在kubernetes 1.8以下的版本滚动升级要求使用deployment

最新推荐文章于 2022-11-22 16:26:41 发布

clear_soul

最新推荐文章于 2022-11-22 16:26:41 发布

阅读量311

点赞数

文章标签： kubernetes 回滚 deployment

本文链接：https://blog.csdn.net/kingrootZ/article/details/96136686

版权

在kubernetes 1.8以下的版本滚动升级要求使用deployment

控制器deployment 是用来管理无状态应用的，面向的集群的管理，而不是面向的是一个不可变的个体。简单来说就一笼包子吃掉一个在放一个进去就行了。

Deployment为Pod和ReplicaSet提供了一个声明式定义(declarative)方法，用来替代以前的ReplicationController来方便的管理应用。典型的应用场景包括：

定义Deployment来创建Pod和ReplicaSet
滚动升级和回滚应用
扩容和缩容
暂停和继续Deployment

网上找个小栗子：

[root@cicp-k8s-master1 test]# cat nginx.yaml 
apiVersion: extensions/v1beta1
kind: Deployment
metadata:
  name: nginx-deployment
  labels:
    app: nginx
spec:
  replicas: 3
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: registry:5000/nginx:1.9.1
        ports:

先创建然后查询下

[root@cicp-k8s-master1 heapster]# kubectl create -f nginx.yaml  --record
[root@cicp-k8s-master1 heapster]# kubectl get deployment  -o wide
NAME               DESIRED   CURRENT   UP-TO-DATE   AVAILABLE   AGE
nginx-deployment   3         3         3            3           46s

先扩个容

[root@cicp-k8s-master1 heapster]# kubectl scale deployment nginx-deployment --replicas 4
deployment "nginx-deployment" scaled
[root@cicp-k8s-master1 heapster]# kubectl get deployment  -o wide
NAME               DESIRED   CURRENT   UP-TO-DATE   AVAILABLE   AGE
nginx-deployment   4         4         4            4           4m
[root@cicp-k8s-master1 heapster]# kubectl get pod   -o wide
NAME                                READY     STATUS    RESTARTS   AGE       IP            NODE
mongo-6sv3i                         1/1       Running   9          1y        172.17.56.3   cicp-k8s-node2
nginx-deployment-1969799642-3l01p   1/1       Running   0          9m        172.17.83.9   cicp-k8s-node5
nginx-deployment-1969799642-9s5tv   1/1       Running   0          9m        172.17.39.5   cicp-k8s-node6
nginx-deployment-1969799642-nunr4   1/1       Running   0          5m        172.17.8.4    cicp-k8s-node1
nginx-deployment-1969799642-sa1tg   1/1       Running   0          9m        172.17.63.5   cicp-k8s-node3

更新镜像:

kubectl set image deployment/nginx-deployment nginx=registry:5000/nginx:1.7.9 --record

nginx=registry:5000/nginx:1.7.9 这个前一个参数是app名字，后面是镜像版本

查看状态和服务：

kubectl rollout status  deployment/nginx-deployment
镜像改变了
[root@cicp-k8s-master1 ~]# kubectl describe pod nginx-deployment-2457125344-3hzc5
Name:		nginx-deployment-2457125344-3hzc5
Namespace:	default
Node:		cicp-k8s-node1/10.169.2.27
Start Time:	Mon, 15 Jul 2019 18:05:48 +0800
Labels:		app=nginx
		pod-template-hash=2457125344
Status:		Running
IP:		172.17.8.4
Controllers:	ReplicaSet/nginx-deployment-2457125344
Containers:
  nginx:
    Container ID:		docker://1fd3ac70081be7465cd4e2ba98ca9470ebc6676cf9824365b37a8ecdbd37c965
    Image:			registry:5000/nginx:1.7.9
    Image ID:			docker://sha256:84581e99d807a703c9c03bd1a31cd9621815155ac72a7365fd02311264512656
    Port:			80/TCP
    State:			Running
      Started:			Mon, 15 Jul 2019 18:05:48 +0800
    Ready:			True
    Restart Count:		0
    Volume Mounts:		<none>
    Environment Variables:	<none>
Conditions:
  Type		Status
  Initialized 	True 
  Ready 	True 
  PodScheduled 	True 
No volumes.
QoS Class:	BestEffort
Tolerations:	<none>
Events:
  FirstSeen	LastSeen	Count	From				SubobjectPath		Type	      Reason		Message
  ---------	--------	-----	----				-------------		--------      ------		-------
  32s		32s		1	{default-scheduler }					Normal	      Scheduled	Successfully assigned nginx-deployment-2457125344-3hzc5 to cicp-k8s-node1
  32s		32s		1	{kubelet cicp-k8s-node1}	spec.containers{nginx}	Normal	      Pulled		Container image "registry:5000/nginx:1.7.9" already present on machine
  32s		32s		1	{kubelet cicp-k8s-node1}	spec.containers{nginx}	Normal	      Created		Created container with docker id 1fd3ac70081b; Security:[seccomp=unconfined]
  32s		32s		1	{kubelet cicp-k8s-node1}	spec.containers{nginx}	Normal	      Started		Started container with docker id 1fd3ac70081b

