Guides
Getting Started
Deploy your first PostgreSQL cluster on Kubernetes in under five minutes.
This guide walks you through the fastest path to a running PostgreSQL cluster on Kubernetes. By the end, you’ll have a three-node highly available PostgreSQL setup with automated failover and backups configured.
Prerequisites
Before you begin, make sure you have:
- A running Kubernetes cluster (v1.28 or later)
kubectlconfigured to access the cluster- Cluster-admin privileges to install CRDs
Install the operator
Apply the operator manifest from the latest release:
kubectl apply --server-side -f \
https://releases.keldon.io/operator/latest.yaml
Verify the operator is running:
kubectl get pods -n cnb-system
You should see the operator pod in the Running state.
Deploy your first cluster
Create a file named cluster.yaml:
apiVersion: postgresql.cnb.io/v1
kind: Cluster
metadata:
name: my-first-cluster
spec:
instances: 3
storage:
size: 10Gi
bootstrap:
initdb:
database: app
owner: app
Apply it:
kubectl apply -f cluster.yaml
Verify the cluster
Watch the cluster come online:
kubectl get cluster my-first-cluster -w
Once ready, connect to it:
kubectl port-forward svc/my-first-cluster-rw 5432:5432
psql -h localhost -U app app
Next steps
- Configure backups to S3-compatible storage
- Tune PostgreSQL parameters
- Set up monitoring