Skip to main content

Deploy a PostgreSQL Helm Chart Using an Existing EBS Volume

Deploy the Bitnami PostgreSQL Helm chart inside a SleakOps cluster pointing to a pre-existing EBS volume, by creating the required PersistentVolume and PersistentVolumeClaim resources.

info

This tutorial assumes you have already followed Migrate EBS Volumes to a New Account and have a migrated volume ID available.

Prerequisites​

  • A Cluster configured in SleakOps with EBS CSI Driver installed
  • An Environment and Project deployed
  • The EBS Volume ID and its availability zone
  • kubectl and helm installed locally
  • The SleakOps cluster's NodePool name

Let's Start​

Step 1 — (Optional) Identify the volume ID​

If you migrated the volume using the EBS migration guide, run this command from a Pod with the required IAM permissions to list migrated volumes:

aws ec2 describe-volumes \
  --filters Name=status,Values=available \
  --query "Volumes[?Tags[?Key=='Name'] && Tags[?Key=='Pod'] && Tags[?Key=='OriginalVolumeId']].[{VolumeId: VolumeId, Name: Tags[?Key=='Name']|[0].Value, Pod: Tags[?Key=='Pod']|[0].Value, OriginalVolumeId: Tags[?Key=='OriginalVolumeId']|[0].Value, Size: Size}]" \
  --output table

Copy the VolumeId of the target volume.

Step 2 — Create the PV, PVC, and deploy with Helm​

Set the variables, then run the script to create the Kubernetes resources and deploy the chart:

volume_id="vol-111aaa222ddd3333fff44"
pv_name="your-pv-name"
storage_size="8Gi"
availability_zone="us-east-1a"
nodepool_name="your-nodepool-name"

pvc_name="${pv_name}-claim"

# Create PV and PVC
kubectl apply -f - <<EOF
---
apiVersion: v1
kind: PersistentVolume
metadata:
  name: $pv_name
spec:
  capacity:
    storage: $storage_size
  volumeMode: Filesystem
  accessModes:
    - ReadWriteOnce
  persistentVolumeReclaimPolicy: Retain
  storageClassName: ebs-csi-default-sc
  csi:
    driver: ebs.csi.aws.com
    volumeHandle: $volume_id
    fsType: ext4
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: $pvc_name
spec:
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: $storage_size
  storageClassName: ebs-csi-default-sc
  volumeName: $pv_name
EOF

# Create Helm values file
cat > values.yaml <<EOF
primary:
  persistence:
    existingClaim: $pvc_name

  affinity:
    nodeAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:
        nodeSelectorTerms:
          - matchExpressions:
              - key: topology.kubernetes.io/zone
                operator: In
                values:
                  - $availability_zone

  tolerations:
    - key: "karpenter.sh/nodepool"
      operator: "Equal"
      value: $nodepool_name
EOF

# Deploy chart
helm repo add bitnami https://charts.bitnami.com/bitnami
helm repo update
helm install my-postgres bitnami/postgresql -f values.yaml
warning

Node affinity is required. If you omit the nodeAffinity configuration, Kubernetes may schedule the Pod in a different availability zone from the volume, causing the Pod to remain in a Pending state indefinitely.

warning

Tolerations are required. SleakOps clusters use Karpenter node pools. The tolerations block ensures the Pod is scheduled on the correct nodes.