k8s 重启策略、健康检查、环境变量、初始化容器

- Pod基本概念
- - Pod特点
  - Pod存在意义
- Pod资源共享实现机制
- 重启策略
- 健康检查
- 环境变量
- init Container(初始化容器)
- haproxy

Pod基本概念

Pod是Kubernetes创建和管理的最小单元，一个Pod由一个容器或多个容器组成，这些容器共享存储、网络。

Pod特点

一个Pod可以理解为是一个应用实例，提供服务.
Pod中容器始终部署在一个Node上
Pod中容器共享网络、存储资源
Kubernetes直接管理Pod，而不是容器

Pod存在意义

Pod主要用法:

运行单个容器:最常见的用法，在这种情况下，可以将Pod看做是单个容器的抽象封装
运行多个容器:封装多个紧密耦合且需要共享资源的应用程序

如果有这些需求，你可以运行多个容器:

两个应用之间发生文件交互
两个应用需要通过127.0.0.1或者socket通信
两个应用需要发生频繁的调用

Pod资源共享实现机制

重启策略

Always：当容器终止退出后，总是重启容器，默认策略（总是重启）
OnFailure：当容器异常退出(退出状态码非0)时，才重启容器（非正常退出，比如stop、kill）
Never：当容器终止退出，从不重启容器（永不重启）

[root@master manifest]# cat test.yml 
apiVersion: v1
kind: Pod
metadata: 
  name: web
spec: 
  containers:		//一个pod多个容器
  - name: nginx
    image: nginx
    imagePullPolicy: IfNotPresent
  - name: test
    image: busybox
    imagePullPolicy: IfNotPresent
    command: ["bin/sh","-c","sleep 450"]
  restartPolicy: Never		#默认的话不用修改，改为never后停止容器不会重启

[root@master manifest]# kubectl apply -f test.yml 		//创建pod
pod/web created
[root@master manifest]# kubectl get pod 
NAME   READY   STATUS    RESTARTS   AGE
web    2/2     Running   0          92s
[root@master manifest]# kubectl get pod -o wide -w		//在node2上停止其中一个，发现并不会重启
NAME   READY   STATUS   RESTARTS   AGE     IP            NODE                NOMINATED NODE   READINESS GATES
web    1/2     NotReady    0          4m51s   10.244.2.10   node2.example.com

健康检查

livenessProbe(存活检查)︰如果检查失败，将杀死容器，根据Pod的- restartPolicy来操作
readinessProbe(就绪检查)︰如果检查失败，Kubernetes会把Pod从service endpoints中剔除

支持的检查方式:

httpGet:发送HTTP请求，返回200-400范围状态码为成功
exec: 执行hell命令返回状态码是0为成功
tcpSocket:发起TCP Socket建立成功
与重启策略相结合使用

重启策略+健康检查

//端口探测
[root@master ~]# cat test.yml 
---
apiVersion: v1
kind: Pod
metadata:
  name: web
spec:
  containers:
  - name: nginx
    image: nginx
    imagePullPolicy: IfNotPresent
    ports:
    - containerPort: 80
      hostPort: 80
    livenessProbe:
      tcpSocket:
        port: 80
      initialDelaySeconds: 20 		#启动容器后多少秒健康检查
      periodSeconds: 10 			#以后间隔多少秒检查一次
    readinessProbe:
      httpGet:
        port: 80
      initialDelaySeconds: 20
      periodSeconds: 10


[root@master ~]# kubectl apply -f test.yml 
pod/web created

//查看pod,发现在进行初始化
[root@master ~]# kubectl get pod
NAME   READY（就绪状态）   STATUS（存活状态）    RESTARTS   AGE
web    0/1     Running   0          18s

//等待一定时间后会进入运行
[root@master ~]# kubectl get pod
NAME   READY   STATUS    RESTARTS   AGE
web    1/1     Running   0          34s

环境变量

变量值几种定义方式：

自定义变量值
变量值从Pod属性获取
变量值从Secrt，ConfigMap

自定义变量值

---
apiVersion: v1
kind: Pod
metadata:
  name: test
spec:
  containers:
  - name: test
    image: busybox
    imagePullPolicy: IfNotPresent
    command: ["bin/sh","-c","sleep 45"]
    env:
    - name: HN
      value: tom

[root@master ~]# kubectl apply -f test.yml 
pod/test created
[root@master ~]# kubectl get pod
NAME   READY   STATUS    RESTARTS   AGE
test   1/1     Running   0          21s
[root@master ~]# kubectl exec -it test -- /bin/sh
/ # echo $HN
tom

变量值从Pod属性获取

[root@master ~]# cat test.yml 
---
apiVersion: v1
kind: Pod
metadata:
  name: test
spec:
  containers:
  - name: test
    image: busybox
    imagePullPolicy: IfNotPresent
    command: ["bin/sh","-c","sleep 45"]
    env:
    - name: HN
      valueFrom:
        fieldRef:
          fieldPath: metadata.name
          
