以下部署过程基于官方手册，稍作调整以便理解。
若还未搭建环境，可参考CentOS7 Kubernetes minikube本地环境搭建一文搭建环境。
本示例包含以下内容：
由于本篇内容是按实际操作顺序编排，如要实践本篇中的事例，推荐从头到尾按顺序一步步执行，以免“跳戏”。
创建Deployment
$ vim nginx-deployment.yaml
创建nginx-deployment.yaml文件，声明deployment的各项参数，文件内容如下：
apiVersion: extensions/v1beta1
kind: Deployment
metadata:
#指定deployment的名称
name: nginx-deployment
labels:
app: nginx
spec:
#期望创建3个nginx实例（3个pod）
replicas: 3
selector:
#选择label：app=nginx的pod来创建实例
matchLabels:
app: nginx
template:
metadata:
labels:
app: nginx
spec:
containers:
- name: nginx
image: nginx:1.7.9
ports:
- containerPort: 80
该文件表示要创建3个pod，每个pod运行nginx1.7.9镜像。
执行创建
$ kubectl create -f nginx-deployment.yaml $ deployment.apps "nginx-deployment" created
创建成功，查看deployment和pod的运行状况：
$ kubectl get deploy NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE nginx-deployment 3 3 3 3 57s $ kubectl get po -o wide NAME READY STATUS RESTARTS AGE IP NODE nginx-deployment-75675f5897-jgl27 1/1 Running 0 12s 172.17.0.5 minikube nginx-deployment-75675f5897-rrnsk 1/1 Running 0 12s 172.17.0.7 minikube nginx-deployment-75675f5897-txlkv 1/1 Running 0 12s 172.17.0.6 minikube
可以看到deployment已经创建成功，并创建了3个pod，IP地址分别是172.17.0.5，172.17.0.6，172.17.0.7。
分别对三个IP地址使用curl命令，均能返回nginx的欢迎页内容，表示创建成功。
$ curl http://172.17.0.5
<!DOCTYPE html>
<html>
<head>
<title>Welcome to nginx!</title>
<style>
body {
width: 35em;
margin: 0 auto;
font-family: Tahoma, Verdana, Arial, sans-serif;
}
</style>
</head>
<body>
<h1>Welcome to nginx!</h1>
<p>If you see this page, the nginx web server is successfully installed and
working. Further configuration is required.</p>
<p>For online documentation and support please refer to
<a href="http://nginx.org/">nginx.org</a>.<br/>
Commercial support is available at
<a href="http://nginx.com/">nginx.com</a>.</p>
<p><em>Thank you for using nginx.</em></p>
</body>
</html>
更新Deployment
我们将nginx:1.7.9版本更新到nginx:1.9.1。
手动设置升级的nginx版本：
$ kubectl set image deployment/nginx-deployment nginx=nginx:1.9.1 --record deployment.apps "nginx-deployment" image updated
查看rollout，pod和rs状态：
$ kubectl rollout status deploy nginx-deployment Waiting for rollout to finish: 2 out of 3 new replicas have been updated... $ kubectl get po NAME READY STATUS RESTARTS AGE nginx-deployment-75675f5897-rrnsk 1/1 Terminating 0 1m nginx-deployment-75675f5897-txlkv 0/1 Terminating 0 1m nginx-deployment-c4747d96c-bf97v 1/1 Running 0 4s nginx-deployment-c4747d96c-ctpl2 1/1 Running 0 5s nginx-deployment-c4747d96c-rthg5 1/1 Running 0 6s $ get rs NAME DESIRED CURRENT READY AGE nginx-deployment-75675f5897 0 0 0 1m nginx-deployment-c4747d96c 3 3 3 9s
可以看到旧的pod正在被终止，新的pod被创建出来，这是滚动升级的过程。
等升级完成再来看看rollout和pod的状态：
