NSEA key terms

A design that combines compute, storage, and virtualization resources in the same nodes, managed as one system, instead of separate server, storage, and network tiers.

A set of nodes joined by shared network links that present pooled compute and storage as a single managed system with a common management interface.

A single physical server in a cluster, contributing CPU, memory, and local disks to the shared pool and running the platform software.

A virtual machine running on each node that serves storage I/O for all workloads, handles data locality and replication, and participates in cluster-wide operations.

The software layer spanning all nodes that aggregates local disks into one logical store, routes I/O to the right node, and enforces redundancy policies.

A grouping of physical disks on a node from which containers are carved; pools determine capacity, tiering, and failure-domain behavior for the data they hold.

A logical namespace presented to hosts and VMs that defines policies such as replication factor, compression, deduplication, and encryption for the data inside it.

The practice of keeping a VM's read traffic on the same node as its data when possible, reducing network hops and latency across the cluster.

The number of copies of each data extent kept in the cluster, typically two or three, so that one or more component failures do not cause data loss.

A scheme that stores data as fragments plus parity across nodes, giving redundancy with less capacity overhead than full copies, at some cost in write performance.

The ability of the cluster to keep serving workloads and preserve data when a disk, node, or other component fails, usually expressed as how many simultaneous failures are survivable.

The overall capacity of a cluster to absorb failures and rebuild lost redundancy automatically, determined by replication choices, capacity headroom, and node count.

The background process that recreates lost data copies on surviving nodes after a failure, consuming network and disk resources until redundancy is restored.

A logical grouping used in data protection features to apply snapshot and replication schedules consistently to the VMs and files assigned to it.

A point-in-time, space-efficient copy of a VM or volume that allows rollback or cloning without duplicating all data at creation time.

Copying protected data to a remote cluster on a schedule, so the remote site lags the source and may lose recent changes after a disaster.

A replication mode with a very short recovery point objective, typically seconds to minutes, that sits between scheduled async replication and fully synchronous mirroring.

The maximum acceptable amount of data, measured in time, that a workload can lose in a disaster; it drives the choice of replication method and schedule.

The target time within which a workload must be restored or failed over after an outage, shaping design choices such as standby clusters and automated failover.

The native hypervisor included with the platform, managed through the same interfaces as the cluster and supporting live migration and high availability without a separate license.

A software switch on each node that connects VM network interfaces to physical uplinks and enforces VLAN and bonding configuration.

Bonding multiple physical network interfaces into one logical link to increase throughput and provide redundancy if a cable or switch port fails.

A logical segmentation of a physical network that isolates traffic groups; on the platform it is used to separate management, storage, and workload traffic.

Moving a running VM from one node to another with little or no interruption to its workload, used for maintenance and load balancing.

Automatic restart of VMs on surviving nodes when their host fails, governed by configured policies and reservation capacity in the cluster.

The web-based management interface and API set used to configure, monitor, and operate the cluster, including alerts, tasks, and capacity views.

A separate management appliance that oversees multiple clusters, offering centralized inventory, categories, policies, and cross-cluster operations.

A key-value label applied to VMs and other objects so that policies, permissions, and reports can target groups without naming each object individually.

A diagnostic tool that runs a library of checks against the cluster and produces findings with severity and remediation guidance.

A notification generated when a monitored condition crosses a threshold or a fault occurs, carrying severity, affected entity, and suggested resolution.

The deployment tool used to image nodes, define network settings, and create the initial cluster before it is registered for management.

Adding nodes to an existing cluster to increase compute and storage capacity, after which data is rebalanced across the new hardware.

The one-time setup step that forms a new cluster from imaged nodes, elects cluster services, and makes the cluster reachable for management.

The process of calculating node count, CPU, memory, and usable capacity needed to run a workload set with headroom for growth and failure.

The storage space actually available to workloads after accounting for replication or erasure coding overhead, reserved space, and rebuild headroom.

A copy of a VM or disk created from a snapshot, often space-efficient, used to deploy many similar workloads quickly.