终止升级

kubectl rollout pause deployment/nginx-deployment

继续升级

kubectl rollout resume deployment/nginx-deployment

操作记录

[root@cicp-k8s-master1 test]# kubectl rollout history deployments
deployments "nginx-deployment"
REVISION	CHANGE-CAUSE
1		kubectl scale deployment nginx-deployment --replicas 4
2		kubectl set image deployment/nginx-deployment nginx=registry:5000/nginx:1.7.9 --record

回退到指定版本

回退到第一个版本

kubectl rollout undo deployment/nginx-deployment --to-revision=1

查看现在的pod

[root@cicp-k8s-master1 ~]# kubectl get pod 
NAME                                READY     STATUS    RESTARTS   AGE
mongo-6sv3i                         1/1       Running   9          1y
nginx-deployment-1969799642-6vgw5   1/1       Running   0          26s
nginx-deployment-1969799642-902r4   1/1       Running   0          27s
nginx-deployment-1969799642-fgwk1   1/1       Running   0          27s
nginx-deployment-1969799642-znrpi   1/1       Running   0          26s
tyk-dashboard-cmqvy                 1/1       Running   9          1y
tyk-gateway-r6eap                   1/1       Running   9          1y
[root@cicp-k8s-master1 ~]# kubectl describe pod nginx-deployment-1969799642-znrpi 
Name:		nginx-deployment-1969799642-znrpi
Namespace:	default
Node:		cicp-k8s-node5/10.169.2.31
Start Time:	Mon, 15 Jul 2019 18:11:23 +0800
Labels:		app=nginx
		pod-template-hash=1969799642
Status:		Running
IP:		172.17.83.9
Controllers:	ReplicaSet/nginx-deployment-1969799642
Containers:
  nginx:
    Container ID:		docker://85bb72d79329af37497f5b61d96acc5ea3793e2e46b9b45badabb31af40b5e76
    Image:			registry:5000/nginx:1.9.1
    Image ID:			docker://sha256:94ec7e53edfc793d6d8412b4748cd84270da290ce9256730eb428574f98f7c95
    Port:			80/TCP
    State:			Running
      Started:			Mon, 15 Jul 2019 18:11:24 +0800
    Ready:			True
    Restart Count:		0
    Volume Mounts:		<none>
    Environment Variables:	<none>
Conditions:
  Type		Status
  Initialized 	True 
  Ready 	True 
  PodScheduled 	True 
No volumes.
QoS Class:	BestEffort
Tolerations:	<none>
Events:
  FirstSeen	LastSeen	Count	From				SubobjectPath		Type	      Reason		Message
  ---------	--------	-----	----				-------------		--------      ------		-------
  50s		50s		1	{default-scheduler }					Normal	      Scheduled	Successfully assigned nginx-deployment-1969799642-znrpi to cicp-k8s-node5
  49s		49s		1	{kubelet cicp-k8s-node5}	spec.containers{nginx}	Normal	      Pulled		Container image "registry:5000/nginx:1.9.1" already present on machine
  49s		49s		1	{kubelet cicp-k8s-node5}	spec.containers{nginx}	Normal	      Created		Created container with docker id 85bb72d79329; Security:[seccomp=unconfined]
  49s		49s		1	{kubelet cicp-k8s-node5}	spec.containers{nginx}	Normal	      Started		Started container with docker id 85bb72d79329

其他修改方式:

kubectl edit deployment/nginx-deployment

# Please edit the object below. Lines beginning with a '#' will be ignored,
# and an empty file will abort the edit. If an error occurs while saving this file will be
# reopened with the relevant failures.
#
apiVersion: extensions/v1beta1
kind: Deployment
metadata:
 annotations:
   deployment.kubernetes.io/revision: "3"
   kubernetes.io/change-cause: kubectl edit deployment/nginx-deployment
 creationTimestamp: 2019-07-15T09:50:28Z
 generation: 10
 labels:
   app: nginx
 name: nginx-deployment
 namespace: default
 resourceVersion: "125366735"
 selfLink: /apis/extensions/v1beta1/namespaces/default/deployments/nginx-deployment
 uid: fa30a2da-a6e5-11e9-adca-000c296454ba
spec:
 replicas: 2
 selector:
   matchLabels:
     app: nginx
 strategy:
   rollingUpdate:
     maxSurge: 1
     maxUnavailable: 1
   type: RollingUpdate


