Epic Goal

Continue to refine the pattern we can re-use for managing resources in the fleet of clusters.  Some examples could be Hosted Control Planes on Multi-Cluster Engine clusters.  Another is Virtual Machines on a ManagedCluster.

Link to board 

Why is this important?

The pattern replays for all of our core competencies. It is also the pattern for fleet management of kubernetes resources.

Scenarios

1. Can I discover a resource kind and its depended resources. Example: VirtualMachine
2. Can I list that resource kind across the fleet
3. Can I take a specific action on that resource
4. Can I take a bulk action on many similar resources

Acceptance Criteria

• Documentation of the pattern
• [SPIKE] Implementation of the pattern for a resource type (already in use for Virtual Machines)

Dependencies (internal and external)

1. Discovery = Search
2. Action = kubernetes api via reverse proxy, ManifestWork, and Maestro

Previous Work (Optional):

1. Search discovery for Applicaitons
2. Maestro
3. ManifestWork, WorkAction, ManifestWorkReplicaSet
4. logs and YAML via kubernetes api via reverse proxy

Open questions:

1. Should we pick one action type?

Done Checklist

• Pattern is agreed upon
• Pattern is documented
•  [SPIKE]Examples are available