helm部署redis主从和哨兵模式
1. helm部署redis主从节点
1.1. helm部署redis条件
a. 准备storageClass服务作为永久存储，名称为nfs
b. helm Client："v2.13.1"， Server:"v2.13.1"
c. 仓库源为http://mirror.azure.cn/kubernetes/charts/
d. 确认仓库里有chart，redis chart版本号为10.5.7
[root@ops1 redis]# helm version
Client: &version.Version{SemVer:"v2.13.1", GitCommit:"618447cbf203d147601b4b9bd7f8c37a5d39fbb4", GitTreeState:"clean"}
Server: &version.Version{SemVer:"v2.13.1", GitCommit:"0e7f3b6637f7af8fcfddb3d2941fcc7cbebb0085", GitTreeState:"clean"}
#添加新的远程仓库
[root@ops1 redis]# helm repo add stable http://mirror.azure.cn/kubernetes/charts/
[root@ops1 redis]# helm repo list
NAME URL
local http://127.0.0.1:8879/charts
stable http://mirror.azure.cn/kubernetes/charts/
incubator http://mirror.azure.cn/kubernetes/charts-incubator/
[root@ops1 redis]# helm repo update
[root@ops1 redis]# helm search redis
NAME CHART VERSION APP VERSION DESCRIPTION
stable/redis 10.5.7 5.0.7 DEPRECATED Open source, advanced key-value store. It is o...
stable/redis-ha 4.4.3 5.0.6 Highly available Kubernetes implementation of Redis
[root@ops1 redis]# helm fetch stable/redis # 拉取代码到本地，如有问题，查看编写规则
1.2.部署redis一主二从节点
[root@ops1 redis]# cat <<EOF> my-values.yaml
password: "redis1"
master:
service:
type: NodePort
persistence:
storageClass: "nfs"
slave:
service:
type: NodePort
persistence:
storageClass: "nfs"
EOF
[root@ops1 redis]# helm install --name redis-service -f my-values.yaml --namespace=course stable/redis
NAME: redis-service
LAST DEPLOYED: Sun Mar 29 12:17:33 2020
NAMESPACE: course
STATUS: DEPLOYED
[root@ops1 test]# kubectl get svc | grep redis-service
redis-service-headless ClusterIP None <none> 6379/TCP 3m45s
redis-service-master NodePort 10.96.140.206 <none> 6379:20986/TCP 3m45s
redis-service-slave NodePort 10.96.98.91 <none> 6379:35090/TCP 3m45s
#有时候我们创建的服务不想走 负载均衡，想直接通过 pod-ip 链接后端， 怎么办呢， 使用headless service接可以解决。
所以在微服务相关场景 如果需要直通pod的时候 我们就可以使用headless service 绕过 k8s的转发机制，直接访问pod了。
[root@ops1 redis]# redis-cli -p 20986 -a redis1
127.0.0.1:20986>
2. 部署redis 高可用哨兵模式
本文中的Redis高可用方案采用Sentinel(哨兵)模式，在集群出现故障的时候自动进行故障转移，保证集群的可用性。
2.1.确认仓库里有chart，redis chart版本号为10.5.7
a. 准备storageClass服务作为永久存储，名称为nfs
b. helm Client："v2.13.1"， Server:"v2.13.1"
c. 仓库源为http://mirror.azure.cn/kubernetes/charts/
d. 确认仓库里有chart，redis chart版本号为4.4.3
[root@ops1 redis]# helm repo list
NAME URL
local http://127.0.0.1:8879/charts
stable http://mirror.azure.cn/kubernetes/charts/
incubator http://mirror.azure.cn/kubernetes/charts-incubator/
[root@ops1 redis]# helm repo update
[root@ops1 redis]# helm search redis
NAME CHART VERSION APP VERSION DESCRIPTION
stable/redis 10.5.7 5.0.7 DEPRECATED Open source, advanced key-value store. It is o...
