kernel-based hypervisor, included with AOS at no additional licensing cost. It runs VMs, supports live migration, and integrates directly with the distributed storage fabric and Prism.

A service virtual machine running on every Nutanix node. It owns the node's local disks, serves storage I/O, and coordinates with controller VMs on other nodes to form the cluster.

The controller VM service that acts as the data I/O manager: it accepts read and write requests, decides where data is placed, and talks to the rest of the storage fabric.

A distributed NoSQL database that stores cluster metadata and configuration. Its data is replicated across controller VMs so the cluster keeps its identity when a node is lost.

A process on each controller VM that launches and monitors the other Nutanix services, handles cluster initialization, and reports service state to the cluster.

A background controller VM service that scans stored data, repairs inconsistencies, rebalances disks, and executes lifecycle decisions such as moving or deleting cold data.

Nutanix's software-defined storage layer. It pools local disks from all cluster nodes into one namespace and manages placement, replication, tiering, and data services transparently.

A logical grouping of physical disks drawn from cluster nodes. Storage containers are created inside a pool, which defines the fault-tolerance model for the capacity beneath them.

A logical partition inside a storage pool that is presented to VMs and hosts. Each container carries its own replication, compression, deduplication, and reservation policies.

The number of copies Nutanix keeps of each piece of data. A factor of two keeps two copies and three keeps three, directly setting how many failures the data survives.

A capacity-saving protection scheme that spreads data and parity fragments across nodes instead of keeping full replicas, cutting overhead while still tolerating disk or node loss.

The policy engine that watches how often data is accessed and moves cold data to slower, cheaper media while keeping frequently used data on fast tiers.

A read-only local copy of a remote vDisk, created on the node running the VM after repeated remote reads, so later reads are served locally.

A small fixed-size unit of stored data that the distributed storage fabric replicates and places independently, letting the system balance load and rebuild after failures.

A storage object that presents block devices to a VM or external host as virtual disks, commonly used when a guest needs direct iSCSI access to Nutanix storage.

A package installed inside guest VMs that lets Prism and the controller VM coordinate with the guest, enabling quiesced snapshots, self-service restore, and file-level recovery.

Key-value labels attached to VMs, hosts, or other objects in Prism Central. They drive policy assignment, search, reporting, and grouping across clusters.

A Prism Central construct that places VMs, users, and infrastructure inside an ownership boundary, making quotas and role-based access simpler to manage for individual teams.

A rule that controls which hosts a VM may run on, keeping chosen VMs together or apart to satisfy performance, licensing, or availability requirements.

The virtual switch AHV uses to connect VM network interfaces to host uplinks and VLANs, supporting bonding, tagging, and traffic shaping.

A logical slice of a physical network used to separate traffic. On Nutanix, VLAN IDs are applied to virtual switches and VM NICs through Prism.

Nutanix's distributed firewall. Policies are defined centrally in Prism Central and enforced at each VM's virtual NIC, isolating workloads without extra hardware.

A set of VMs, volume groups, and storage containers that are snapshotted and replicated together as one unit, giving related workloads a consistent recovery point.

A replication mode that streams changes continuously and reaches a recovery point objective measured in minutes rather than hours, sitting between snapshot scheduling and synchronous replication.

A stretched-cluster design in which two sites hold synchronously replicated copies of the same data, allowing workloads to keep running and fail over with almost no data loss.

Nutanix disaster recovery orchestration software. It uses protection policies and ordered runbooks to automate failover and failback of multi-tier applications between sites.

A collection of VMs and disks snapshotted at a single point in time so that a multi-VM application can recover in a consistent state.

A built-in diagnostic tool that runs a broad set of health checks on cluster services, configuration, and hardware, reporting problems before they turn into outages.

An automated telemetry service that periodically sends cluster health and configuration data to Nutanix support, enabling proactive diagnostics and recommendations.

A permission model in Prism Central that binds users or directory groups to roles, defining which actions they may take on which objects.

A security setting that restricts SSH and console access to the controller VMs, permitting connections only from approved addresses or through an authenticated access path.

Encryption of data written to cluster disks so that removed or stolen drives cannot be read. Keys are held by an external key management server or native key manager.

The Prism framework that inventories cluster software and firmware, checks compatibility, and applies updates in a validated order, including AOS and hypervisor upgrades.

A deployment tool used to image bare-metal nodes, set networking and cluster parameters, and either create a new Nutanix cluster or add nodes to an existing one.

A cluster setting that spreads data replicas across separate physical blocks, so the loss of an entire block does not take the data offline.