$ kubectl rollout status deploy nginx-deployment deployment "nginx-deployment" successfully rolled out $ kubectl get po NAME READY STATUS RESTARTS AGE nginx-deployment-c4747d96c-bf97v 1/1 Running 0 4m nginx-deployment-c4747d96c-ctpl2 1/1 Running 0 4m nginx-deployment-c4747d96c-rthg5 1/1 Running 0 4m
可以看到已经升级完成了，依旧是3个pod。
我们可以通过describe命令查看deployment的升级过程：
$ kubectl describe deployments
Name: nginx-deployment
Namespace: default
CreationTimestamp: Mon, 31 Dec 2018 16:52:13 +0800
Labels: app=nginx
Annotations: deployment.kubernetes.io/revision=2
kubernetes.io/change-cause=kubectl set image deployment/nginx-deployment nginx=nginx:1.9.1 --record=true
Selector: app=nginx
Replicas: 3 desired | 3 updated | 3 total | 3 available | 0 unavailable
StrategyType: RollingUpdate
MinReadySeconds: 0
RollingUpdateStrategy: 25% max unavailable, 25% max surge
Pod Template:
Labels: app=nginx
Containers:
nginx:
Image: nginx:1.9.1
Port: 80/TCP
Host Port: 0/TCP
Environment: <none>
Mounts: <none>
Volumes: <none>
Conditions:
Type Status Reason
---- ------ ------
Available True MinimumReplicasAvailable
Progressing True NewReplicaSetAvailable
OldReplicaSets: <none>
NewReplicaSet: nginx-deployment-c4747d96c (3/3 replicas created)
Events:
Type Reason Age From Message
---- ------ ---- ---- -------
Normal ScalingReplicaSet 12m deployment-controller Scaled up replica set nginx-deployment-75675f5897 to 3
Normal ScalingReplicaSet 10m deployment-controller Scaled up replica set nginx-deployment-c4747d96c to 1
Normal ScalingReplicaSet 10m deployment-controller Scaled down replica set nginx-deployment-75675f5897 to 2
Normal ScalingReplicaSet 10m deployment-controller Scaled up replica set nginx-deployment-c4747d96c to 2
Normal ScalingReplicaSet 10m deployment-controller Scaled down replica set nginx-deployment-75675f5897 to 1
Normal ScalingReplicaSet 10m deployment-controller Scaled up replica set nginx-deployment-c4747d96c to 3
Normal ScalingReplicaSet 10m deployment-controller Scaled down replica set nginx-deployment-75675f5897 to 0
从Event.Message这一列可以看出，滚动升级的过程就是把新的deployment的ReplicaSet设置从0慢慢上升到期望，这里也就是3，同时把旧deployment的ReplicaSet设置从3慢慢降到0。
根据RollingUpdateStrategy的设置，保证在滚动升级中可用pod的数量。
回滚Deployment
当更新失败，或者是更新版本不稳定，k8s也可以方便的回滚到指定的历史版本。
比如，当我们更新的时候，把1.9.1误输成了1.91
$ kubectl set image deployment/nginx-deployment nginx=nginx:1.91 --record deployment.apps "nginx-deployment" image updated
查看deploy，pod和rollout状态：
$ kubectl get deploy NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE nginx-deployment 3 4 1 3 45m $ kubectl get po NAME READY STATUS RESTARTS AGE nginx-deployment-595696685f-q8zrr 0/1 ImagePullBackOff 0 29s nginx-deployment-c4747d96c-bf97v 1/1 Running 0 44m nginx-deployment-c4747d96c-ctpl2 1/1 Running 0 44m nginx-deployment-c4747d96c-rthg5 1/1 Running 0 44m $ kubectl rollout status deploy nginx-deployment Waiting for rollout to finish: 1 out of 3 new replicas have been updated...