[root@master ~]# kubectl delete -f test.yml 
pod "test" deleted
[root@master ~]# kubectl apply -f test.yml 
pod/test created
[root@master ~]# kubectl get pod
NAME   READY   STATUS    RESTARTS   AGE
test   1/1     Running   0          21s
[root@master ~]# kubectl exec -it test -- /bin/sh
/ # echo $HN
test

变量值从Secrt

[root@master ~]# cat test.yml 
---
apiVersion: v1
kind: Pod
metadata:
  name: test
spec:
  containers:
  - name: test
    image: busybox
    imagePullPolicy: IfNotPresent
    command: ["bin/sh","-c","sleep 45"]
    env:
    - name: HN
      valueFrom:
        fieldRef:
          fieldPath: spec.nodeName

[root@master ~]# kubectl apply -f test.yml 
pod/test created

[root@master ~]# kubectl get pod
NAME   READY   STATUS    RESTARTS   AGE
test   1/1     Running   0          17s
[root@master ~]# kubectl exec -it test -- /bin/sh
/ #  echo $HN
node1

init Container(初始化容器)

初始化容器

Init Container:用于初始化工作，执行完就结束(一次性任务)
支持大部分应用容器配置，但不支持健康检查
优先应用容器执行

应用场景:

环境检查:例如确保应用容器依赖的服务启动后再启动应用容器
初始化配置:例如给应用容器准备配置文件

[root@master manifest]# cat test.yml 
---
apiVersion: v1
kind: Pod
metadata:
  name: web
  namespace: default
spec:
  initContainers:
  - name: download
    image: busybox
    imagePullPolicy: IfNotPresent
    volumeMounts:
    - name: data
      mountPath: /tmp
  containers:
  - name: nginx
    image: nginx
    imagePullPolicy: IfNotPresent
    ports:
    - containerPort: 80
      hostPort: 80
    volumeMounts:
    - name: data
      mountPath: /usr/share/nginx/html
  volumes:
  - name: data
    hostPath:
      path: /var/www/html

//创建目录
[root@node1 ~]# mkdir /var/www/html/ -p
[root@node1 ~]# cd /var/www/html/
[root@node1 html]# echo "hello world" > index.html

[root@node2 ~]# mkdir /var/www/html/ -p
[root@node2 ~]# cd /var/www/html/
[root@node2 html]# echo "123" > index.html

[root@master manifest]# kubectl apply -f test.yml 
pod/web created
[root@master manifest]# kubectl get pod -o wide
NAME   READY   STATUS    RESTARTS   AGE   IP            NODE                NOMINATED NODE   READINESS GATES
web    1/1     Running   0          21s   10.244.2.11   node2.example.com              
[root@master manifest]# curl 10.244.2.11
123

//pod详细信息
[root@master manifest]# kubectl describe pod web
Name:         web
Namespace:    default
Priority:     0
Node:         node2.example.com/192.168.237.141
Start Time:   Thu, 23 Dec 2021 11:15:45 +0800
Labels:       
Annotations:  
Status:       Running
IP:           10.244.2.11
IPs:
  IP:  10.244.2.11
Init Containers:
  download:
    Container ID:   docker://ee2cad1f98d7cc104a0a9e8463d5dc3d4790693de38df6c5b0454bb08d76338a
    Image:          busybox
    Image ID:       docker-pullable://busybox@sha256:b5cfd4befc119a590ca1a81d6bb0fa1fb19f1fbebd0397f25fae164abe1e8a6a
    Port:           
    Host Port:      
    State:          Terminated
      Reason:       Completed
      Exit Code:    0
      Started:      Thu, 23 Dec 2021 11:15:46 +0800
      Finished:     Thu, 23 Dec 2021 11:15:46 +0800
    Ready:          True
    Restart Count:  0
    Environment:    
    Mounts:
      /tmp from data (rw)
      /var/run/secrets/kubernetes.io/serviceaccount from default-token-r4pr2 (ro)
Containers:
  nginx:
    Container ID:   docker://3ccb5aaf458a5b7e6afe4c60a24ea8314690261e342525d0be6f92341723df90
    Image:          nginx
    Image ID:       docker-pullable://nginx@sha256:1f105601bfded0fa298d8c5efd5569f4ed3bf53dc7f4c41c691c29999550f6a3
    Port:           80/TCP
    Host Port:      80/TCP
    State:          Running
      Started:      Thu, 23 Dec 2021 11:15:47 +0800
    Ready:          True
    Restart Count:  0
    Environment:    
    Mounts:
      /usr/share/nginx/html from data (rw)
      /var/run/secrets/kubernetes.io/serviceaccount from default-token-r4pr2 (ro)
Conditions:
  Type              Status
  Initialized       True 
  Ready             True 
  ContainersReady   True 
  PodScheduled      True 
Volumes:
  data:
    Type:          HostPath (bare host directory volume)
    Path:          /var/www/html
    HostPathType:  
  default-token-r4pr2:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  default-token-r4pr2
    Optional:    false
QoS Class:       BestEffort
Node-Selectors:  
Tolerations:     node.kubernetes.io/not-ready:NoExecute op=Exists for 300s
                 node.kubernetes.io/unreachable:NoExecute op=Exists for 300s
Events:
  Type    Reason     Age   From               Message
  ----    ------     ----  ----               -------
  Normal  Scheduled  74s   default-scheduler  Successfully assigned default/web to node2.example.com
  Normal  Pulled     73s   kubelet            Container image "busybox" already present on machine
  Normal  Created    73s   kubelet            Created container download
  Normal  Started    73s   kubelet            Started container download
  Normal  Pulled     73s   kubelet            Container image "nginx" already present on machine
  Normal  Created    72s   kubelet            Created container nginx
  Normal  Started    72s   kubelet            Started container nginx

