Edit

Share via

Configure and deploy a Valkey cluster on Azure Kubernetes Service (AKS)

  • Article

In this article, we configure and deploy a Valkey cluster on Azure Kubernetes Service (AKS).

Note

This article contains references to the terms master and slave, which are terms that Microsoft no longer uses. When the term is removed from the Valkey software, we’ll remove it from this article.

Create a namespace

  1. Create a namespace for the Valkey cluster using the kubectl create namespace command.

    kubectl create namespace ${SERVICE_ACCOUNT_NAMESPACE} --dry-run=client --output yaml | kubectl apply -f -

    Example output:

    namespace/valkey created

Create secrets

  1. Generate a random password for the Valkey cluster using openssl and store it in your Azure key vault using the az keyvault secret set command. Set the policy to allow the user-assigned identity to get the secret using the az keyvault set-policy command.

    SECRET=$(openssl rand -base64 32)
echo requirepass $SECRET > /tmp/valkey-password-file.conf
echo primaryauth $SECRET >> /tmp/valkey-password-file.conf
az keyvault secret set --vault-name $MY_KEYVAULT_NAME --name valkey-password-file --file /tmp/valkey-password-file.conf --output table
rm /tmp/valkey-password-file.conf
az keyvault set-policy --name $MY_KEYVAULT_NAME --object-id $userAssignedObjectID --secret-permissions get --output table

  2. Create a SecretProviderClass resource to access the Valkey password stored in your key vault using the kubectl apply command.

    kubectl apply -f - <<EOF
---
apiVersion: secrets-store.csi.x-k8s.io/v1
kind: SecretProviderClass
metadata:
  name: valkey-password
  namespace: valkey
spec:
  provider: azure
  parameters:
    usePodIdentity: "false"
    useVMManagedIdentity: "true"
    userAssignedIdentityID: "${userAssignedIdentityID}"
    keyvaultName: ${MY_KEYVAULT_NAME}              # the name of the AKV instance
    objects: |
      array:
        - |
          objectName: valkey-password-file
          objectAlias: valkey-password-file.conf
          objectType: secret
    tenantId: "${TENANT_ID}" # the tenant ID of the AKV instance
EOF

Deploy the Valkey cluster

  1. Create a ConfigMap resource to store the Valkey configuration file.

    kubectl apply -f - <<EOF
apiVersion: v1
kind: ConfigMap
metadata:
  name: valkey-cluster
  namespace: valkey
data:
  valkey.conf:  |+
    cluster-enabled yes
    cluster-node-timeout 15000
    cluster-config-file /data/nodes.conf
    appendonly yes
    protected-mode yes
    dir /data
    port 6379
    include /etc/valkey-password/valkey-password-file.conf
EOF

    Example output:

    configmap/valkey-cluster created

  2. Create a StatefulSet resource with a spec.affinity goal is to keep all primaries in zone 1 and zone 2, preferably in different nodes, using the kubectl apply command.

