KubeVirt often spans more than one cluster
Production VM platforms rarely stay in a single homogeneous cluster. Teams separate environments, hardware pools, regions, or tenants across Kubernetes clusters, then need one operating model for requests, ownership, and review.
Cluster-aware self-service
Shepherd lets platform teams decide which clusters, namespaces, templates, and instance sizes are visible for a given service or environment.
- Expose multiple KubeVirt clusters through one management surface.
- Keep production and non-production access scoped separately.
- Route user requests through policy and approval before delivery.
Governance without broad cluster admin access
A multi-cluster KubeVirt platform should not require every application team to hold direct administrative access to each cluster. Shepherd provides a higher-level service interface while preserving infrastructure-level control for platform operators.