Skip to main content
Version: main 🚧
Enterprise
Available in these plansFreeDevProdScale
Auto Sleep for Virtual Clusters
info

Platform auto sleep has been integrated into the standard sleep configuration and is no longer a separate feature.

Auto sleep and auto delete for tenant clusters

The platform provides two features to reduce Kubernetes costs:

  • Auto Sleep puts tenant clusters to sleep when nobody is using them. This removes all pods while keeping all resources inside the tenant clusters during periods of inactivity.
  • Auto-delete deletes tenant clusters that have been idle for a while.

Both features rely on the platform's inactivity detection.

Configure auto sleep​

Auto sleep can be configured per tenant cluster instance. When enabled, the platform monitors activity and puts the tenant cluster to sleep after a configured period of inactivity.

To configure auto sleep in the Platform UI:

  1. Navigate to the Tenant Clusters view.
  2. Select the tenant cluster you want to configure.
  3. Click Edit and navigate to the Sleep Mode section.
  4. Enable auto sleep and configure the inactivity timeout.
  5. Click Save.

You can also start sleep manually from the tenant cluster's ellipsis menu in the Platform UI.

For detailed vCluster-specific sleep mode configuration and examples, including vcluster.yaml settings, scheduled sleep, and advanced configuration, see the vCluster sleep mode documentation.

Configure auto-delete​

The platform lets you configure auto-delete for tenant clusters that have not been used for a certain period of time.

To configure auto-delete in the Platform UI:

  1. Navigate to the Tenant Clusters view.
  2. Select the tenant cluster you want to configure.
  3. Click Edit and navigate to the Auto-Delete section.
  4. Enable auto-delete and configure the inactivity timeout.
  5. Click Save.

For detailed configuration options, see the vCluster auto-delete documentation.

Inactivity detection​

All requests that are made through the platform count as activity in the tenant cluster. This includes kubectl requests, API requests, and access through the platform UI.

The platform tracks activity based on requests that are proxied through the platform. Direct access to the tenant cluster API server (for example, via ingress access) is not tracked by the platform's inactivity detection.

If you need more advanced activity detection, see the vCluster sleep mode documentation for configuration options including custom activity annotations.