    kubectl apply -f - <<EOF
---
apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: valkey-masters
  namespace: valkey
spec:
  serviceName: "valkey-masters"
  replicas: 3
  selector:
    matchLabels:
      app: valkey
  template:
    metadata:
      labels:
        app: valkey
        appCluster: valkey-masters
    spec:
      terminationGracePeriodSeconds: 20
      affinity:
        nodeAffinity:
          requiredDuringSchedulingIgnoredDuringExecution:
            nodeSelectorTerms:
            - matchExpressions:
              - key: agentpool
                operator: In
                values:
                - valkey
              - key: topology.kubernetes.io/zone
                operator: In
                values:
                - ${MY_LOCATION}-1
            - matchExpressions:
              - key: agentpool
                operator: In
                values:
                - valkey
              - key: topology.kubernetes.io/zone
                operator: In
                values:
                - ${MY_LOCATION}-2
        podAntiAffinity:
          preferredDuringSchedulingIgnoredDuringExecution:
          - weight: 100
            podAffinityTerm:
              labelSelector:
                matchExpressions:
                - key: app
                  operator: In
                  values:
                  - valkey
              topologyKey: topology.kubernetes.io/zone
          - weight: 90
            podAffinityTerm:
              labelSelector:
                matchExpressions:
                - key: app
                  operator: In
                  values:
                  - valkey
              topologyKey: kubernetes.io/hostname
      containers:
      - name: role-master-checker
        image: "${MY_ACR_REGISTRY}.azurecr.io/valkey:latest"
        command:
          - "/bin/bash"
          - "-c"
        args:
          [
            "while true; do role=\$(valkey-cli --pass \$(cat /etc/valkey-password/valkey-password-file.conf | awk '{print \$2; exit}') role | awk '{print \$1; exit}');     if [ \"\$role\" = \"slave\" ]; then valkey-cli --pass \$(cat /etc/valkey-password/valkey-password-file.conf | awk '{print \$2; exit}') cluster failover; fi; sleep 30; done"
          ]
        volumeMounts:
        - name: valkey-password
          mountPath: /etc/valkey-password
          readOnly: true
      - name: valkey
        image: "${MY_ACR_REGISTRY}.azurecr.io/valkey:latest"
        env:
        - name: VALKEY_PASSWORD_FILE
          value: "/etc/valkey-password/valkey-password-file.conf"
        - name: MY_POD_IP
          valueFrom:
            fieldRef:
              fieldPath: status.podIP
        command:
          - "valkey-server"
        args:
          - "/conf/valkey.conf"
          - "--cluster-announce-ip"
          - "\$(MY_POD_IP)"
        resources:
          requests:
            cpu: "100m"
            memory: "100Mi"
        ports:
            - name: valkey
              containerPort: 6379
              protocol: "TCP"
            - name: cluster
              containerPort: 16379
              protocol: "TCP"
        volumeMounts:
        - name: conf
          mountPath: /conf
          readOnly: false
        - name: data
          mountPath: /data
          readOnly: false
        - name: valkey-password
          mountPath: /etc/valkey-password
          readOnly: true
      volumes:
      - name: valkey-password
        csi:
          driver: secrets-store.csi.k8s.io
          readOnly: true
          volumeAttributes:
            secretProviderClass: valkey-password
      - name: conf
        configMap:
          name: valkey-cluster
          defaultMode: 0755
  volumeClaimTemplates:
  - metadata:
      name: data
    spec:
      accessModes: [ "ReadWriteOnce" ]
      storageClassName: managed-csi-premium
      resources:
        requests:
          storage: 20Gi
EOF

    Example output:

    statefulset.apps/valkey-masters created

  3. Create a second StatefulSet resource for the Valkey secondaries with a spec.affinity goal to keep all replicas in zone 3, preferably in different nodes, using the kubectl apply command.

    kubectl apply -f - <<EOF
---
apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: valkey-replicas
  namespace: valkey
spec:
  serviceName: "valkey-replicas"
  replicas: 3
  selector:
    matchLabels:
      app: valkey
  template:
    metadata:
      labels:
        app: valkey
        appCluster: valkey-replicas
    spec:
      terminationGracePeriodSeconds: 20
      affinity:
        nodeAffinity:
          requiredDuringSchedulingIgnoredDuringExecution:
            nodeSelectorTerms:
            - matchExpressions:
              - key: agentpool
                operator: In
                values:
                - valkey
              - key: topology.kubernetes.io/zone
                operator: In
                values:
                - ${MY_LOCATION}-3
        podAntiAffinity:
          preferredDuringSchedulingIgnoredDuringExecution:
          - weight: 90
            podAffinityTerm:
              labelSelector:
                matchExpressions:
                - key: app
                  operator: In
                  values:
                  - valkey
              topologyKey: kubernetes.io/hostname
      containers:
      - name: valkey
        image: "${MY_ACR_REGISTRY}.azurecr.io/valkey:latest"
        env:
        - name: VALKEY_PASSWORD_FILE
          value: "/etc/valkey-password/valkey-password-file.conf"
        - name: MY_POD_IP
          valueFrom:
            fieldRef:
              fieldPath: status.podIP
        command:
          - "valkey-server"
        args:
          - "/conf/valkey.conf"
          - "--cluster-announce-ip"
          - "\$(MY_POD_IP)"
        resources:
          requests:
            cpu: "100m"
            memory: "100Mi"
        ports:
            - name: valkey
              containerPort: 6379
              protocol: "TCP"
            - name: cluster
              containerPort: 16379
              protocol: "TCP"
        volumeMounts:
        - name: conf
          mountPath: /conf
          readOnly: false
        - name: data
          mountPath: /data
          readOnly: false
        - name: valkey-password
          mountPath: /etc/valkey-password
          readOnly: true
      volumes:
      - name: valkey-password
        csi:
          driver: secrets-store.csi.k8s.io
          readOnly: true
          volumeAttributes:
            secretProviderClass: valkey-password
      - name: conf
        configMap:
          name: valkey-cluster
          defaultMode: 0755
  volumeClaimTemplates:
  - metadata:
      name: data
    spec:
      accessModes: [ "ReadWriteOnce" ]
      storageClassName: managed-csi-premium
      resources:
        requests:
          storage: 20Gi
EOF

