Learn how much memory and CPU usage is consumed by the cluster agent extensions that are installed by VMware Tanzu Mission Control on managed Kubernetes clusters, both attached and provisioned.

When you attach a cluster, register a management cluster, or create a workload cluster in a registered management cluster in Tanzu Mission Control, the cluster agent service installs a set of cluster agent extensions and custom resource definitions into your cluster to enable Tanzu Mission Control to communicate with your cluster.

In the following table, memory is expressed in mebibytes/gibibytes and CPU is expressed in cores/millicores.
Table 1. Memory and CPU Reservations and Limits for Cluster Agent Extensions
Extension Node Memory Reservation Memory Limit CPU Reservation CPU Limit
agent-updater Master 100Mi 150Mi 100m 100m
agentupdater-workload Worker 100Mi 150Mi 100m 100m
cluster-health-extension Worker 128Mi 2Gi 100m 1000m
manager (data-protection) Worker 128Mi 512Mi 50m 100m
extension-manager Master 100Mi 150Mi 100m 100m
extension-updater Master 128Mi 512Mi 50m 100m
manager (inspection-extension) Worker 128Mi 256Mi 10m 500m
intent-agent Worker 150Mi 150Mi 100m 100m
manager (policy-sync-extension) Worker 128Mi 256Mi 10m 500m
manager (policy-webhook) Worker 128Mi 256Mi 100m 100m
sync-agent Worker 128Mi 2Gi 100m 2000m
tmc-observer Master 100Mi 150Mi 50m 100m

Note that information about all the extensions is shown in the agent and extensions health section of the cluster details page in the Tanzu Mission Control console. Some extensions consist of more than one pod, for example, the policy-webhook pod is part of the policy-sync-extension. You can check number of extensions running using this command: 

kubectl get extensions

Kubernetes Resource Units

For detailed information about kubernetes resource units and measurements, see https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/#resource-units-in-kubernetes.