High Availability (HA) Control Plane on LKE

In Kubernetes, the control plane is the set of components that orchestrate the cluster and manage the worker nodes (Compute Instances) in that cluster and the pods (containers) within the worker nodes. The control plane components include the Kubernetes API server (kube-api-server), etcd, the Kubernetes scheduler (kube-scheduler), the cloud control manager (cloud-control-manager), and the Kubernetes controller manager (kube-controller-manager).

Within the Linode Platform, this control plane is fully managed by LKE (the Linode Kubernetes Engine). By default, these components are not replicated. If any of them fail, it’s possible that there may be issues with your cluster. Enabling HA (High Availability) Control Plane adds an additional layer of redundancy by replicating these control plane components. In doing so, this feature can ensure that the cluster experiences maximum uptime (99.99% guaranteed) and is recommended for all production applications running on LKE.

Cost

The HA control plane feature is an optional billable service that starts at $60/month per cluster, depending on region. For more information, see our pricing page.

Important

While upgrading to an HA cluster is always possible, downgrading your cluster is not currently supported. Enabling HA is an irreversible change for your cluster.

Replication Details

When HA control plane is enabled on a cluster, the following control plane components are replicated:

etcd and kube-api-server increases from one to three replicas.
All other components, including the Cloud Controller Manager, kube-scheduler, and kube-controller-manager, increase from one to two replicas, with leader election put in place.

When multiple replicas are created, they are always placed on separate infrastructure to better support uptime and redundancy. This configuration maintains a guaranteed 99.99% uptime for the control plane and worker nodes.

Enable During Cluster Creation

To enable this feature when creating a cluster, check the Enable HA Control Plane box in the Cluster Summary section. Review the Create a Cluster guide for full instructions.

Enable on Existing Clusters

If a cluster was created without this feature, you can enable it at any time.

Warning

Enabling the high availability control plane feature may result in brief and temporary downtime while the cluster is upgraded and the following actions are performed:

All nodes will be deleted and new nodes are created to replace them.
Any local storage (such as hostPath volumes) will be erased.

The upgrade process may take several minutes to complete, as nodes are replaced on a rolling basis.

High Availability can be added to pre-existing clusters at any given time through the cluster’s Summary Page.

To reach the summary page for the cluster, navigate first to the Kubernetes section of the Cloud Manager.
Select the Cluster by label that you would like to enable HA for. The summary page for the cluster appears.
To enable HA, select the Upgrade to HA button at the top of the page.
A new window appears, asking you to confirm all of the changes that come with High Availability. Read through the message and select the Enable HA Control Plane checkbox to confirm that you agree to the changes. Then click the Upgrade to HA button.