    Example output:

    statefulset.apps/valkey-replicas created

  4. Verify that master-N and replica-N are running in different nodes and zones using the kubectl get nodes and kubectl get pods commands.

    kubectl get pods -n valkey -o wide
kubectl get node -o custom-columns=Name:.metadata.name,Zone:".metadata.labels.topology\.kubernetes\.io/zone"

    Example output:

    NAME                READY   STATUS    RESTARTS   AGE     IP             NODE                             NOMINATED NODE   READINESS GATES
valkey-masters-0    1/1     Running   0          2m55s   10.224.0.4     aks-valkey-18693609-vmss000004   <none>           <none>
valkey-masters-1    1/1     Running   0          2m31s   10.224.0.137   aks-valkey-18693609-vmss000000   <none>           <none>
valkey-masters-2    1/1     Running   0          2m7s    10.224.0.222   aks-valkey-18693609-vmss000001   <none>           <none>
valkey-replicas-0   1/1     Running   0          88s     10.224.0.237   aks-valkey-18693609-vmss000005   <none>           <none>
valkey-replicas-1   1/1     Running   0          70s     10.224.0.18    aks-valkey-18693609-vmss000002   <none>           <none>
valkey-replicas-2   1/1     Running   0          48s     10.224.0.242   aks-valkey-18693609-vmss000005   <none>           <none>
Name                                Zone
aks-nodepool1-17621399-vmss000000   centralus-1
aks-nodepool1-17621399-vmss000001   centralus-2
aks-nodepool1-17621399-vmss000003   centralus-3
aks-valkey-18693609-vmss000000      centralus-1
aks-valkey-18693609-vmss000001      centralus-2
aks-valkey-18693609-vmss000002      centralus-3
aks-valkey-18693609-vmss000003      centralus-1
aks-valkey-18693609-vmss000004      centralus-2
aks-valkey-18693609-vmss000005      centralus-3

    Wait for all pods to be running before proceeding to the next step.

  5. Create three headless Service resources (the first for the entire cluster, the second for the primaries, and the third for the secondaries) to use to get the IP addresses of the Valkey pods using the kubectl apply command.

    kubectl apply -f - <<EOF
apiVersion: v1
kind: Service
metadata:
  name: valkey-cluster
  namespace: valkey
spec:
  clusterIP: None
  ports:
  - name: valkey-port
    port: 6379
    protocol: TCP
    targetPort: 6379
  selector:
    app: valkey
  sessionAffinity: None
  type: ClusterIP
EOF

kubectl apply -f - <<EOF
apiVersion: v1
kind: Service
metadata:
  name: valkey-masters
  namespace: valkey
spec:
  clusterIP: None
  ports:
  - name: valkey-port
    port: 6379
    protocol: TCP
    targetPort: 6379
  selector:
    app: valkey
    appCluster: valkey-masters
  sessionAffinity: None
  type: ClusterIP
EOF

kubectl apply -f - <<EOF
apiVersion: v1
kind: Service
metadata:
  name: valkey-replicas
  namespace: valkey
spec:
  clusterIP: None
  ports:
  - name: valkey-port
    port: 6379
    protocol: TCP
    targetPort: 6379
  selector:
    app: valkey
    appCluster: valkey-replicas
  sessionAffinity: None
  type: ClusterIP
EOF

    Example output:

    service/valkey-cluster created
service/valkey-masters created
service/valkey-replicas created

  6. Create a Pod Disruption Budget (PDB) to ensure always that one pod at most is unavailable.

    kubectl apply -f - <<EOF
apiVersion: policy/v1
kind: PodDisruptionBudget
metadata:
  name: valkey
  namespace: valkey
spec:
  maxUnavailable: 1
  selector:
    matchLabels:
      app: valkey
EOF

    Example output:

    poddisruptionbudget.policy/valkey created

Run the Valkey cluster

  1. Add the Valkey primaries, in zone 1 and 2, to the cluster using the kubectl exec command.

    kubectl exec -it -n valkey valkey-masters-0 -- valkey-cli --cluster create --cluster-yes --cluster-replicas 0 \
                    valkey-masters-0.valkey-masters.valkey.svc.cluster.local:6379 \
                    valkey-masters-1.valkey-masters.valkey.svc.cluster.local:6379 \
                    valkey-masters-2.valkey-masters.valkey.svc.cluster.local:6379 \
                    --pass ${SECRET}

    Example output:

