k8s中安装redis6集群

注意:第二步不用操作了,yaml文件仅供参考

1.创建Namespace

kubectl apply -f Namespace.yaml

apiVersion: v1

kind: Namespace

metadata:

  name: cluster-redis

2.创建PersistentVolumeClaim #这一步不用操作了

kubectl apply -f PersistentVolumeClaim.yaml

---

apiVersion: v1

kind: PersistentVolumeClaim

metadata:

  name: redis-pvc1

  namespace: cluster-redis

spec:

  accessModes:

    - ReadWriteMany

  resources:

    requests:

      storage: 1Gi

  storageClassName: "storage"

---

---

apiVersion: v1

kind: PersistentVolumeClaim

metadata:

  name: redis-pvc2

  namespace: cluster-redis

spec:

  accessModes:

    - ReadWriteMany

  resources:

    requests:

      storage: 1Gi

  storageClassName: "storage"

---

apiVersion: v1

kind: PersistentVolumeClaim

metadata:

  name: redis-pvc3

  namespace: cluster-redis

spec:

  accessModes:

    - ReadWriteMany

  resources:

    requests:

      storage: 1Gi

  storageClassName: "storage"

---

apiVersion: v1

kind: PersistentVolumeClaim

metadata:

  name: redis-pvc4

  namespace: cluster-redis

spec:

  accessModes:

    - ReadWriteMany

  resources:

    requests:

      storage: 1Gi

  storageClassName: "storage"

---

apiVersion: v1

kind: PersistentVolumeClaim

metadata:

  name: redis-pvc5

  namespace: cluster-redis

spec:

  accessModes:

    - ReadWriteMany

  resources:

    requests:

      storage: 1Gi

  storageClassName: "storage"

---

apiVersion: v1

kind: PersistentVolumeClaim

metadata:

  name: redis-pvc6

  namespace: cluster-redis

spec:

  accessModes:

    - ReadWriteMany

  resources:

    requests:

      storage: 1Gi

  storageClassName: "storage"

3.创建statefulset

kubectl apply -f ConfigMap.yaml

kubectl apply -f StatefulSet.yaml

apiVersion: v1

kind: ConfigMap

metadata:

  name: redis-cluster-configmap

  namespace: cluster-redis

data:

  update-node.sh: |

    #!/bin/sh

    REDIS_NODES="/data/nodes.conf"

    sed -i -e "/myself/ s/[0-9]\{1,3\}\.[0-9]\{1,3\}\.[0-9]\{1,3\}\.[0-9]\{1,3\}/${POD_IP}/" ${REDIS_NODES}

    exec "$@"

  redis.conf: |+

    cluster-enabled yes

    cluster-require-full-coverage no

    cluster-node-timeout 15000

    cluster-config-file /data/nodes.conf

    cluster-migration-barrier 1

    appendonly yes

    protected-mode no

apiVersion: apps/v1

kind: StatefulSet

metadata:

  name: redis-pod

  namespace: cluster-redis

spec:

  serviceName: redis-cluster-service

  replicas: 6

  selector:

    matchLabels:

      app: redis-pod

  template:

    metadata:

      labels:

        app: redis-pod

    spec:

      containers:

      - name: redis

        image: redis:6.2.1

        ports:

        - containerPort: 6379

          name: client

        - containerPort: 16379

          name: gossip

        command: ["/conf/update-node.sh", "redis-server", "/conf/redis.conf"]

        env:

        - name: POD_IP

          valueFrom:

            fieldRef:

              fieldPath: status.podIP

        volumeMounts:

        - name: conf

          mountPath: /conf

          readOnly: false

        - name: data

          mountPath: /data

          readOnly: false

      volumes:

      - name: conf

        configMap:

          name: redis-cluster-configmap

          defaultMode: 0755

  volumeClaimTemplates:

  - metadata:

      name: data

    spec:

      accessModes: [ "ReadWriteMany" ]

      resources:

        requests:

          storage: 1Gi

      storageClassName: "storage" # 注意这个,事先创建好存储卷

4.创建service

kubectl apply -f Service.yaml

apiVersion: v1

kind: Service

metadata:

  name: redis-cluster-service

  namespace: cluster-redis

spec:

  type: ClusterIP

  ports:

  - port: 6379

    targetPort: 6379

    name: client

  - port: 16379

    targetPort: 16379

    name: gossip

  selector:

    app: redis-pod

5.初始化 Redis Cluster

注意如下这一步,分开来看:

(1) 命令: kubectl get pods -n cluster-redis -l app=redis-pod -o jsonpath='{range.items[*]}{.status.podIP}:6379 '|awk '{print $1" "$2" "$3" "$4" "$5" "$6" "}'

这个命令是获取到StatefulSet.yaml文件中生成的redis pod ip,最后那个awk是过滤到不需要的信息

比如执行kubectl get pods -n cluster-redis -l app=redis-pod -o jsonpath='{range.items[*]}{.status.podIP}:6379 '得到的结果如下:

10.0.2.122:6379 10.0.0.65:6379 10.0.2.123:6379 10.0.0.66:6379 10.0.0.69:6379 10.0.2.125:6379 :6379

可以看到最后有个多余的:6379是没用的,awk的作用就是去掉这个

(2) 命令:kubectl exec -it redis-pod-0 -n cluster-redis -- redis-cli --cluster create (pod ip) --cluster-replicas 1

这个命令是登录到其中一个pod中创建集群使用的

kubectl exec -it redis-pod-0 -n cluster-redis -- redis-cli --cluster create `kubectl get pods -n cluster-redis -l app=redis-pod -o jsonpath='{range.items[*]}{.status.podIP}:6379 '|awk '{print $1" "$2" "$3" "$4" "$5" "$6" "}'` --cluster-replicas 1

6.验证集群部署

kubectl exec -it redis-pod-0 -n cluster-redis -- redis-cli cluster info

for x in $(seq 0 5); do echo "redis-pod-$x"; kubectl exec redis-pod-$x -n cluster-redis -- redis-cli role; echo; done

7.注意事项

(1)StatefulSet类型的资源不能通过修改'replicas', 'template', and 'updateStrategy'等字段来进行更新,否则就会报错,具体如下:

The StatefulSet "redis-pod" is invalid: spec: Forbidden: updates to statefulset spec for fields other than 'replicas', 'template', and 'updateStrategy' are forbidden

(2)生成的pod编号是从0开始的,依次是 redis-pod-0, redis-pod-1, redis-pod-2, redis-pod-3, redis-pod-4, redis-pod-5.

当删除一个pod时则会自动生成一个同名的pod.当然了,新生成的pod ip会发生变化

(3)查看pod对应的DNS域名

# for x in $(seq 0 5); do kubectl exec redis-pod-$x -n cluster-redis -- hostname -f; done

redis-pod-0.redis-cluster-service.cluster-redis.svc.cluster.local

redis-pod-1.redis-cluster-service.cluster-redis.svc.cluster.local

redis-pod-2.redis-cluster-service.cluster-redis.svc.cluster.local

redis-pod-3.redis-cluster-service.cluster-redis.svc.cluster.local

redis-pod-4.redis-cluster-service.cluster-redis.svc.cluster.local

redis-pod-5.redis-cluster-service.cluster-redis.svc.cluster.local

8.使用

项目中代码连接redis集群的话,使用的参数是: ClusterIP:6379

k8s中安装redis6集群的相关教程结束。