# 实例数我改成了 2

[root@cicp-k8s-master1 ~]# kubectl get pod 
NAME                                READY     STATUS    RESTARTS   AGE
mongo-6sv3i                         1/1       Running   9          1y
nginx-deployment-1969799642-902r4   1/1       Running   0          6m
nginx-deployment-1969799642-fgwk1   1/1       Running   0          6m
tyk-dashboard-cmqvy                 1/1       Running   9          1y
tyk-gateway-r6eap                   1/1       Running   9          1y

deployment文件说明

apiVersion: extensions/v1beta1
kind: Deployment
metadata:
  name: nginx
  namespace: kube-system
  labels:
    app: nginx
spec:
  replicas: 3
#  minReadySeconds: 60     #滚动升级时60s后认为该pod就绪
  strategy:
    rollingUpdate:  ##由于replicas为3,则整个升级,pod个数在2-4个之间
      maxSurge: 1      #滚动升级时会先启动1个pod
      maxUnavailable: 1 #滚动升级时允许的最大Unavailable的pod个数
  template:
    metadata:
      labels:
        app: nginx
    spec:
      terminationGracePeriodSeconds: 60 ##k8s将会给应用发送SIGTERM信号，可以用来正确、优雅地关闭应用,默认为30秒
      containers:
      - name: nginx
        image: nginx
        imagePullPolicy: IfNotPresent
        livenessProbe: #kubernetes认为该pod是存活的,不存活则需要重启
          httpGet:
            path: /health
            port: 80
            scheme: HTTP
          initialDelaySeconds: 60 ## equals to the maximum startup time of the application + couple of seconds
          timeoutSeconds: 5
          successThreshold: 1
          failureThreshold: 5
        readinessProbe: #kubernetes认为该pod是启动成功的
          httpGet:
            path: /health
            port: 80
            scheme: HTTP
          initialDelaySeconds: 30 ## equals to minimum startup time of the application
          timeoutSeconds: 5
          successThreshold: 1
          failureThreshold: 5
        resources:
          # keep request = limit to keep this container in guaranteed class
          requests:
            cpu: 50m
            memory: 200Mi
          limits:
            cpu: 500m
            memory: 500Mi
        ports:
          - name: http
            containerPort: 80

几个重要参数说明

maxSurge与maxUnavailable

maxSurge: 1 表示滚动升级时会先启动1个pod
maxUnavailable: 1 表示滚动升级时允许的最大Unavailable的pod个数
由于replicas为3,则整个升级,pod个数在2-4个之间

terminationGracePeriodSeconds

k8s将会给应用发送SIGTERM信号，可以用来正确、优雅地关闭应用,默认为30秒。
如果需要更优雅地关闭，则可以使用k8s提供的pre-stop lifecycle hook 的配置声明，将会在发送SIGTERM之前执行。

livenessProbe与readinessProbe

livenessProbe是kubernetes认为该pod是存活的，不存在则需要kill掉，然后再新启动一个，以达到replicas指定的个数。
readinessProbe是kubernetes认为该pod是启动成功的，这里根据每个应用的特性，自己去判断，可以执行command，也可以进行httpGet。比如对于使用java web服务的应用来说，并不是简单地说tomcat启动成功就可以对外提供服务的，还需要等待spring容器初始化，数据库连接连接上等等。对于spring boot应用，默认的actuator带有/health接口，可以用来进行启动成功的判断。
其中readinessProbe.initialDelaySeconds可以设置为系统完全启动起来所需的最少时间，livenessProbe.initialDelaySeconds可以设置为系统完全启动起来所需的最大时间+若干秒

clear_soul

关注

0
点赞
踩
0

收藏

觉得还不错? 一键收藏
0
评论
在kubernetes 1.8以下的版本滚动升级要求使用deployment

在kubernetes 1.8以下的版本滚动升级要求使用deployment控制器deployment 是用来管理无状态应用的，面向的集群的管理，而不是面向的是一个不可变的个体。简单来说就一笼包子吃掉一个在放一个进去就行了。Deployment为Pod和ReplicaSet提供了一个声明式定义(declarative)方法，用来替代以前的ReplicationController来方便的管理...
复制链接

扫一扫