命令kubectl rollout status deploy nginx-deployment会阻塞等待deployment更新完成，由于我们输入了错误的nginx版本，无法更新完成，这条命令会一直阻塞，所以我们先CTRL+C终止执行。
UP-TO-DATE=1表示创建了一个用于更新的Pod。
从kubectl get po的结果我们得知有一个pod创建不成功。
我们再通过describe命令看一下升级不成功的pod发生了什么：
$kubectl describe po nginx-deployment-595696685f-q8zrr
Name: nginx-deployment-595696685f-q8zrr
Namespace: default
Node: minikube/192.168.0.62
Start Time: Mon, 31 Dec 2018 17:37:07 +0800
Labels: app=nginx
pod-template-hash=1512522419
Annotations: <none>
Status: Pending
IP: 172.17.0.7
Controlled By: ReplicaSet/nginx-deployment-595696685f
Containers:
nginx:
Container ID:
Image: nginx:1.91
Image ID:
Port: 80/TCP
Host Port: 0/TCP
State: Waiting
Reason: ImagePullBackOff
Ready: False
Restart Count: 0
Environment: <none>
Mounts:
/var/run/secrets/kubernetes.io/serviceaccount from default-token-q244m (ro)
Conditions:
Type Status
Initialized True
Ready False
PodScheduled True
Volumes:
default-token-q244m:
Type: Secret (a volume populated by a Secret)
SecretName: default-token-q244m
Optional: false
QoS Class: BestEffort
Node-Selectors: <none>
Tolerations: node.kubernetes.io/not-ready:NoExecute for 300s
node.kubernetes.io/unreachable:NoExecute for 300s
Events:
Type Reason Age From Message
---- ------ ---- ---- -------
Normal Scheduled 12m default-scheduler Successfully assigned nginx-deployment-595696685f-q8zrr to minikube
Normal SuccessfulMountVolume 12m kubelet, minikube MountVolume.SetUp succeeded for volume "default-token-q244m"
Normal Pulling 10m (x4 over 12m) kubelet, minikube pulling image "nginx:1.91"
Warning Failed 10m (x4 over 12m) kubelet, minikube Failed to pull image "nginx:1.91": rpc error: code = Unknown desc = Error response from daemon: manifest for nginx:1.91 not found
Warning Failed 10m (x4 over 12m) kubelet, minikube Error: ErrImagePull
Normal BackOff 9m (x6 over 11m) kubelet, minikube Back-off pulling image "nginx:1.91"
Warning Failed 1m (x40 over 11m) kubelet, minikube Error: ImagePullBackOff
可以看到，由于镜像nginx:1.91是不存在的，所以pod报错：Failed to pull image "nginx:1.91":xxxx，导致升级不成功。
通过以下命令查看升级历史记录，就跟svn，git差不多。
$ kubectl rollout history deploy nginx-deployment deployments "nginx-deployment" REVISION CHANGE-CAUSE 1 <none> 2 kubectl set image deployment/nginx-deployment nginx=nginx:1.9.1 --record=true 3 kubectl set image deployment/nginx-deployment nginx=nginx:1.91 --record=true
由于我们之前升级加了--record参数，可以看到升级当时所使用命令。
使用rollout history xxx --revision=n可以查看明细：
$ kubectl rollout history deploy nginx-deployment --revision=2
deployments "nginx-deployment" with revision #2
Pod Template:
Labels: app=nginx
pod-template-hash=703038527
Annotations: kubernetes.io/change-cause=kubectl set image deployment/nginx-deployment nginx=nginx:1.9.1 --record=true
Containers:
nginx:
Image: nginx:1.9.1
Port: 80/TCP
Host Port: 0/TCP
Environment: <none>
Mounts: <none>
Volumes: <none>
用undo操作回滚：
$ kubectl rollout undo deploy nginx-deployment deployment.apps "nginx-deployment"
或者使用--to-revision来回滚到指定版本：
$ kubectl rollout undo deploy nginx-deployment --to-revision=2 deployment.apps "nginx-deployment"
看一下rollout,pod,rs的状态：
$ get deployment NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE nginx-deployment 3 3 3 3 1h $ kubectl get pod NAME READY STATUS RESTARTS AGE nginx-deployment-c4747d96c-bf97v 1/1 Running 0 1h nginx-deployment-c4747d96c-ctpl2 1/1 Running 0 1h nginx-deployment-c4747d96c-rthg5 1/1 Running 0 1h $ kubectl get rs NAME DESIRED CURRENT READY AGE nginx-deployment-595696685f 0 0 0 1h nginx-deployment-75675f5897 0 0 0 1h nginx-deployment-c4747d96c 3 3 3 1h $ kubectl rollout status deploy nginx-deployment deployment "nginx-deployment" successfully rolled out
