使用redis-exporter對redis集羣進行性能監控

redis集羣有三種模式，主從，sentinel（哨兵），cluster。關於這三種模式的介紹和部署，其它文章也有相關介紹，本人蔘考了他人的博客完成了三種模式的部署，博客地址：https://blog.csdn.net/miss1181248983/article/details/90056960。非常感謝此大佬的分享
目前redis使用主從模式，集羣信息如下：

[root@master redis]# redis-cli -h 192.168.122.7 -a 123456 info replication
Warning: Using a password with '-a' or '-u' option on the command line interface may not be safe.
# Replication
role:master
connected_slaves:2
slave0:ip=192.168.122.8,port=6379,state=online,offset=28,lag=0
slave1:ip=192.168.122.9,port=6379,state=online,offset=28,lag=1
master_replid:950b3c2b60b56148577e566ac53af00911527c4e
master_replid2:0000000000000000000000000000000000000000
master_repl_offset:28
second_repl_offset:-1
repl_backlog_active:1
repl_backlog_size:1048576
repl_backlog_first_byte_offset:1
repl_backlog_histlen:28
[root@master redis]#
[root@master ~]#

以上信息可以看出，此時集羣爲主從模式，192.168.122.7爲master節點，192.168.122.8和192.168.122.9爲slave節點。關於redis的具體操作這兒不做闡述，相關文章很多。下面步驟就是如何利用redis-exporter收集redis性能監控。

1 創建redis-exporter

使用下面的yaml文件創建一個redis-exporter的Deployment‘

[root@master redis-local]# cat redis-exporter.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
  name: redis-exporter001
  namespace: monitoring
  labels:
    k8s-app: redis-exporter001
spec:
  selector:
    matchLabels:
      k8s-app: redis-exporter001
  template:
    metadata:
      labels:
        k8s-app: redis-exporter001
    spec:
      containers:
      - name: redis-exporter001
        image: oliver006/redis_exporter:latest
        args: ["-redis.addr", "redis://192.168.122.7:6379", "-redis.password", "123456"]
        ports:
        - containerPort: 9121
          name: http
[root@master redis-local]# kubectl apply -f redis-exporter.yaml
deployment.apps/redis-exporter001 created
[root@master redis-local]# kubectl logs -f -n monitoring redis-exporter001-5dc78569b5-5lmtf
time="2020-05-26T02:49:21Z" level=info msg="Redis Metrics Exporter v1.6.1    build date: 2020-05-13-02:16:26    sha1: 34c779deb5f9a1e023166b5cf3717d53cdc97053    Go: go1.14.2    GOOS: linux    GOARCH: amd64"
time="2020-05-26T02:49:21Z" level=info msg="Providing metrics at :9121/metrics"

此時redis-exporter創建完成

2 創建redis-exporter服務

使用下面的yaml文件創建redis-exporter服務

[root@master redis-local]# cat redis-exporter-service.yaml
apiVersion: v1
kind: Service
metadata:
  labels:
    k8s-app: redis-exporter001
  name: redis-exporter001
  namespace: monitoring
spec:
  type: NodePort
  ports:
  - name: http
    port: 9121
    nodePort: 30019
    targetPort: http
  selector:
    k8s-app: redis-exporter001
[root@master redis-local]# kubectl apply -f redis-exporter-service.yaml
service/redis-exporter001 created
[root@master redis-local]# kubectl describe svc -n monitoring redis-exporter001
Name:                     redis-exporter001
Namespace:                monitoring
Labels:                   k8s-app=redis-exporter001
Annotations:              kubectl.kubernetes.io/last-applied-configuration:
                            {"apiVersion":"v1","kind":"Service","metadata":{"annotations":{},"labels":{"k8s-app":"redis-exporter001"},"name":"redis-exporter001","name...
Selector:                 k8s-app=redis-exporter001
Type:                     NodePort
IP:                       10.10.126.113
Port:                     http  9121/TCP
TargetPort:               http/TCP
NodePort:                 http  30019/TCP
Endpoints:                10.124.187.225:9121
Session Affinity:         None
External Traffic Policy:  Cluster
Events:                   <none>
[root@master redis-local]#

使用nodeip:nodeport進行訪問

此時redis的metrics部署完成

3創建redis-exporter-serverMonitor

使用下面的文件創建redis-exporter-serverMonitor

[root@master redis-local]# cat redis-exporter-serviceMonitor.yaml
apiVersion: monitoring.coreos.com/v1
kind: ServiceMonitor
metadata:
  labels:
    k8s-app: redis-exporter001
  name: redis-exporter001
  namespace: monitoring
spec:
  endpoints:
  - interval: 30s
    port: http
  jobLabel: k8s-app
  selector:
    matchLabels:
      k8s-app: redis-exporter001
[root@master redis-local]# kubectl apply -f redis-exporter-serviceMonitor.yaml
servicemonitor.monitoring.coreos.com/redis-exporter001 created
[root@master redis-local]# kubectl describe servicemonitors.monitoring.coreos.com -n monitoring redis-exporter001
Name:         redis-exporter001
Namespace:    monitoring
Labels:       k8s-app=redis-exporter001
Annotations:  kubectl.kubernetes.io/last-applied-configuration:
                {"apiVersion":"monitoring.coreos.com/v1","kind":"ServiceMonitor","metadata":{"annotations":{},"labels":{"k8s-app":"redis-exporter001"},"na...
API Version:  monitoring.coreos.com/v1
Kind:         ServiceMonitor
Metadata:
  Creation Timestamp:  2020-05-26T02:53:46Z
  Generation:          1
  Resource Version:    574501
  Self Link:           /apis/monitoring.coreos.com/v1/namespaces/monitoring/servicemonitors/redis-exporter001
  UID:                 065dc9cc-a872-4630-a387-f4b11193aab6
Spec:
  Endpoints:
    Interval:  30s
    Port:      http
  Job Label:   k8s-app
  Selector:
    Match Labels:
      k8s-app:  redis-exporter001
Events:         <none>

最後創建一個serviceaccount

[root@master yaml]# kubectl create serviceaccount -n monitoring redis-exporter001
serviceaccount/redis-exporter001 created

以上部署完成後，在prometheus頁面中查看對應的targets

4 grafana面板創建dashboard來監控redis集羣

點擊Import

監控效果如上圖

5 redis寫入數據進行測試

[root@master redis-local]# cat 123.txt |redis-cli -h 192.168.122.7 -a 123456

批量插入數據 ，查看grafana監控

此時redis性能使用率已經在上升。
關於如何監控redis集羣——主從模式到此完成。關於如何監控另外兩種模式，後續文檔會繼續更新。