目录

Kubernetes 的安装和使用（一）
Kubernetes 的安装和使用（二）
Kubernetes 的安装和使用（三）

k8s 使用 Service

在前面的例子中我们使用 port-forward 进行端口转发，这样会存在两个问题：

  1. pod 重启后 ip 地址发生了变化怎么办
  2. 如何进行负载均衡

k8s 的 Service 就是用来解决这个问题的，它包含 ClusterIP、NodePort 和 Headless 等模块。

ClusterIP

ClusterIP 就是将多个 pod 用一个 ip 进行访问的服务，这个 ip 只能在集群内访问，我们创建如下的程序

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
package main

import (
	"fmt"
	"io"
	"net/http"
	"os"
)

func main() {
	http.HandleFunc("/", func(w http.ResponseWriter, r *http.Request) {
		host, _ := os.Hostname()
		io.WriteString(w, fmt.Sprintf("[v3] Hello, Kubernetes!, From host: %s\n", host))
	})
	http.ListenAndServe(":3000", nil)
}

我们将如上程序打包成镜像

docker build . -t derobukal/hellok8s:v3_hostname
docker push derobukal/hellok8s:v3_hostname

然后启动 k8s 的 deployment，可以得到 3 个运行的 pod

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
apiVersion: apps/v1
kind: Deployment
metadata:
  # deployment 唯一名称
  name: hellok8s-go-http
spec:
  replicas: 3 # 副本数量
  selector:
    matchLabels:
      app: hellok8s # 管理template下所有 app=hellok8s的pod，（要求和template.metadata.labels完全一致！！！否则无法部署deployment）
  template: # template 定义一组容器
    metadata:
      labels:
        app: hellok8s
    spec:
      containers:
        - image: derobukal/hellok8s:v3_hostname
          name: hellok8s

之后我们创建 service-clusterip.yaml

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
apiVersion: v1
kind: Service
metadata:
  name: service-hellok8s-clusterip
spec:
  type: ClusterIP  # 这行是默认的，可省略
#  sessionAffinity: ClientIP # or None, 设置会话亲和性（ClientIP表示同一客户端ip的请求会路由到同个Pod）
#  sessionAffinityConfig:
#    clientIP:
#      timeoutSeconds: 3600 # 范围 0~86400，默认10800（3h）
  selector:
    app: hellok8s  # 通过selector关联pod组
  ports:
    - port: 3000 # service端口
      targetPort: 3000 # 后端pod端口

随后启动 service，并查看 service 所生成的 clusterIp 的值，以及 clusterIp 后面的 endpoints 的详细信息

1
2
3
4
5
6
7
8
9
10
~ kc apply -f service-clusterip.yaml 
service/service-hellok8s-clusterip created
~ kc get svc                        
NAME                         TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)    AGE
kubernetes                   ClusterIP   10.96.0.1        <none>        443/TCP    7d1h
service-hellok8s-clusterip   ClusterIP   10.111.240.227   <none>        3000/TCP   10s
~ kc get endpoints
NAME                         ENDPOINTS                                            AGE
kubernetes                   192.168.49.2:8443                                    7d1h
service-hellok8s-clusterip   10.244.0.67:3000,10.244.0.68:3000,10.244.0.69:3000   16s

有了 service 之后，我们就可以很方便的通过 clusterIp 的地址访问服务了。因为我们使用的是 minikube，为了能正常访问 clusterIp，我们先创建一个 nginx 的 pod

1
2
3
4
5
6
7
8
9
10
apiVersion: v1
kind: Pod
metadata:
  name: nginx
  labels:
    app: nginx
spec:
  containers:
    - name: nginx-container
      image: nginx

之后我们进入 nginx 的 pod 测试 clusterIp

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
~ kc apply -f pod_nginx.yaml        
pod/nginx created
~ kc get pods
NAME                                READY   STATUS    RESTARTS   AGE
hellok8s-go-http-6f5d68bc64-6xrdx   1/1     Running   0          23m
hellok8s-go-http-6f5d68bc64-b7wq2   1/1     Running   0          23m
hellok8s-go-http-6f5d68bc64-rk59k   1/1     Running   0          23m
nginx                               1/1     Running   0          98s
~ kc exec nginx -it -- bash                            
root@nginx:/# curl 10.111.240.227:3000
[v3] Hello, Kubernetes!, From host: hellok8s-go-http-6f5d68bc64-b7wq2
root@nginx:/# curl 10.111.240.227:3000
[v3] Hello, Kubernetes!, From host: hellok8s-go-http-6f5d68bc64-rk59k
root@nginx:/# curl 10.111.240.227:3000
[v3] Hello, Kubernetes!, From host: hellok8s-go-http-6f5d68bc64-rk59k
root@nginx:/# curl 10.111.240.227:3000
[v3] Hello, Kubernetes!, From host: hellok8s-go-http-6f5d68bc64-b7wq2
root@nginx:/# curl 10.111.240.227:3000
[v3] Hello, Kubernetes!, From host: hellok8s-go-http-6f5d68bc64-6xrdx
root@nginx:/# curl 10.111.240.227:3000
[v3] Hello, Kubernetes!, From host: hellok8s-go-http-6f5d68bc64-b7wq2
root@nginx:/# curl 10.111.240.227:3000
[v3] Hello, Kubernetes!, From host: hellok8s-go-http-6f5d68bc64-rk59k
root@nginx:/#

