K8S上部署redis集群

Redis 介紹

Redis 代表REmote DIctionary Server是一種開源的內存中數據存儲，通常用作數據庫，緩存或消息代理。它可以存儲和操作高級數據類型，例如列表，地圖，集合和排序集合。由於Redis接受多種格式的密鑰，因此可以在服務器上執行操作，從而減少了客戶端的工作量。它僅將磁盤用於持久性，而將數據庫完全保存在內存中。Redis是一種流行的數據存儲解決方案，並被GitHub，Pinterest，Snapchat，Twitter，StackOverflow，Flickr等技術巨頭所使用。

為什么使用 Redis

• 它的速度非常快。它是用 ANSI C 編寫的，並且可以在 POSIX 系統上運行，例如 Linux，Mac OS X 和 Solaris。
• Redis 通常被排名為最流行的鍵/值數據庫和最流行的與容器一起使用的 NoSQL 數據庫。
• 其緩存解決方案減少了對雲數據庫后端的調用次數。
• 應用程序可以通過其客戶端 API 庫對其進行訪問。
• 所有流行的編程語言都支持 Redis。
• 它是開源且穩定的。

什么是 Redis 集群

Redis Cluster 是一組 Redis 實例，旨在通過對數據庫進行分區來擴展數據庫，從而使其更具彈性。群集中的每個成員（無論是主副本還是輔助副本）都管理哈希槽的子集。如果主機無法訪問，則其從機將升級為主機。在由三個主節點組成的最小 Redis 群集中，每個主節點都有一個從節點（以實現最小的故障轉移），每個主節點都分配有一個介於 0 到 16,383 之間的哈希槽范圍。節點 A 包含從 0 到 5000 的哈希槽，節點 B 從 5001 到 10000，節點 C 從 10001 到 16383。群集內部的通信是通過內部總線進行的，使用協議傳播有關群集的信息或發現新節點。

在 Kubernetes 中部署 Redis 集群

在Kubernetes中部署Redis集群面臨挑戰，因為每個 Redis 實例都依賴於一個配置文件，該文件可以跟蹤其他集群實例及其角色。為此，我們需要結合使用Kubernetes StatefulSets和PersistentVolumes。

克隆部署文件

git clone https://github.com/llmgo/redis-sts.git  

創建 statefulset 類型資源

[root@node01 redis-sts]# cat redis-sts.yml 
---
apiVersion: v1
kind: ConfigMap
metadata:
  name: redis-cluster
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-cluster
spec:
  serviceName: redis-cluster
  replicas: 6
  selector:
    matchLabels:
      app: redis-cluster
  template:
    metadata:
      labels:
        app: redis-cluster
    spec:
      containers:
      - name: redis
        image: redis:5.0.5-alpine
        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
          defaultMode: 0755
  volumeClaimTemplates:
  - metadata:
      name: data
    spec:
      accessModes: [ "ReadWriteOnce" ]
      resources:
        requests:
          storage: 5Gi
      storageClassName: standard
$ kubectl apply -f redis-sts.yml
configmap/redis-cluster created
statefulset.apps/redis-cluster created
$ kubectl get pods -l app=redis-cluster
NAME              READY   STATUS    RESTARTS   AGE
redis-cluster-0   1/1     Running   0          53s
redis-cluster-1   1/1     Running   0          49s
redis-cluster-2   1/1     Running   0          46s
redis-cluster-3   1/1     Running   0          42s
redis-cluster-4   1/1     Running   0          38s
redis-cluster-5   1/1     Running   0          34s

創建 service

[root@node01 redis-sts]# cat redis-svc.yml   
---  
apiVersion: v1  
kind: Service  
metadata:  
  name: redis-cluster  
spec:  
  type: ClusterIP  
  clusterIP: 10.96.0.100  
  ports:  
  - port: 6379  
    targetPort: 6379  
    name: client  
  - port: 16379  
    targetPort: 16379  
    name: gossip  
  selector:  
    app: redis-cluster  
$ kubectl apply -f redis-svc.yml  
service/redis-cluster created  
$ kubectl get svc redis-cluster  
NAME            TYPE        CLUSTER-IP   EXTERNAL-IP   PORT(S)              AGE  
redis-cluster   ClusterIP   10.96.0.100   <none>        6379/TCP,16379/TCP   35s  

初始化 redis cluster

“

下一步是形成Redis集群。為此，我們運行以下命令並鍵入yes以接受配置。前三個節點成為主節點，后三個節點成為從節點。

$ kubectl exec -it redis-cluster-0 -- redis-cli --cluster create --cluster-replicas 1 $(kubectl get pods -l app=redis-cluster -o jsonpath='{range.items[*]}{.status.podIP}:6379 ')  
  
