Skip to content

High Availability

High availability (HA) helps keep your cluster’s Kubernetes API available during maintenance or a single-node failure. It is optional and set when you create a cluster.

Enable HA

Section titled “Enable HA”

In the console, open Clusters → Create Cluster, enter the cluster details, and select High Availability under Reliability.

With the CLI:

Terminal window
kupe cluster create prod \
  --cpu-limit 8 --memory-limit 32Gi --storage-limit 200Gi \
  --high-availability --wait

With Terraform, set high_availability = true on kupe_cluster at creation.

Existing clusters

Section titled “Existing clusters”

You cannot change HA on an existing cluster. Create a new cluster with the setting you want, migrate your applications and data, test them, then switch traffic. In Terraform, changing this setting replaces the cluster and destroys its workloads and storage. Use a separate resource for the new cluster during migration.

Check status

Section titled “Check status”

Check the cluster in the console or run kupe cluster get prod. ha-healthy means HA is ready. If the cluster remains degraded or unavailable, contact support.

HA covers the cluster API within one region. Your applications still need suitable replica counts, health checks, and backups. HA does not increase workload capacity or provide application data recovery.