Glossary

This glossary defines canonical platform-wide terms used across documentation. It focuses on concepts that appear in multiple sections of the product. Terms that apply only to a single workflow or subsystem should remain documented in their local pages.

Platform and Cluster Terms

TermDefinitionRelated doc
Global ClusterThe centralized management and control hub of . In the platform's hub-and-spoke architecture, it provides platform-wide services such as authentication, policy management, cluster lifecycle operations, and observability.Architecture
Workload ClusterA Kubernetes-based environment managed by the global cluster. A workload cluster runs isolated application workloads and inherits governance and configuration from the central control plane.Architecture
Platform-Provisioned InfrastructureA cluster management model in which the platform provisions both machines and node operating systems, and manages the full cluster lifecycle. In this model, all nodes use an immutable operating system.Clusters Overview
User-Provisioned InfrastructureA cluster management model in which users provide pre-provisioned physical or virtual machines. The platform manages Kubernetes on those nodes, while node operating system management remains under user control.Clusters Overview
Hosted Control Plane (HCP)A deployment model in which each cluster has its own dedicated control plane, while multiple control planes are hosted as workloads on a dedicated management cluster. This model separates the control plane from worker nodes to reduce resource consumption and improve multi-cluster scalability.About Hosted Control Plane
Managed ClusterAn existing cluster brought under the platform for centralized governance and operations. In ACP, managed clusters include existing standard Kubernetes clusters and selected public cloud clusters that are onboarded through import or registration workflows.Managed Clusters Overview
Immutable OSAn immutable operating system used for platform-managed nodes in platform-provisioned environments. Node state is kept consistent and recoverable by treating the operating system layer as read-only and centrally managed.Clusters Overview
Immutable InfrastructureA cluster provisioning and operating model in which node configurations are baked into images and remain unchanged after deployment. Cluster upgrades and configuration changes are applied by replacing nodes with new images.About Immutable Infrastructure
ProjectA platform governance unit that isolates resources and personnel for a tenant or team. A project can span multiple associated clusters and acts as the management boundary for quotas, policies, and namespace ownership.Create Project
NamespaceA Kubernetes namespace managed directly or indirectly by the platform. In ACP, a namespace can be created within or imported into a project so that it inherits project-level governance and visibility.Importing Namespaces
Control PlaneThe Kubernetes management layer that runs core cluster components such as the API server, scheduler, and controller manager.Architecture
Control Plane NodeA node that runs Kubernetes control plane components used for cluster management. Use this term instead of outdated alternatives such as "master node".Architecture
Worker NodeA node that runs application workloads and supporting platform components. Use this term instead of outdated alternatives such as "slave node".Architecture

Identity and Access Terms

TermDefinitionRelated doc
Identity Provider (IdP)An external identity system that authenticates users for the platform, such as LDAP, Active Directory, or an OpenID Connect provider.Accessing the Web Console
OpenID Connect (OIDC)An identity layer built on OAuth 2.0 that ACP uses in several authentication and authorization scenarios.Managing OIDC

Extension and Packaging Terms

TermDefinitionRelated doc
OperatorAn extension mechanism built on Kubernetes custom resources and controllers that automates lifecycle management for complex applications or services. In , Operators are managed through Operator Lifecycle Manager.Operator
Operator Lifecycle Manager (OLM)The operator management framework that handles Operator installation, upgrades, channel subscriptions, dependency resolution, and related custom resources such as CatalogSource, Subscription, and InstallPlan.Operator
OperatorHubThe platform interface for discovering, installing, upgrading, and managing Operators through OLM.Operator
Cluster PluginThe platform's extension mechanism for chart-based plugins. Cluster plugins are managed through the ModulePlugin, ModuleConfig, and ModuleInfo custom resources.Cluster Plugin

Networking and Access Terms

TermDefinitionRelated doc
IngressA Kubernetes resource that exposes HTTP and HTTPS routes from outside the cluster to internal services. ACP uses Ingress as one of its main north-south traffic entry models.Configure Ingresses
Gateway APIThe Kubernetes networking API family that defines role-oriented resources for advanced L4 and L7 routing. In ACP, Gateway API is positioned as a next-generation traffic management model alongside Service and Ingress.Networking Overview
ServiceIn Kubernetes, a Service is a method for exposing a network application that runs as one or more Pods in a cluster. In ACP, Service is a core service-discovery and traffic-exposure primitive, including ClusterIP, NodePort, and LoadBalancer types.Configure Services
LoadBalancerA Service type that exposes a Service through an external load balancer. This usually requires either a cloud-provider integration or a separately provided load-balancing component.Configure Services
Platform Access AddressThe external address used to access platform services such as the web console and platform APIs. It can be the same as the Cluster Endpoint or a separate address for external access scenarios.Install
Cluster EndpointThe address used by cluster components and administrators to reach the target cluster control plane endpoint. It is the primary control-plane access entry during installation and later operations.Install
Self-built VIPThe built-in virtual IP option used when an external load balancer is not provided for the Cluster Endpoint.Install

Disaster Recovery and Upgrade Terms

TermDefinitionRelated doc
Global Cluster Disaster RecoveryThe disaster recovery model for the global cluster in which a primary global cluster and a standby global cluster are kept ready for failover through etcd data synchronization and coordinated operational procedures.Global Cluster Disaster Recovery
Cluster Version Operator (CVO)The operator-based upgrade workflow and controller used to coordinate target version, preflight status, and execution progress for global and workload cluster upgrades.Upgrade Overview

Usage Notes

  • Use this page as the canonical source for ACP-wide terms that appear across multiple documentation sections.
  • Keep page-local ## Terminology sections for workflow-specific or subsystem-specific terms that are not reused broadly across the product.
  • The Term column uses a normalized display style for readability.
  • Keep official feature names, protocol names, UI labels, and API-facing names in their official capitalization, such as OperatorHub, Platform Access Address, ClusterIP, Self-built VIP, and OpenID Connect (OIDC).
  • Favor product concepts, platform models, and high-value cross-section entry terms over generic engineering vocabulary.
  • Expand an acronym on first mention when needed, then use the acronym consistently.
  • When a term is already defined by Kubernetes or OpenShift, use the upstream meaning first and add ACP-specific context only when needed.