[gianlucam76]

Expose the health of the resources Sveltos deploys (profiles with spec.healthMonitoring.enabled) so the dashboard can show it per cluster and per profile.

ui-backend joins three things: the ClusterSummaries (which profile has monitoring enabled), the HealthChecks addon-controller generates for them (found by label and annotations, never by name) and the HealthCheckReports. The join is the shared addon-controller lib/healthmonitoring package, so the cluster list, the profile rows and the detail view cannot diverge.

Cache: a new healthCache in internal/server with its own lock, filled by two new controllers (HealthCheck, metadata only and filtered to generated checks; and HealthCheckReport) plus the existing ClusterSummary controller. Events can arrive in any order: each updates its own maps and recomputes only the cluster it touches. The report cache drops the collected resources (ResourceStatuses[].Resource), which are never read and are the only large field.

API:

• ManagedCluster.healthStatus on /capiclusters and /sveltosclusters.
• healthStatus on each /getClusterStatus row, for the profile.
• New GET /clusterHealth: the resources that are not Healthy, worst first, each with group, version, kind, namespace, name, status, message and the profiles owning the check. Filters profile_kind, profile_namespace, profile_name, resource_kind, resource_namespace, resource_name, plus limit and skip. Resources reported by a HealthCheck no profile owns are listed without profiles and carry the HealthCheck name.
• healthStatus is omitted when no profile on the cluster has monitoring enabled. Values: NoData, Healthy, Suspended, Progressing, Degraded.

RBAC: get/list/watch on healthchecks and healthcheckreports (role.yaml and manifest.yaml). The helm chart needs the same rules.