Managed Add-ons
Managed Add-ons extend a Managed Kubernetes cluster with NetActuate-operated capabilities that you turn on from the portal instead of installing and maintaining yourself. Once enabled, NetActuate installs the add-on into your cluster, keeps it running, and reconciles it, so you get storage, DNS, and load balancing that are integrated with the platform without hand-managing controllers or CRDs. Manage them from the Managed Add-ons tab of your cluster in the portal.
What Managed Add-ons are
Each add-on is a controller or driver that NetActuate installs and owns inside your cluster. You request an add-on and provide any inputs it needs; NetActuate handles the installation, upgrades, and health. Because the platform owns the add-on, you should not hand-edit the resources it manages: change configuration through the portal, and use the resource's own Kubernetes API (a StorageClass, a Service annotation, an Ingress) as documented on each add-on page.
The available add-ons are:
- Storage - dynamic block volumes backed by NetActuate Ceph, via a managed StorageClass
- DNS - automatic DNS records for Services and Ingresses in a zone you own
- Load Balancing - Layer 4 and Layer 7 load balancing through the VPC gateway (in-VPC clusters only)
Add-on lifecycle
An add-on moves through a small set of states, surfaced on the Managed Add-ons tab:
- requested - you have enabled the add-on and NetActuate has accepted the request.
- installing - the controller or driver and its resources are being applied to the cluster.
- ready - the platform has completed the add-on installation or reconciliation request.
- degraded - the add-on installation or reconciliation request did not complete cleanly.
- failed - installation or reconciliation could not complete.
The add-on state field reflects the platform installation and reconciliation state. It does not report Kubernetes workload health today. An add-on can show ready while its controller pods, driver pods, or provisioner pods are failing inside the cluster. Monitor the add-on workloads with Kubernetes tools and alerts, not the add-on state alone.
When an add-on is degraded or failed, the tab shows a Repair action. Repair re-runs the add-on installation and reconciliation workflow and reapplies the resources the platform owns. Use it when an add-on is stuck, missing managed resources, or has drifted from the expected installed state. Repair does not replace normal Kubernetes workload monitoring.
Availability
Add-on availability depends on the cluster type. Storage works on any cluster. DNS works on both edge and in-VPC clusters. Load Balancing depends on the VPC gateway and is available only on in-VPC clusters.
| Add-on | Edge cluster | In-VPC cluster |
|---|---|---|
| Storage | Available | Available |
| DNS | Available | Available |
| Load Balancing | Not available | Available |
Some add-on status responses include podDistruptionDrainFailures. That spelling is the field name returned by the API today. The spelling is known to be wrong, but clients must use the returned field name until the API changes.
See Kubernetes in a VPC for the difference between edge and in-VPC clusters.
Need Help?
Contact support@netactuate.com or open a support ticket from the portal.