Uploads to Rancher

Rancher manages clusters through its control plane. Managed clusters send data to Rancher's central management servers. This includes "always-on" data, exchanged with Rancher whenever the cluster has Internet access, and "on-demand" data, which should be explicitly requested by GetVisibility Support via the Rancher UI.

Always-On Data sent to Rancher includes:

1. Cluster Metadata:

   • Information about the cluster

   • Nodes list and metadata (IP address, hostname, cluster role, etc.)

   • K3s version

2. Health and Monitoring Data:

   • CPU and RAM usage on each cluster node

   • Current Metrics (via Prometheus)

   • Fleet agent heartbeat

   • Fleet bundle synchronization data

   • Current cluster status (healthy/unhealthy)

On-Demand Data:

1. Cluster Metadata:

   • Resource allocation (which Kubernetes resource runs on which node)

   • Current cluster-level Alerts

   • Current cluster-level Events

2. Kubernetes Objects:

   • List Kubernetes objects (usually Pods and Configurations)

   • Delete one or more objects

   • Create a new Kubernetes definition

   • Update existing definition (limited)

3. Kubernetes Container Logs:

   • Current logs via Rancher UI

   • Historical logs via Loki and Prometheus

4. Health and Monitoring Data:

   • Historical metrics via Grafana

kubectl Commands Output:

Rancher allows running kubectl exec into running containers, but this feature is blocked by our WAF. Support needs SSH access or screen-sharing with the customer to execute these commands.

None of those categories are critical for operation, and access to Rancher can be disabled after deployment.