kubenetes中的pod删除策略 级联删除与非级联删除

2021-07-21 00:55

阅读:1074

标签:实例   svc   基本   start   存储   容器化   less   之间   mina   

StatefulSet 有状态应用[有状态应用]

有状态:StatefulSet
  - 集群节点之间的关系。
  - 数据不完全一致。
  - 实例之间不对等的关系。
  - 依靠外部存储的应用。
  - 通过dns维持身份
  - 每个pod都有特定的名称和网络标识(如pod名是由statefulSet名+有序的数字组成(0、1、2..))
 
 
 redis是有状态应用
StatefulSet(有状态集,缩写为sts)常用于部署有状态的且需要有序启动的应用程序,比如在进行SpringCloud项目容器化时,Eureka的部署是比较适合用StatefulSet部署方式的,可以给每个Eureka实例创建一个唯一且固定的标识符,并且每个Eureka实例无需配置多余的Service,其余Spring Boot应用可以直接通过Eureka的Headless Service即可进行注册。
Eureka的statefulset的资源名称是
eureka,eureka-0
        eureka-1
        eureka-2
Service:headless service,没有ClusterIP [每个]	
eureka-svc Eureka-0.eureka-svc.NAMESPACE_NAME  eureka-1.eureka-svc …
连接 eureka的资源名称为:eureka

statefullset示例

apiVersion: v1
kind: Service
metadata:
  name: nginx
  labels:
    app: nginx
spec:
  ports:
  - port: 80
    name: web
  clusterIP: None
  selector:
    app: nginx
---
apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: web
spec:
  serviceName: "nginx"
  replicas: 2
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx
        ports:
        - containerPort: 80
          name: web


[root@k8s-master01 statefullset]# kubectl create -f statefullset.yaml
service/nginx created
statefulset.apps/web created



#扩容:
replicas: 2  #[定义副本集数量]

[root@k8s-master01 statefullset]# kubectl get pod
web-0                    1/1     Running   0          1m
web-1                    1/1     Running   0          1m


#命令扩容
[root@k8s-master01 statefullset]# kubectl scale --replicas=3 sts web
statefulset.apps/web scaled


#检查扩容
[root@k8s-master01 statefullset]# kubectl get pod
NAME                     READY   STATUS    RESTARTS   AGE
web-0                    1/1     Running   0          15m
web-1                    1/1     Running   0          12m
web-2                    1/1     Running   0          35s



#检查扩容流程:
[root@k8s-master01 ~]# kubectl scale --replicas=5 sts web
statefulset.apps/web scaled

[root@k8s-master01 ~]# kubectl get pod -l app=nginx -w
NAME                     READY   STATUS    RESTARTS   AGE
nginx-68db656dd8-cc4jv   1/1     Running   0          36h
nginx-68db656dd8-hvj8x   1/1     Running   0          36h
web-0                    1/1     Running   0          18h
web-1                    1/1     Running   0          18h
web-2                    0/1     Pending   0          0s
web-2                    0/1     Pending   0          0s
web-2                    0/1     ContainerCreating   0          0s
web-2                    1/1     Running             0          18s
web-3                    0/1     Pending             0          0s
web-3                    0/1     Pending             0          0s
web-3                    0/1     ContainerCreating   0          0s
web-3                    1/1     Running             0          18s
web-4                    0/1     Pending             0          0s
web-4                    0/1     Pending             0          0s
web-4                    0/1     ContainerCreating   0          0s
web-4                    1/1     Running             0          17s

statefulset 更新策略

#默认更新策略:
  updateStrategy:
    rollingUpdate:
      partition: 0         # 设置为1 一次更新一个,如果设置为 0  则是随机更新
    type: RollingUpdate    #--- statefulset默认更新策略 滚动更新,有一个更新失败就不会继续更新,deployment是随机更新模式
    

statefulset更新策略: 
  1. RollingUpdate  滚动更新 [从下往上更新,倒序更新]
  2. OnDelete       手动更新 [需要删除一个pod才会触发更新策略]

默认更新策略RollingUpdate示例:

 执行命令: kubectl edit sts web
 找到:
 containers:
   - image: nginx
 改为:
  containers:
   - image: nginx:1.15.2
   
   
   