    >>> Performing hash slots allocation on 3 nodes...
Master[0] -> Slots 0 - 5460
Master[1] -> Slots 5461 - 10922
Master[2] -> Slots 10923 - 16383
M: ee6ac1d00d3f016b6f46c7ce11199bc1a7809a35 valkey-masters-0.valkey-masters.valkey.svc.cluster.local:6379
   slots:[0-5460] (5461 slots) master
M: fd1fb98db83976478e05edd3d2a02f9a13badd80 valkey-masters-1.valkey-masters.valkey.svc.cluster.local:6379
   slots:[5461-10922] (5462 slots) master
M: ea47bf57ae7080ef03164a4d48b662c7b4c8770e valkey-masters-2.valkey-masters.valkey.svc.cluster.local:6379
   slots:[10923-16383] (5461 slots) master
>>> Nodes configuration updated
>>> Assign a different config epoch to each node
>>> Sending CLUSTER MEET messages to join the cluster
Waiting for the cluster to join
...
>>> Performing Cluster Check (using node valkey-masters-0.valkey-masters.valkey.svc.cluster.local:6379)
M: ee6ac1d00d3f016b6f46c7ce11199bc1a7809a35 valkey-masters-0.valkey-masters.valkey.svc.cluster.local:6379
   slots:[0-5460] (5461 slots) master
M: ea47bf57ae7080ef03164a4d48b662c7b4c8770e 10.224.0.176:6379
   slots:[10923-16383] (5461 slots) master
M: fd1fb98db83976478e05edd3d2a02f9a13badd80 10.224.0.247:6379
   slots:[5461-10922] (5462 slots) master
[OK] All nodes agree about slots configuration.
>>> Check for open slots...
>>> Check slots coverage...
[OK] All 16384 slots covered.

  2. Add the Valkey replicas, in zone 3, to the cluster using the kubectl exec command.

    kubectl exec -ti -n valkey valkey-masters-0 -- valkey-cli --cluster add-node \
                    valkey-replicas-0.valkey-replicas.valkey.svc.cluster.local:6379 \
                    valkey-masters-0.valkey-masters.valkey.svc.cluster.local:6379  --cluster-slave \
                    --pass ${SECRET}

kubectl exec -ti -n valkey valkey-masters-0 -- valkey-cli --cluster add-node \
                    valkey-replicas-1.valkey-replicas.valkey.svc.cluster.local:6379 \
                    valkey-masters-1.valkey-masters.valkey.svc.cluster.local:6379  --cluster-slave \
                    --pass ${SECRET}

kubectl exec -ti -n valkey valkey-masters-0 -- valkey-cli --cluster add-node \
                    valkey-replicas-2.valkey-replicas.valkey.svc.cluster.local:6379 \
                    valkey-masters-2.valkey-masters.valkey.svc.cluster.local:6379  --cluster-slave \
                    --pass ${SECRET}

    Example output:

    >>> Adding node valkey-replicas-0.valkey-replicas.valkey.svc.cluster.local:6379 to cluster valkey-masters-0.valkey-masters.valkey.svc.cluster.local:6379
>>> Performing Cluster Check (using node valkey-masters-0.valkey-masters.valkey.svc.cluster.local:6379)
M: ee6ac1d00d3f016b6f46c7ce11199bc1a7809a35 valkey-masters-0.valkey-masters.valkey.svc.cluster.local:6379
   slots:[0-5460] (5461 slots) master
M: ea47bf57ae7080ef03164a4d48b662c7b4c8770e 10.224.0.176:6379
   slots:[10923-16383] (5461 slots) master
M: fd1fb98db83976478e05edd3d2a02f9a13badd80 10.224.0.247:6379
   slots:[5461-10922] (5462 slots) master
[OK] All nodes agree about slots configuration.
>>> Check for open slots...
>>> Check slots coverage...
[OK] All 16384 slots covered.
Automatically selected master valkey-masters-0.valkey-masters.valkey.svc.cluster.local:6379
>>> Send CLUSTER MEET to node valkey-replicas-0.valkey-replicas.valkey.svc.cluster.local:6379 to make it join the cluster.
Waiting for the cluster to join

