Vertical Scaling for ZooKeeper Clusters with KubeBlocks

This guide demonstrates how to vertically scale a ZooKeeper cluster managed by KubeBlocks by adjusting compute resources (CPU and memory).

Vertical scaling modifies compute resources (CPU and memory) for ZooKeeper instances while maintaining replica count. Key characteristics:

• Non-disruptive: Maintains ensemble availability during scaling
• Granular: Adjust CPU, memory, or both independently
• Reversible: Scale up or down as needed

KubeBlocks orchestrates scaling with minimal impact — followers are updated first, then the leader.

Prerequisites

Before proceeding, verify your environment meets these requirements:

• A functional Kubernetes cluster (v1.21+ recommended)
• kubectl v1.21+ installed and configured with cluster access
• Helm installed (installation guide)
• KubeBlocks installed (installation guide)
• ZooKeeper Add-on installed and a ZooKeeper cluster running (see Quickstart)

Vertical Scale

apiVersion: operations.kubeblocks.io/v1alpha1
kind: OpsRequest
metadata:
  name: zookeeper-verticalscaling
  namespace: demo
spec:
  clusterName: zookeeper-cluster
  type: VerticalScaling
  verticalScaling:
  - componentName: zookeeper
    requests:
      cpu: '1'
      memory: 1Gi
    limits:
      cpu: '1'
      memory: 1Gi

kubectl apply -f https://raw.githubusercontent.com/apecloud/kubeblocks-addons/refs/heads/main/examples/zookeeper/verticalscale.yaml

kubectl get ops zookeeper-verticalscaling -n demo -w

NAME                        TYPE              CLUSTER             STATUS    PROGRESS   AGE
zookeeper-verticalscaling   VerticalScaling   zookeeper-cluster   Running   0/3        14s
zookeeper-verticalscaling   VerticalScaling   zookeeper-cluster   Running   1/3        69s
zookeeper-verticalscaling   VerticalScaling   zookeeper-cluster   Running   2/3        2m49s
zookeeper-verticalscaling   VerticalScaling   zookeeper-cluster   Succeed   3/3        3m55s

apiVersion: apps.kubeblocks.io/v1
kind: Cluster
metadata:
  name: zookeeper-cluster
  namespace: demo
spec:
  terminationPolicy: Delete
  componentSpecs:
    - name: zookeeper
      componentDef: zookeeper
      serviceVersion: 3.9.2
      replicas: 3
      resources:
        requests:
          cpu: "1"       # Update as needed
          memory: "1Gi"  # Update as needed
        limits:
          cpu: "1"       # Update as needed
          memory: "1Gi"  # Update as needed
      volumeClaimTemplates:
        - name: data
          spec:
            storageClassName: ""
            accessModes:
              - ReadWriteOnce
            resources:
              requests:
                storage: 20Gi
        - name: snapshot-log
          spec:
            storageClassName: ""
            accessModes:
              - ReadWriteOnce
            resources:
              requests:
                storage: 20Gi

Verify

After scaling, confirm the updated resources:

kubectl get cluster zookeeper-cluster -n demo

NAME                CLUSTER-DEFINITION   TERMINATION-POLICY   STATUS    AGE
zookeeper-cluster                        Delete               Running   7m

kubectl get pods -n demo -l app.kubernetes.io/instance=zookeeper-cluster -o jsonpath='{range .items[*]}{.metadata.name}{"\t"}{.spec.containers[0].resources.limits}{"\n"}{end}'

Best Practices

• Scale during maintenance windows or low-traffic periods
• Verify Kubernetes cluster has sufficient resources before scaling
• Set identical requests/limits for guaranteed QoS
• Monitor ZooKeeper performance metrics after scaling

Cleanup

kubectl delete cluster zookeeper-cluster -n demo
kubectl delete ns demo