回滚成功。
describe查看一下deployment：
$ kubectl describe deploy nginx-deployment
Name: nginx-deployment
Namespace: default
CreationTimestamp: Mon, 31 Dec 2018 16:52:13 +0800
Labels: app=nginx
Annotations: deployment.kubernetes.io/revision=4
kubernetes.io/change-cause=kubectl set image deployment/nginx-deployment nginx=nginx:1.9.1 --record=true
Selector: app=nginx
Replicas: 3 desired | 3 updated | 3 total | 3 available | 0 unavailable
StrategyType: RollingUpdate
MinReadySeconds: 0
RollingUpdateStrategy: 25% max unavailable, 25% max surge
Pod Template:
Labels: app=nginx
Containers:
nginx:
Image: nginx:1.9.1
Port: 80/TCP
Host Port: 0/TCP
Environment: <none>
Mounts: <none>
Volumes: <none>
Conditions:
Type Status Reason
---- ------ ------
Available True MinimumReplicasAvailable
Progressing True NewReplicaSetAvailable
OldReplicaSets: <none>
NewReplicaSet: nginx-deployment-c4747d96c (3/3 replicas created)
Events:
Type Reason Age From Message
---- ------ ---- ---- -------
Normal DeploymentRollback 4m deployment-controller Rolled back deployment "nginx-deployment" to revision 2
Normal ScalingReplicaSet 4m deployment-controller Scaled down replica set nginx-deployment-595696685f to 0
从Events.Message中得知，deployment-controller将版本回滚到revision 2了。
Deployment扩容和缩容
Deployment扩容缩容的本质其实就是改变ReplicaSet的数量来控制Pod的数量，增加就是扩容，缩小就是缩容。
扩容
假设我们要将nginx从3个pod扩展到10个pod：
$ kubectl scale deploy nginx-deployment --replicas=10 deployment.extensions "nginx-deployment" scaled
查看po,rs,deployment状态：
$ kubectl get po NAME READY STATUS RESTARTS AGE nginx-deployment-c4747d96c-2zw9w 0/1 ContainerCreating 0 2s nginx-deployment-c4747d96c-bf97v 1/1 Running 0 2h nginx-deployment-c4747d96c-ctpl2 1/1 Running 0 2h nginx-deployment-c4747d96c-fr6b6 0/1 ContainerCreating 0 2s nginx-deployment-c4747d96c-grgr6 0/1 ContainerCreating 0 2s nginx-deployment-c4747d96c-hh9tk 0/1 ContainerCreating 0 2s nginx-deployment-c4747d96c-pdz4q 0/1 ContainerCreating 0 2s nginx-deployment-c4747d96c-rthg5 1/1 Running 0 2h nginx-deployment-c4747d96c-rvfs7 0/1 ContainerCreating 0 2s nginx-deployment-c4747d96c-s5mrc 0/1 ContainerCreating 0 2s $ kubectl get rs NAME DESIRED CURRENT READY AGE nginx-deployment-595696685f 0 0 0 1h nginx-deployment-75675f5897 0 0 0 2h nginx-deployment-c4747d96c 10 10 7 2h $ kubectl get deploy NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE nginx-deployment 10 10 10 10 2h
可以看到，pod已经从3个变成10个了，这里pod的状态还在启动中，启动完成之后10个pod都将会是Running状态。
缩容
假设我们要将nginx从10个pod减少为3个：
$ kubectl scale deploy nginx-deployment --replicas=3 deployment.extensions "nginx-deployment" scaled $ kubectl get deploy NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE nginx-deployment 3 3 3 3 2h $ kubectl get rs NAME DESIRED CURRENT READY AGE nginx-deployment-595696685f 0 0 0 1h nginx-deployment-75675f5897 0 0 0 2h nginx-deployment-c4747d96c 3 3 3 2h $ kubeget po NAME READY STATUS RESTARTS AGE nginx-deployment-c4747d96c-2zw9w 1/1 Terminating 0 3m nginx-deployment-c4747d96c-bf97v 1/1 Running 0 2h nginx-deployment-c4747d96c-ctpl2 1/1 Running 0 2h nginx-deployment-c4747d96c-hh9tk 1/1 Terminating 0 3m nginx-deployment-c4747d96c-pdz4q 0/1 Terminating 0 3m nginx-deployment-c4747d96c-rthg5 1/1 Running 0 2h
