Sveltos Custom Resource Definitions (CRDs)
Sveltos exposes its functionality through a set of Kubernetes CRDs. Users create and manage these resources through the management cluster API, and a dedicated controller reconciles each of them. The table below maps every CRD to what it defines and the controller responsible for it. For a quick reference, check out the visual representation.
CRD Reference
| CRD | What it defines | Reconciled by |
|---|---|---|
| ClusterProfile / Profile | What needs to be deployed on which sets of clusters. | addon-controller |
| ClusterPromotion | How and when the deployments safely roll out across different environments. | addon-controller |
| Classifier | Classify a managed cluster based on its live state. | classifier-manager |
| ManagementClusterClassifier | Classify a managed cluster based on resources in the control cluster. | classifier-manager |
| EventSource | Defines what an event is. | event-manager |
| EventTrigger | When the event happens, it triggers Sveltos to deploy resources. | event-manager |
| HealthCheck | Defines the criteria for what healthy looks like for resources. | healthcheck-manager |
| ClusterHealthCheck | Send notifications based on the health of Kubernetes resources. | healthcheck-manager |
| Techsupport | Defines which resources/logs to collect and where to send them. | techsupport |
| RoleRequest | Define tenant permissions. | access-manager |
Grouping by Controller
The CRDs group naturally around the controller that owns them.
- addon-controller:
ClusterProfile/Profile,ClusterPromotion. The deployment and rollout layer. What is deployed, where, and how it is promoted across environments. - classifier-manager:
Classifier,ManagementClusterClassifier. Cluster classification based on either the managed cluster's live state or resources in the control cluster. - event-manager:
EventSource,EventTrigger. The event-driven layer defines events and the actions taken when they fire. - healthcheck-manager:
HealthCheck,ClusterHealthCheck. Health criteria for resources and notifications based on cluster health. - techsupport:
Techsupport. Collection of logs and resources for support bundles. - access-manager:
RoleRequest. Tenant permissions for multi-tenancy.