Kubestellar

Kubestellar is an open-source project that provides multi-cluster workload placement and control for Kubernetes environments (multi-cluster management / edge and fleet orchestration).

• Decouples Kubernetes control planes from target clusters to support centralized workload placement and policy management (multi-cluster orchestration).
• Enables distribution of workloads and configurations from one or more “home” clusters to multiple “edge” or remote clusters (edge computing / fleet management).
• Implements declarative placement rules to control where applications, configurations, and resources should run across clusters (GitOps / policy-based management).
• Integrates with existing Kubernetes toolchains and APIs, operating as an overlay on top of standard clusters (Kubernetes platform integration).
• Targets use cases such as edge, telco, and geographically distributed deployments where clusters may be numerous, remote, or intermittently connected (distributed infrastructure management).

More About Kubestellar

Kubestellar addresses the problem of managing and operating Kubernetes workloads across large numbers of distributed clusters, with a focus on edge, telecom, and geographically dispersed environments (multi-cluster management). In these scenarios, organizations may run many small clusters at remote locations with intermittent connectivity, and need a way to control workloads and configurations from central sites without tightly coupling control planes to each remote environment.

The project introduces an architecture that separates the “home” clusters, where operators and automation live, from the “edge” or target clusters, where workloads actually run (edge and fleet orchestration). Kubestellar uses Kubernetes-native APIs and declarative resources to represent placements, policies, and desired state distributions. Operators define which applications, configuration objects, or policies should be present on which clusters, and Kubestellar propagates those resources according to that placement intent.

Core capabilities include workload and configuration distribution based on placement rules (policy-based management), synchronization of selected Kubernetes objects from home clusters to edge clusters (configuration management), and support for clusters that may be occasionally connected (edge operations). The project is designed to work with standard Kubernetes clusters without requiring intrusive changes, and it interoperates with common patterns such as GitOps repositories and Continuous Integration and Continuous Deployment (CI/CD) pipelines (platform integration).

In enterprise environments, Kubestellar is used to manage fleets of clusters across retail locations, branches, cell sites, or industrial facilities (fleet management). Central teams define policies and application manifests once, and use Kubestellar to place them on the appropriate subset of clusters based on labels, topology, or other placement criteria. This approach aims to reduce per-cluster manual configuration and allows organizations to keep central governance while respecting connectivity and autonomy constraints at the edge.

From a technical categorization perspective, Kubestellar fits into multi-cluster Kubernetes management, edge and distributed systems orchestration, and policy-driven configuration distribution. It focuses on overlay control and placement rather than replacing the underlying cluster lifecycle tooling, and it aligns with existing CNCF ecosystem components by leveraging standard Kubernetes concepts, APIs, and workflows.