>>> Configure node as replica of valkey-masters-0.valkey-masters.valkey.svc.cluster.local:6379.
[OK] New node added correctly.
>>> Adding node valkey-replicas-1.valkey-replicas.valkey.svc.cluster.local:6379 to cluster valkey-masters-1.valkey-masters.valkey.svc.cluster.local:6379
>>> Performing Cluster Check (using node valkey-masters-1.valkey-masters.valkey.svc.cluster.local:6379)
M: fd1fb98db83976478e05edd3d2a02f9a13badd80 valkey-masters-1.valkey-masters.valkey.svc.cluster.local:6379
   slots:[5461-10922] (5462 slots) master
S: 0ebceb60cbcc31da9040159440a1f4856b992907 10.224.0.224:6379
   slots: (0 slots) slave
   replicates ee6ac1d00d3f016b6f46c7ce11199bc1a7809a35
M: ea47bf57ae7080ef03164a4d48b662c7b4c8770e 10.224.0.176:6379
   slots:[10923-16383] (5461 slots) master
M: ee6ac1d00d3f016b6f46c7ce11199bc1a7809a35 10.224.0.14:6379
   slots:[0-5460] (5461 slots) master
   1 additional replica(s)
[OK] All nodes agree about slots configuration.
>>> Check for open slots...
>>> Check slots coverage...
[OK] All 16384 slots covered.
Automatically selected master valkey-masters-1.valkey-masters.valkey.svc.cluster.local:6379
>>> Send CLUSTER MEET to node valkey-replicas-1.valkey-replicas.valkey.svc.cluster.local:6379 to make it join the cluster.
Waiting for the cluster to join

>>> Configure node as replica of valkey-masters-1.valkey-masters.valkey.svc.cluster.local:6379.
[OK] New node added correctly.
>>> Adding node valkey-replicas-2.valkey-replicas.valkey.svc.cluster.local:6379 to cluster valkey-masters-2.valkey-masters.valkey.svc.cluster.local:6379
>>> Performing Cluster Check (using node valkey-masters-2.valkey-masters.valkey.svc.cluster.local:6379)
M: ea47bf57ae7080ef03164a4d48b662c7b4c8770e valkey-masters-2.valkey-masters.valkey.svc.cluster.local:6379
   slots:[10923-16383] (5461 slots) master
S: 0ebceb60cbcc31da9040159440a1f4856b992907 10.224.0.224:6379
   slots: (0 slots) slave
   replicates ee6ac1d00d3f016b6f46c7ce11199bc1a7809a35
S: fa44edff683e2e01ee5c87233f9f3bc35c205dce 10.224.0.103:6379
   slots: (0 slots) slave
   replicates fd1fb98db83976478e05edd3d2a02f9a13badd80
M: ee6ac1d00d3f016b6f46c7ce11199bc1a7809a35 10.224.0.14:6379
   slots:[0-5460] (5461 slots) master
   1 additional replica(s)
M: fd1fb98db83976478e05edd3d2a02f9a13badd80 10.224.0.247:6379
   slots:[5461-10922] (5462 slots) master
   1 additional replica(s)
[OK] All nodes agree about slots configuration.
>>> Check for open slots...
>>> Check slots coverage...
[OK] All 16384 slots covered.
Automatically selected master valkey-masters-2.valkey-masters.valkey.svc.cluster.local:6379
>>> Send CLUSTER MEET to node valkey-replicas-2.valkey-replicas.valkey.svc.cluster.local:6379 to make it join the cluster.
Waiting for the cluster to join

>>> Configure node as replica of valkey-masters-2.valkey-masters.valkey.svc.cluster.local:6379.
[OK] New node added correctly.

  3. Verify the roles of the pods using the following commands:

    for x in $(seq 0 2); do echo "valkey-masters-$x"; kubectl exec -n valkey valkey-masters-$x  -- valkey-cli --pass ${SECRET} role; echo; done
for x in $(seq 0 2); do echo "valkey-replicas-$x"; kubectl exec -n valkey valkey-replicas-$x -- valkey-cli --pass ${SECRET} role; echo; done

    Example output:

    valkey-masters-0
master
84
10.224.0.224
6379
84

valkey-masters-1
master
84
10.224.0.103
6379
84

valkey-masters-2
master
70
10.224.0.200
6379
70

valkey-replicas-0
slave
10.224.0.14
6379
connected
98

valkey-replicas-1
slave
10.224.0.247
6379
connected
98

valkey-replicas-2
slave
10.224.0.176
6379
connected
84

Next steps

Validate the resiliency of the Valkey cluster on AKS

To learn more about deploying open-source software on Azure Kubernetes Service (AKS), see the following articles:

Contributors

Microsoft maintains this article. The following contributors originally wrote it:

  • Nelly Kiboi | Service Engineer
  • Saverio Proto | Principal Customer Experience Engineer
  • Don High | Principal Customer Engineer
  • LaBrina Loving | Principal Service Engineer
  • Ken Kilty | Principal TPM
  • Russell de Pina | Principal TPM
  • Colin Mixon | Product Manager
  • Ketan Chawda | Senior Customer Engineer
  • Naveed Kharadi | Customer Experience Engineer
  • Erin Schaffer | Content Developer 2