NCSE-Core key terms

Nutanix software layer that pools local disks across cluster nodes into one shared datastore, using a Controller VM on each node to serve I/O and replicate data for resilience.

Per-node virtual machine that owns the node's local disks, runs the DSF services, and handles all storage I/O on behalf of guest VMs through a pass-through mechanism.

A grouping of physical disks on a node from which vDisks (extents) are carved; typically one pool per node, and the unit across which data is distributed by DSF.

The unit of data placement and replication in DSF: a set of extents (vDisk slices) written together and placed on multiple nodes according to the replication factor.

Number of copies of data DSF maintains, commonly 2 or 3; RF3 tolerates two simultaneous component failures, while RF2 tolerates one. Chosen per container.

A logical namespace within the distributed datastore that holds vDisks and carries policy settings such as replication factor, compression, deduplication, and erasure coding.

Space-efficient redundancy scheme that stores data plus parity fragments across nodes instead of full copies, reducing capacity overhead versus replication at the cost of higher rebuild work.

Background DSF service that scans metadata, rebalances data after failures or additions, and performs disk balancing and garbage collection to keep placement healthy.

CVM service that exposes the storage layer to hypervisors and handles read/write I/O routing, caching, and coordination with other nodes' Stargate instances.

Nutanix's native, KVM-based hypervisor included with the platform; managed through Prism, it hosts VMs and integrates directly with DSF without a separate hypervisor license.

The AHV management and orchestration layer that schedules VMs, manages clusters and images, and exposes APIs used by Prism and automation tools.

Rule that keeps specified VMs together on the same host or apart on different hosts, used for performance, licensing, or availability requirements within a cluster.

Cluster feature that detects host failure and restarts affected VMs on surviving nodes, subject to reserved capacity and configured policies such as host affinity.

Per-cluster management interface covering storage, VMs, hosts, alerts, and health; the local control plane for a single Nutanix cluster.

Multi-cluster management plane that aggregates registered Prism Elements, adds categories, policies, RBAC, and cross-cluster operations such as reporting and orchestration.

Key-value label applied to VMs, images, or clusters in Prism Central; used to scope policies, searches, and automation instead of naming individual objects.

Health-check framework that runs diagnostic checks against cluster components, scores findings by severity, and produces remediation guidance for support and operations.

Optional telemetry service that periodically sends anonymized health and configuration data from a cluster to Nutanix support for proactive analysis and recommendations.

Prism Central automation construct that triggers a sequence of actions on objects matching a category or condition, such as VM power operations or alerts.

Nutanix software-defined networking feature that provides overlay networks, distributed firewalling, and centralized policy management for VMs across clusters.

Virtual network built on top of the physical underlay, using encapsulation so VM subnets can span hosts and clusters independent of physical switch configuration.

Flow rule set that permits or denies traffic between VM groups defined by categories, applied at the virtual NIC level and enforced in a distributed manner.

Security approach that applies fine-grained firewall rules between individual workloads rather than at network perimeter, limiting lateral movement after a compromise.

Layer 2 network segmentation identifier used to isolate traffic on shared physical infrastructure; Nutanix hosts carry VM and storage traffic on tagged VLANs.

Aggregation of multiple physical NICs on a host into one logical interface for redundancy or throughput, configured with a supported mode such as active-backup or LACP.

Logical grouping of VMs and volume groups used for data protection; snapshots and replication are scheduled per protection domain rather than per individual VM.

Point-in-time, space-efficient copy of VM or volume data in DSF, used as a restore point or as the source for replication to a remote cluster.

Scheduled transfer of snapshot data from a source cluster to a remote cluster, meeting a recovery point objective rather than providing zero-data-loss failover.

Replication mode with very short recovery point objective for AHV workloads, continuously shipping changes so failover loses only seconds to minutes of data.

Maximum acceptable amount of data loss measured in time; determines how frequently snapshots or replication must occur to meet business continuity requirements.

Maximum acceptable time to restore a service after disruption; drives decisions about failover automation, standby capacity, and recovery runbooks.

Nutanix disaster recovery service that orchestrates replication, runbooks, and failover/failback across on-premises and cloud sites from a central console.

Nutanix migration tool that replicates and cuts over workloads from legacy hypervisors or public cloud into AHV with minimal downtime and scheduled cutover windows.

Service that runs the Nutanix stack natively on public cloud bare metal, enabling workload mobility and a consistent operating model between on-premises and cloud.

Adding nodes to a running cluster; DSF rebalances data to the new nodes while the cluster stays online, increasing capacity and performance.

Ability of the cluster to keep serving data despite component failure, determined by replication factor, data placement across nodes, and available spare capacity.

Process of restoring configured redundancy after a disk or node loss by recreating missing data copies from surviving replicas onto healthy components.

Nutanix sizing tool that models workload requirements, replication, and growth to recommend node models and cluster configurations for a proposed solution.