 # 查看更新过程:
 [root@k8s-master01 ~]# kubectl get pod -l app=nginx -w 
NAME                     READY   STATUS              RESTARTS   AGE
web-0                    1/1     Running             0          20h
web-1                    1/1     Running             0          20h
web-2                    1/1     Running             0          89m
web-2                    1/1     Terminating         0          90m
web-2                    0/1     Terminating         0          90m
web-2                    0/1     Terminating         0          90m
web-2                    0/1     Terminating         0          90m
web-2                    0/1     Pending             0          0s
web-2                    0/1     Pending             0          0s
web-2                    0/1     ContainerCreating   0          0s
web-2                    1/1     Running             0          19s
web-1                    1/1     Terminating         0          20h
web-1                    0/1     Terminating         0          20h
web-1                    0/1     Terminating         0          20h
web-1                    0/1     Terminating         0          20h
web-1                    0/1     Pending             0          0s
web-1                    0/1     Pending             0          0s
web-1                    0/1     ContainerCreating   0          0s
web-1                    1/1     Running             0          20s
web-0                    1/1     Terminating         0          20h
web-0                    0/1     Terminating         0          20h
web-0                    0/1     Terminating         0          20h
web-0                    0/1     Terminating         0          20h
web-0                    0/1     Pending             0          0s
web-0                    0/1     Pending             0          0s
web-0                    0/1     ContainerCreating   0          0s
web-0                    1/1     Running             0          19s

 # 查看更新结果:
[root@k8s-master01 ~]# kubectl get pod -l app=nginx
NAME                     READY   STATUS    RESTARTS   AGE
web-0                    1/1     Running   0          59s
web-1                    1/1     Running   0          88s
web-2                    1/1     Running   0          118s

# 验证结果:
[root@k8s-master01 ~]# kubectl exec -it web-0 -- sh
# nginx -v
nginx version: nginx/1.15.2

OnDelete策略示例:

# 修改为 OnDelete 策略
# kubectl edit sts web

找到: 
updateStrategy:
  rollingUpdate:
    partition: 0
  type: RollingUpdate
  
改为:     
 updateStrategy:
   type: OnDelete

# 修改镜像版本来查看 是否此策略会更新版本
image: nginx
#改为:
image: nginx:1.15.3

#保存 [删除 rollingUpdate: 与 partition: 0 并且将 type: RollingUpdate 改为 type: OnDelete ]


#检查:
[root@k8s-master01 ~]# kubectl get pod |grep web
NAME                     READY   STATUS    RESTARTS   AGE
web-0                    1/1     Running   0          6m58s
web-1                    1/1     Running   0          7m17s  # 

灰度发布常用策略 partition

#OnDelete与策略解释:
updateStrategy:
  type: OnDelete    # 删除pod才会更新



#partition策略解释 RollingUpdate :
updateStrategy:
  rollingUpdate:
    partition: 2    #大于2的才会被更新
  type: RollingUpdate   #自动更新

级联删除和非级联删除

#级联删除:
#删除 statefullset 时同时删除 pod
[root@k8s-master01 statefullset]# kubectl get pod
NAME                     READY   STATUS    RESTARTS   AGE
nginx-68db656dd8-cc4jv   1/1     Running   0          10d
nginx-68db656dd8-hvj8x   1/1     Running   0          10d
web-0                    1/1     Running   0          3m2s
web-1                    1/1     Running   0          2m42s

[root@k8s-master01 statefullset]# kubectl get sts
NAME   READY   AGE
web    2/2     39s

# kubectl delete sts web 
[root@k8s-master01 statefullset]# kubectl get sts
No resources found in default namespace.



#非级联删除 [极少使用] 删除 statefullset 不删除 pod
[root@k8s-master01 statefullset]# kubectl get sts
NAME   READY   AGE
web    2/2     3s

[root@k8s-master01 statefullset]# kubectl delete sts web --cascade=orphan
statefulset.apps "web" deleted

[root@k8s-master01 statefullset]# kubectl get pod
NAME                     READY   STATUS    RESTARTS   AGE
nginx-68db656dd8-cc4jv   1/1     Running   0          10d
nginx-68db656dd8-hvj8x   1/1     Running   0          10d
web-0                    1/1     Running   0          3m43s
web-1                    1/1     Running   0          3m35s


kubectl delete sts web --cascade=orphan #非级联删除 ,但是pod不会被删除,会变成孤儿pod
kubectl delete pod web-0 web-1     #这时候删除的pod都不会被重建。


#注意 非级联删除 pod是不会被删除的,他会变成孤儿pod,此时使用 kubectl delete pod web-0 web-1 删除不会再创建

在日常中基本不会用到非级联删除,因为非级联删除还会遗留pod来手工操作删除

kubenetes中的pod删除策略 级联删除与非级联删除

标签:实例   svc   基本   start   存储   容器化   less   之间   mina   

原文地址:https://www.cnblogs.com/superlinux/p/15024680.html


评论


亲,登录后才可以留言!