stable/redis-ha 4.4.3 5.0.6 Highly available Kubernetes implementation of Redis
[root@ops1 redis]# helm fetch stable/redis-ha # 拉取代码到本地，如有问题，查看编写规则
2.2. 部署redis 高可用
[root@ops1 redis]# cat <<EOF> my-values.yaml
auth: true
redisPassword: "redis1"
hardAntiAffinity: false #(仅限当replicas > worker node 节点数时修改)
haproxy:
service:
type: NodePort
persistence:
storageClass: "nfs"
persistentVolume:
storageClass: "nfs"
EOF
[root@ops1 redis-ha]# helm install --name redis-ha -f my-values.yaml --namespace=course stable/redis-ha
NAME: redis-ha
LAST DEPLOYED: Sun Mar 29 14:40:53 2020
NAMESPACE: course
STATUS: DEPLOYED
RESOURCES:
==> v1/ConfigMap
NAME DATA AGE
redis-ha-configmap 4 2s
==> v1/Pod(related)
NAME READY STATUS RESTARTS AGE
redis-ha-server-0 0/2 Init:0/1 0 1s
==> v1/Role
NAME AGE
redis-ha 2s
==> v1/RoleBinding
NAME AGE
redis-ha 1s
==> v1/Secret
NAME TYPE DATA AGE
redis-ha Opaque 1 3s
==> v1/Service
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
redis-ha ClusterIP None <none> 6379/TCP,26379/TCP 1s
redis-ha-announce-0 ClusterIP 10.96.208.41 <none> 6379/TCP,26379/TCP 1s
redis-ha-announce-1 ClusterIP 10.96.59.19 <none> 6379/TCP,26379/TCP 1s
redis-ha-announce-2 ClusterIP 10.96.95.191 <none> 6379/TCP,26379/TCP 1s
==> v1/ServiceAccount
NAME SECRETS AGE
redis-ha 1 2s
==> v1/StatefulSet
NAME READY AGE
redis-ha-server 0/3 1s
NOTES:
Redis can be accessed via port 6379 and Sentinel can be accessed via port 26379 on the following DNS name from within your cluster:
redis-ha.course.svc.cluster.local
To connect to your Redis server:
1. To retrieve the redis password:
echo $(kubectl get secret redis-ha -o "jsonpath={.data['auth']}" | base64 --decode)
2. Connect to the Redis master pod that you can use as a client. By default the redis-ha-server-0 pod is configured as the master:
kubectl exec -it redis-ha-server-0 sh -n course
3. Connect using the Redis CLI (inside container):
redis-cli -a <REDIS-PASS-FROM-SECRET>
[root@ops1 redis-ha]# kubectl get all
NAME READY STATUS RESTARTS AGE
pod/redis-ha-server-0 2/2 Running 0 74s
pod/redis-ha-server-1 2/2 Running 0 63s
pod/redis-ha-server-2 2/2 Running 0 53s
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
service/redis-ha ClusterIP None <none> 6379/TCP,26379/TCP 74s
service/redis-ha-announce-0 ClusterIP 10.96.162.123 <none> 6379/TCP,26379/TCP 74s
service/redis-ha-announce-1 ClusterIP 10.96.198.108 <none> 6379/TCP,26379/TCP 74s
service/redis-ha-announce-2 ClusterIP 10.96.96.97 <none> 6379/TCP,26379/TCP 74s
NAME READY AGE
statefulset.apps/redis-ha-server 3/3 74s
2.3.如果需要暴露给外部使用则需要再部署一个 NodePort Service
[root@ops1 redis]# cat <<EOF> redis-service.yaml
apiVersion: v1
kind: Service
metadata:
name: redis-ha-service #名称：随意
labels:
app: redis-ha #部署的 redis-ha 名称
spec:
ports:
- name: redis-ha #部署的 redis-ha 名称
protocol: "TCP" #TCP 协议
port: 26379
targetPort: 6379
nodePort: 30379 #此为外部连接k8s redis-ha 服务的端口
selector:
statefulset.kubernetes.io/pod-name: redis-ha-server-0
type: NodePort
EOF
[root@ops1 redis-ha]# kubectl apply -f redis-service.yaml
[root@ops1 redis-ha]# kubectl get svc | grep redis-ha-service
redis-ha-service NodePort 10.96.152.166 <none> 26379:30379/TCP 65s
[root@ops1 redis-ha]# redis-cli -p 30379 -a redis1
127.0.0.1:30379> info