可以看到pod的数量在减少，当缩容成功之后，会只剩下3个pod。
手动扩容缩绒还不能体现k8s的精髓之处，k8s还能够根据节点的负载情况，自动扩容缩容，这样才能实现自动化，智能化。自动扩容的内容不在本篇中的讨论范围内。大概可能或许假以时日会写一篇关于自动扩容的例子。
Deployment的暂停和恢复
在上述的例子中，当我们输入更新指令之后，滚动更新就立即开始了，但有的时候，我们还需要修改其他一些配置来保证新版本的正确运行，这时就需要先暂停滚动更新，等其他配置工作完成之后，再开始滚动升级。
我们删掉当前deployment，用初始文件重新部署一个:
$ kubectl delete deploy nginx-deployment deployment.extensions "nginx-deployment" deleted $ kubectl create -f nginx-deployment.yaml deployment.apps "nginx-deployment" created
暂停滚动更新：
$ kubectl rollout pause deploy nginx-deployment deployment.apps "nginx-deployment" paused
我们将当前的1.9.1版本“升级”成1.7.9：
$ kubectl set image deploy nginx-deployment nginx=nginx:1.7.9 deployment.apps "nginx-deployment" image updated
查看rollout history
$ kubectl rollout history deploy nginx-deployment deployments "nginx-deployment" REVISION CHANGE-CAUSE 1 <none>
我们发现由于我们使用rollout pause暂停了滚动更新，该deployment并没有更新。
继续修改配置，完成更新操作：
$ kubectl set resources deploy nginx-deployment -c=nginx --limits=cpu=200m,memory=512Mi deployment.apps "nginx-deployment" resource requirements updated $ kubectl rollout resume deploy nginx-deployment deployment.apps/nginx-deployment resumed
配置完之后查看pod状态：
$ kubectl get pod NAME READY STATUS RESTARTS AGE nginx-deployment-6d95f89565-7wgrb 0/1 Pending 0 2m nginx-deployment-75675f5897-9pqz5 1/1 Running 0 18m nginx-deployment-75675f5897-mdclm 1/1 Running 0 18m nginx-deployment-75675f5897-vf4lw 1/1 Running 0 18m
发现第一个Pod一直处于Pending状态，这剧本好像不太对…
有了前面describe查看deployment升级失败的例子，这里我们举一反三一下：
$ kubectl describe pod nginx-deployment-6d95f89565-7wgrb
Name: nginx-deployment-6d95f89565-7wgrb
Namespace: default
Node: <none>
Labels: app=nginx
pod-template-hash=2851945121
Annotations: <none>
Status: Pending
IP:
Controlled By: ReplicaSet/nginx-deployment-6d95f89565
Containers:
nginx:
Image: nginx:1.7.9
Port: 80/TCP
Host Port: 0/TCP
Limits:
cpu: 200m
memory: 512Mi
Requests:
cpu: 200m
memory: 512Mi
Environment: <none>
Mounts:
/var/run/secrets/kubernetes.io/serviceaccount from default-token-q244m (ro)
Conditions:
Type Status
PodScheduled False
Volumes:
default-token-q244m:
Type: Secret (a volume populated by a Secret)
SecretName: default-token-q244m
Optional: false
QoS Class: Guaranteed
Node-Selectors: <none>
Tolerations: node.kubernetes.io/not-ready:NoExecute for 300s
node.kubernetes.io/unreachable:NoExecute for 300s
Events:
Type Reason Age From Message
---- ------ ---- ---- -------
Warning FailedScheduling 4s (x12 over 2m) default-scheduler 0/1 nodes are available: 1 Insufficient cpu.
看到Events知道了，是我的服务器太渣了，分配不起200m的cpu资源，那就分配少一些，重新分配一下：
$ kubectl set resources deploy nginx-deployment -c=nginx --limits=cpu=20m,memory=256Mi deployment.apps "nginx-deployment" resource requirements updated
等一阵子等pod启动完毕：
$ kubectl get po NAME READY STATUS RESTARTS AGE nginx-deployment-6b58c6db49-n9jmf 1/1 Running 0 20s nginx-deployment-6b58c6db49-rbwxw 1/1 Running 0 22s nginx-deployment-6b58c6db49-rklg2 1/1 Running 0 17s
现在能正常更新完成了。
用Service暴露Deployment
到目前为止，上述deployment所生产的Pod都只能被Node内部访问，而且而且每次pod扩容，缩容或者是升级之后，IP地址均可能改变。我们用service将deployment暴露出去，让k8s外部的应用也可以访问：
$ kubectl expose deploy nginx-deployment --type=NodePort --name=nginx --port=80 service "nginx" exposed
使用get svc查看服务：
$ kubectl get svc NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE kubernetes ClusterIP 10.96.0.1 <none> 443/TCP 21h nginx NodePort 10.111.219.247 <none> 80:31302/TCP 13s
这时候外部应用就可以使用31302这个端口来访问内部的包含nginx的pod了。
这里就不着重介绍service了，service的内容就留到下一篇吧。