Pod中会有这几种类型的容器

Infrastructure Container：基础容器
维护整个Pod网络空间
lnitContainers：初始化容器
先于业务容器开始执行
Containers：业务容器
并行启动

haproxy

运行两个pod

[root@master manifest]# cat web.yml 
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: web1
  namespace: default
spec:
  replicas: 1
  selector:
    matchLabels:
      app: web1
  template:  
    metadata:
      labels:
        app: web1
    spec:
      containers:
      - name: web1
        image: wyus/httpd:v0.1
        imagePullPolicy: IfNotPresent
---
apiVersion: v1
kind: Service
metadata:
  name: web1
  namespace: default
spec:
  ports:
  - port: 80
    protocol: TCP
    targetPort: 80
  selector:
    app: web1
  type: NodePort

---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: web2
  namespace: default
spec:
  replicas: 1
  selector:
    matchLabels:
      app: web2
  template:  
    metadata:
      labels:
        app: web2
    spec:
      containers:
      - name: web2
        image: wyus/httpd:v0.2
        imagePullPolicy: IfNotPresent     

---
apiVersion: v1
kind: Service
metadata:
  name: web2
  namespace: default
spec:
  ports:
  - port: 80
    protocol: TCP
    targetPort: 80
  selector:
    app: web2
  type: NodePort

[root@master manifest]# kubectl apply -f web.yml 
deployment.apps/web1 created
service/web1 created
deployment.apps/web2 created
service/web2 created
[root@master ~]# kubectl get pod,svc
NAME                        READY   STATUS    RESTARTS   AGE
pod/web1-855b788957-8fzpg      1/1     Running   0          17m
pod/web2-5f7456967b-t5vqs      1/1     Running   0          17m

NAME                 TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)    AGE
service/kubernetes   ClusterIP   10.96.0.1               443/TCP    3d8h
service/web1         NodePort    10.244.207.28            80:31807/TCP   17m
service/web2         NodePort    10.244.246.130           80:31413/TCP   17m

haproxy

[root@master ~]# cat manifest/haproxy.yml
---
apiVersion: v1
kind: Pod
metadata:
  name: haproxy
  namespace: default
  labels:
    app: haproxy
spec:
  restartPolicy: onFailure
  initContainers:
  - name: data
    volumeMounts:
    - name: data
      mountPath: /tmp
  containers:
  - image: wyus/haproxy:latest
    imagePullPolicy: IfNotPresent
    name: haproxy
    env:
      - name: RSIP
        value: "web1 web2"
    livenessProbe:
      tcpSocket:
        port: 80
      initialDelaySeconds: 20
      periodSeconds: 10
    readinessProbe:
      tcpSocket:
        port: 80
      initialDelaySeconds: 20
      periodSeconds: 10
          
---
apiVersion: v1
kind: Service
metadata:
  name: haproxy
spec:
  ports:
  - port: 80
    protocol: TCP
    targetPort: 80
  selector:
    app: haproxy
  type: NodePort

[root@master ~]# kubectl apply -f manifest/haproxy.yml 
deployment.apps/haproxy created
[root@master ~]# kubectl get pod,svc
NAME                           READY   STATUS    RESTARTS   AGE
pod/haproxy-54c76db7b8-bl44g   1/1     Running   0          111s
pod/web1-855b788957-8fzpg      1/1     Running   0          17m
pod/web2-5f7456967b-t5vqs      1/1     Running   0          17m

NAME                 TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)        AGE
service/haproxy      NodePort    10.106.158.53           8080:30774/TCP,8190:31193/TCP   15m
service/kubernetes   ClusterIP   10.96.0.1                443/TCP        3d9h
service/web1         NodePort    10.244.207.28            80:31807/TCP   17m
service/web2         NodePort    10.244.246.130           80:31413/TCP   17m

[root@master manifest]# curl 10.106.158.53:8080
test1
[root@master manifest]# curl 10.106.158.53:8080
test2

k8s 重启策略、健康检查、环境变量、初始化容器

Linux相关栏目本月热门文章