>>> Performing hash slots allocation on 6 nodes...  
Master[0] -> Slots 0 - 5460  
Master[1] -> Slots 5461 - 10922  
Master[2] -> Slots 10923 - 16383  
Adding replica 10.244.2.11:6379 to 10.244.9.19:6379  
Adding replica 10.244.9.20:6379 to 10.244.6.10:6379  
Adding replica 10.244.8.15:6379 to 10.244.7.8:6379  
M: 00721c43db194c8f2cacbafd01fd2be6a2fede28 10.244.9.19:6379  
   slots:[0-5460] (5461 slots) master  
M: 9c36053912dec8cb20a599bda202a654f241484f 10.244.6.10:6379  
   slots:[5461-10922] (5462 slots) master  
M: 2850f24ea6367de58fb50e632fc56fe4ba5ef016 10.244.7.8:6379  
   slots:[10923-16383] (5461 slots) master  
S: 554a58762e3dce23ca5a75886d0ccebd2d582502 10.244.8.15:6379  
   replicates 2850f24ea6367de58fb50e632fc56fe4ba5ef016  
S: 20028fd0b79045489824eda71fac9898f17af896 10.244.2.11:6379  
   replicates 00721c43db194c8f2cacbafd01fd2be6a2fede28  
S: 87e8987e314e4e5d4736e5818651abc1ed6ddcd9 10.244.9.20:6379  
   replicates 9c36053912dec8cb20a599bda202a654f241484f  
Can I set the above configuration? (type 'yes' to accept): yes  
>>> 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 10.244.9.19:6379)  
M: 00721c43db194c8f2cacbafd01fd2be6a2fede28 10.244.9.19:6379  
   slots:[0-5460] (5461 slots) master  
   1 additional replica(s)  
M: 9c36053912dec8cb20a599bda202a654f241484f 10.244.6.10:6379  
   slots:[5461-10922] (5462 slots) master  
   1 additional replica(s)  
S: 87e8987e314e4e5d4736e5818651abc1ed6ddcd9 10.244.9.20:6379  
   slots: (0 slots) slave  
   replicates 9c36053912dec8cb20a599bda202a654f241484f  
S: 554a58762e3dce23ca5a75886d0ccebd2d582502 10.244.8.15:6379  
   slots: (0 slots) slave  
   replicates 2850f24ea6367de58fb50e632fc56fe4ba5ef016  
S: 20028fd0b79045489824eda71fac9898f17af896 10.244.2.11:6379  
   slots: (0 slots) slave  
   replicates 00721c43db194c8f2cacbafd01fd2be6a2fede28  
M: 2850f24ea6367de58fb50e632fc56fe4ba5ef016 10.244.7.8:6379  
   slots:[10923-16383] (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.  

驗證集群

[root@node01 redis-sts]# kubectl exec -it redis-cluster-0 -- redis-cli cluster info  
cluster_state:ok  
cluster_slots_assigned:16384  
cluster_slots_ok:16384  
cluster_slots_pfail:0  
cluster_slots_fail:0  
cluster_known_nodes:6  
cluster_size:3  
cluster_current_epoch:6  
cluster_my_epoch:1  
cluster_stats_messages_ping_sent:16  
cluster_stats_messages_pong_sent:22  
cluster_stats_messages_sent:38  
cluster_stats_messages_ping_received:17  
cluster_stats_messages_pong_received:16  
cluster_stats_messages_meet_received:5  
cluster_stats_messages_received:38  
[root@node01 redis-sts]# for x in $(seq 0 5); do echo "redis-cluster-$x"; kubectl exec redis-cluster-$x -- redis-cli role; echo; done  
redis-cluster-0  
master  
14  
10.244.2.11  
6379  
14  
  
redis-cluster-1  
master  
28  
10.244.9.20  
6379  
28  
  
redis-cluster-2  
master  
28  
10.244.8.15  
6379  
28  
  
redis-cluster-3  
slave  
10.244.7.8  
6379  
connected  
28  
  
redis-cluster-4  
slave  
10.244.9.19  
6379  
connected  
14  
  
redis-cluster-5  
slave  
10.244.6.10  
6379  
connected  
28  

測試集群

我們想使用集群，然后模擬節點的故障。對於前一項任務，我們將部署一個簡單的 Python 應用程序，而對於后者，我們將刪除一個節點並觀察集群行為。

部署點擊計數器應用

我們將一個簡單的應用程序部署到集群中，並在其前面放置一個負載平衡器。此應用程序的目的是在將計數器值作為 HTTP 響應返回之前，增加計數器並將其存儲在 Redis 集群中。

$ kubectl apply -f app-deployment-service.yml  
service/hit-counter-lb created  
deployment.apps/hit-counter-app created  

在此過程中，如果我們繼續加載頁面，計數器將繼續增加，並且在刪除Pod之后，我們看到沒有數據丟失。

$  curl `kubectl get svc hit-counter-lb -o json|jq -r .spec.clusterIP`  
I have been hit 20 times since deployment.  
$  curl `kubectl get svc hit-counter-lb -o json|jq -r .spec.clusterIP`  
I have been hit 21 times since deployment.  
$ curl `kubectl get svc hit-counter-lb -o json|jq -r .spec.clusterIP`  
I have been hit 22 times since deployment.  
$ kubectl delete pods redis-cluster-0  
pod "redis-cluster-0" deleted  
$ kubectl delete pods redis-cluster-1  
pod "redis-cluster-1" deleted  
$  curl `kubectl get svc hit-counter-lb -o json|jq -r .spec.clusterIP`  
I have been hit 23 times since deployment.

原文鏈接：https://mp.weixin.qq.com/s/xklgkK8ZkUYzFLR_CGrhSA