上面通过 curl 测试 service，可以看到每次访问的后端 pod 都是不一样的。接下来我们增加 pod 的数量，可以看到 endpoint 的数量也发生了变化，同时测试请求也打到了新的 pod 上面

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
~ kc scale deployment/hellok8s-go-http --replicas=10
deployment.apps/hellok8s-go-http scaled
~ kc get endpoints                                                             
NAME                         ENDPOINTS                                                        AGE
kubernetes                   192.168.49.2:8443                                                7d1h
service-hellok8s-clusterip   10.244.0.67:3000,10.244.0.68:3000,10.244.0.69:3000 + 7 more...   22m
~ kc get pods     
NAME                                READY   STATUS    RESTARTS   AGE
hellok8s-go-http-6f5d68bc64-5lxvs   1/1     Running   0          27s
hellok8s-go-http-6f5d68bc64-6xrdx   1/1     Running   0          29m
hellok8s-go-http-6f5d68bc64-b7wq2   1/1     Running   0          29m
hellok8s-go-http-6f5d68bc64-cl56f   1/1     Running   0          27s
hellok8s-go-http-6f5d68bc64-gbn9v   1/1     Running   0          27s
hellok8s-go-http-6f5d68bc64-k7db4   1/1     Running   0          27s
hellok8s-go-http-6f5d68bc64-m4h5s   1/1     Running   0          27s
hellok8s-go-http-6f5d68bc64-rk59k   1/1     Running   0          29m
hellok8s-go-http-6f5d68bc64-whpht   1/1     Running   0          27s
hellok8s-go-http-6f5d68bc64-xnvk2   1/1     Running   0          27s
nginx                               1/1     Running   0          8m13s
~ kc exec nginx -it -- bash
root@nginx:/# curl 10.111.240.227:3000                                 
[v3] Hello, Kubernetes!, From host: hellok8s-go-http-6f5d68bc64-6xrdx
root@nginx:/# curl 10.111.240.227:3000
[v3] Hello, Kubernetes!, From host: hellok8s-go-http-6f5d68bc64-b7wq2
root@nginx:/# curl 10.111.240.227:3000
[v3] Hello, Kubernetes!, From host: hellok8s-go-http-6f5d68bc64-m4h5s
root@nginx:/# curl 10.111.240.227:3000
[v3] Hello, Kubernetes!, From host: hellok8s-go-http-6f5d68bc64-whpht
root@nginx:/# curl 10.111.240.227:3000
[v3] Hello, Kubernetes!, From host: hellok8s-go-http-6f5d68bc64-whpht
root@nginx:/# curl 10.111.240.227:3000
[v3] Hello, Kubernetes!, From host: hellok8s-go-http-6f5d68bc64-cl56f
root@nginx:/# curl 10.111.240.227:3000
[v3] Hello, Kubernetes!, From host: hellok8s-go-http-6f5d68bc64-xnvk2
root@nginx:/# curl 10.111.240.227:3000
[v3] Hello, Kubernetes!, From host: hellok8s-go-http-6f5d68bc64-cl56f
root@nginx:/# curl 10.111.240.227:3000
[v3] Hello, Kubernetes!, From host: hellok8s-go-http-6f5d68bc64-cl56f
root@nginx:/# curl 10.111.240.227:3000
[v3] Hello, Kubernetes!, From host: hellok8s-go-http-6f5d68bc64-m4h5s
root@nginx:/# curl 10.111.240.227:3000
[v3] Hello, Kubernetes!, From host: hellok8s-go-http-6f5d68bc64-rk59k
root@nginx:/# curl 10.111.240.227:3000
[v3] Hello, Kubernetes!, From host: hellok8s-go-http-6f5d68bc64-whpht
root@nginx:/# curl 10.111.240.227:3000
[v3] Hello, Kubernetes!, From host: hellok8s-go-http-6f5d68bc64-k7db4
root@nginx:/#

NodePort

clusterIp 只能在集群内部进行访问，NodePort 在 clusterIp 的基础上，还支持了通过 k8s 集群的节点进行访问。有如下的配置

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
apiVersion: v1
kind: Service
metadata:
  name: service-hellok8s-nodeport
spec:
  type: NodePort
  selector:
    app: hellok8s
  ports:
    - port: 3000  # pod端口
      nodePort: 30000  # 节点固定端口。在NodePort类型中，k8s要求在 30000-32767 范围内，否则apply报错
    # 若需要暴露多个端口，则按下面形式
#    - name: http
#      protocol: TCP
#      port: 80
#      targetPort: 9376
#    - name: https
#      protocol: TCP
#      port: 443
#      targetPort: 9377

之后启动 nodeport 的服务，就可以在集群的节点上访问服务了

1
2
3
4
5
6
7
~ kc apply -f service-nodeport.yaml 
service/service-hellok8s-nodeport created
~ kc get svc                       
NAME                         TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)          AGE
kubernetes                   ClusterIP   10.96.0.1        <none>        443/TCP          7d1h
service-hellok8s-clusterip   ClusterIP   10.111.240.227   <none>        3000/TCP         28m
service-hellok8s-nodeport    NodePort    10.106.138.169   <none>        3000:30000/TCP   17s

因为使用 minikube 创建服务，而非创建了 k8s 的集群，因此暂时不太方便测试该功能。