Nutanix NCP-CN key terms

The dedicated Kubernetes cluster that runs the Nutanix Kubernetes Platform control plane components, hosting the APIs and controllers used to create, upgrade, and delete workload clusters across providers.

A Kubernetes cluster provisioned and managed by the NKP management cluster, which runs the actual application workloads and receives lifecycle operations such as scaling, upgrades, and node pool changes.

A temporary Kubernetes cluster created during NKP installation to host the initial Cluster API controllers; once the management cluster is running, workloads are moved and the bootstrap cluster is deleted.

An upstream Kubernetes project that models clusters as custom resources managed by controllers, letting NKP declare desired cluster state and reconcile infrastructure such as virtual machines and networks.

A Cluster API resource that manages a set of machines for a node pool, maintaining the requested replica count and driving rolling replacement when the machine template changes during upgrades.

The Cluster API controller that owns control plane machines for a workload cluster, handling initialization, scaling to an odd number of nodes, and rolling upgrades of control plane components.

A named group of worker nodes within a workload cluster that shares a VM template, instance type, and labels, allowing independent scaling and heterogeneous hardware within one cluster.

The pluggable Cluster API component that creates the underlying machines and networking for a cluster, such as the Nutanix provider that provisions Prism Central VMs.

Nutanix's centralized management plane that exposes APIs for VMs, images, networks, and storage; NKP uses it as the infrastructure endpoint when provisioning clusters on Nutanix AHV.

Nutanix's native hypervisor, Acropolis Hypervisor, on which NKP nodes run as virtual machines; its networking and storage integrations determine how cluster traffic and volumes are presented.

A deployment method where NKP images, Helm charts, and container images are loaded from a local registry bundle because the environment has no internet access, requiring a private registry and mirrored artifacts.

A private container registry that stores copies of NKP and application images so clusters can pull them without reaching public registries, commonly configured through containerd registry configuration.

The Container Network Interface implementation that assigns pod IP addresses and programs routing and policy; NKP clusters typically use Calico or Cilium, and the choice affects network policy and load balancing features.

A Kubernetes Service type that requests an external virtual IP from an integrated load balancer, giving clients a stable address that forwards traffic to pod endpoints.

A bare-metal load balancer implementation for Kubernetes that assigns addresses from a configured IP pool to LoadBalancer Services and advertises them using ARP or BGP.

A routing configuration in which cluster nodes advertise service or pod networks to upstream routers, allowing external clients to reach LoadBalancer addresses without proxy or NAT devices.

A reverse proxy running in the cluster that routes HTTP and HTTPS traffic to Services based on host and path rules, and typically terminates TLS certificates for multiple applications.

A Kubernetes specification using GatewayClass, Gateway, and HTTPRoute resources to describe L4 and L7 routing, intended to replace Ingress annotations with role-oriented, portable configuration.

A Kubernetes resource selecting pods and declaring which ingress and egress traffic is allowed; enforcement depends on the CNI plugin supporting the policy API.

Nutanix's network security and microsegmentation feature that applies distributed firewall rules to VM traffic at the hypervisor level, complementing Kubernetes network policies with infrastructure-wide enforcement.

The Container Storage Interface driver that lets Kubernetes dynamically provision Nutanix Volumes or Nutanix Files storage and attach it to pods as PersistentVolumes.

A cluster-scoped object naming a provisioner and parameters used to create PersistentVolumes on demand, such as the Nutanix storage container, replication factor, and volume type.

A namespaced request for storage by a pod that specifies size, access mode, and optionally a StorageClass, and is bound to a PersistentVolume before the pod can start.

The field on a PersistentVolumeClaim declaring how many nodes may mount a volume, such as ReadWriteOnce, ReadOnlyMany, or ReadWriteMany, and constraining which storage backends qualify.

Block storage exposed to Kubernetes through the CSI driver, providing iSCSI-backed volumes that are typically attached to a single node and suited to databases and stateful single-writer workloads.

Nutanix's scale-out file storage service that can back Kubernetes persistent volumes requiring shared access, enabling ReadWriteMany claims for multiple pods across nodes.

A cluster-scoped resource that names the CSI snapshotter driver and deletion policy used when creating VolumeSnapshot objects for backup or cloning of persistent data.

Increasing the size of an existing PersistentVolumeClaim, which requires the StorageClass to allow expansion and the CSI driver to support online or offline resize of the backend volume.

A workload controller giving each pod a stable name and its own PersistentVolumeClaim, preserving identity and storage across restarts, which is essential for clustered databases on Kubernetes.

Nutanix's database lifecycle management platform that provisions, clones, patches, and protects databases such as PostgreSQL, MySQL, SQL Server, and Oracle across on-premises and cloud environments.

A registered virtual machine managed by Nutanix Database Service that hosts one or more database instances and provides the compute and storage footprint for provisioning and snapshots.

A reusable template in Nutanix Database Service that captures software version, compute size, storage, and network settings so database provisioning is standardized and repeatable.

A fast, space-efficient copy of a database created by Nutanix Database Service from a snapshot, commonly used to refresh test or development environments without impacting production.

The snapshot and log-capture policy in Nutanix Database Service that enables point-in-time recovery and cloning by retaining scheduled database snapshots on a defined schedule.

Restoring a database to a precise moment by applying archived transaction logs to a base snapshot, minimizing data loss after an accidental change or failure.

Nutanix Cloud Clusters, a service that extends the Nutanix software stack into public clouds such as AWS and Azure, running AHV and Prism Central on cloud infrastructure for hybrid operations.

An operating model spanning private and public infrastructure with consistent management, networking, and identity, allowing workloads to run where latency, cost, or compliance dictates.

A migration tool that moves workloads between platforms, including into NC2, by replicating data and reconfiguring networking so cutover downtime stays short.

The combination of metrics, logs, and traces collected from clusters, typically via Prometheus, Loki, and OpenTelemetry, used to detect failures and diagnose performance problems.

An open-source metrics system that scrapes time-series endpoints and evaluates alerting rules, widely deployed to monitor Kubernetes nodes, pods, and control plane components.

A quantitative measure of service behavior, such as request latency or error rate, used with a target threshold to define an objective and drive alerting decisions.

Kubernetes role-based access control, where Roles or ClusterRoles grant verbs on resources and are bound to users or groups through RoleBindings and ClusterRoleBindings.

A Kubernetes API server plugin that validates or mutates objects before persistence, used to enforce security policy such as image provenance, resource limits, or forbidden privileged pods.

Three built-in Kubernetes policy levels, privileged, baseline, and restricted, that constrain pod fields such as host namespaces, capabilities, and volume types when enforced by admission.

An open-source backup tool that captures Kubernetes resource manifests and, with volume snapshot integration, persistent data, enabling restore of namespaces or whole clusters.

A documented set of recovery objectives, runbooks, and tested procedures describing how applications and data are restored at a secondary site after a site-wide failure.

The maximum acceptable amount of data, measured in time, that may be lost after an outage, which determines snapshot frequency and replication interval for protected workloads.

The maximum acceptable duration to restore a service after disruption, driving choices about standby capacity, automation, and how quickly failover must complete.