在之前的一章中简单的介绍了Yum安装kubernetes的方式，这次的笔记记录kubernetes的一些基础入门的一些操作，更多的是关乎于kubectl的一些使用，如：创建pod，scv，动态的增删，回滚，以及简单的网络等
kubectl
kubectl其实就是kubernetes的一个客户端程序，他通过连接master节点上的api-server，而在api-server上，kubectl就是唯一的一个api-server管理工具，他能够连接到api-server，实现对象资源的增删改查等基本操作kubectl的基本命令中基础命令：create,expose,run,set中级命令：explain,get,edit,delete部署命令，rollout,scale,autoscale集群管理：certificate,cluster-info,top,cordon,uncordon,drain,taint系统：describe,logs,attach,exec,port-forward,porxy等高级命令：apply,patch,replace,wait,convert设置命令：label,annotate,completion其他：alpha,api-resources,api-versions,config,plugin,version
run pod
[root@linuxea ~]# kubectl run nginx-linuxea --image=marksugar/nginx:1.14.a --port=8787 --replicas=1 --dry-run=true
deployment.apps/nginx-linuxea created (dry run)--image：镜象，--port:暴露端口，--replicas：启动的pod，--dry-run=true：干跑测试，-- /bin/sh 则是只运行shell，需要加上-it创建干跑测试
[root@linuxea ~]# kubectl run nginx-linuxea --image=nginx:1.15-alpine --port=8787 --replicas=1 --dry-run=true
deployment.apps/nginx-linuxea created (dry run)没有问题后在运行
[root@linuxea ~]# kubectl run nginx-linuxea --image=marksugar/nginx:1.14.a --port=8787 --replicas=1
deployment.apps/nginx-linuxea created创建之后可以使用kubectl get pods或者-o wide命令查看其中READY为数量(就绪了多少容器和有多少容器)，STATUS状态信息，RESTARTS是否重启，AGE启动时间，PODip地址，NODEnode机器名称。这里的ContainerCreating应该是在Pull镜像
[root@linuxea ~]# kubectl get pod
NAME READY STATUS RESTARTS AGE
nginx-linuxea-5786698598-4sdnt 0/1 ContainerCreating 0 6s
[root@linuxea ~]# kubectl get pod
NAME READY STATUS RESTARTS AGE
nginx-linuxea-5786698598-4sdnt 0/1 ContainerCreating 0 8s
[root@linuxea ~]# kubectl get pod
NAME READY STATUS RESTARTS AGE
nginx-linuxea-5786698598-4sdnt 1/1 Running 0 5m
[root@linuxea ~]# kubectl get pod -o wide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE
nginx-linuxea-5786698598-4sdnt 1/1 Running 0 6m 172.16.2.251 linuxea.node-2.com <none>
[root@linuxea ~]# 也可使用 kubectl get deployment查看其中NAME为指定名称，DESIRED期望的副本个数，CURRENT当前启动的个数，UP-TO-DATE最新状态个数，AVAILABLE就绪的个数，AGE启动时间
[root@linuxea ~]# kubectl get deployment
NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE
nginx-linuxea 1 1 1 1 7m
[root@linuxea ~]# 回到l linuxea.node-2.com，它的cni0的ip段则为pod的ip段，且各机器ip段可能并不一样(在pod中的ip一般和node的ip段在一个段)
[root@linuxea-vm-Node203 ~]# ip a|grep cni0
260: cni0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1450 qdisc noqueue state UP qlen 1000
inet 172.16.2.1/24 scope global cni0
270: veth3cbd5bb9@if3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1450 qdisc noqueue master cni0 state UP 当然，我们可以在kebernetes内进行访问这个pod,当然，在kubernetes之外是不能访问的
[root@linuxea ~]# curl 172.16.2.251
linuxea-nginx-linuxea-5786698598-4sdnt.com-127.0.0.1/8 172.16.2.251/24
[root@linuxea ~]# curl 172.16.2.251/linuxea.html
linuxea-nginx-linuxea-5786698598-4sdnt.com ▍ 95e6756979b83 ▍version number 1.0那如果要从kubernetes外进行访问，就需要其他组件，这样一来就不能用ip地址来访问量，因为我们知道这个Ip地址是随时可以发生改变的但控制器管理pod时，当pod发生改变不足够时，就会启动一个pod来替补,简单测试下get pod
[root@linuxea ~]# kubectl get pod
NAME READY STATUS RESTARTS AGE
nginx-linuxea-5786698598-4sdnt 1/1 Running 0 30m删除掉
[root@linuxea ~]# kubectl delete pods nginx-linuxea-5786698598-4sdnt
pod "nginx-linuxea-5786698598-4sdnt" deleted删除的同时就会run一个新的pod
[root@linuxea ~]# kubectl get pod
NAME READY STATUS RESTARTS AGE
nginx-linuxea-5786698598-48xjg 1/1 Running 0 42s- 注意
假如需要删除就需要删除deployment，它可能大概这样kubectl delete deployment nginx-linuxea
[root@linuxea ~]# kubectl get deployment
NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE
nginx-linuxea 1 1 1 1 22h
[root@linuxea ~]# kubectl delete deployment nginx-linuxea
deployment.extensions "nginx-linuxea" deleted使用-o wide时就能看见ip地址则发生改变，那么使用ip地址是不能够的，由此就需要一个固定端点来提供
[root@linuxea ~]# kubectl get pod -o wide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE
nginx-linuxea-5786698598-48xjg 1/1 Running 0 2m 172.16.3.12 linuxea.node-3.com <none>在curl
[root@linuxea ~]# curl 172.16.3.12
linuxea-nginx-linuxea-5786698598-48xjg.com-127.0.0.1/8 172.16.3.12/24expose
使用kubectl expose创建，其中指明 deployment nginx-linuxea， deployment nginx-linuxea是deployment中定义过的，而后指明名称，server端口，pod端口，以及协议
[root@linuxea ~]# kubectl expose deployment nginx-linuxea --name=www --port=8580 --target-port=80 --protocol=TCP
service/www exposedget services可以看到www服务名称，默认的type ClusterIP信息，10.105.49.157是动态生成的ip，也是10.96.0.1/12位掩码中的地址，,以及8580映射的暴露端口，此刻仍然可以通过10.105.49.157ip在集群内来访问，也就是kebernetes集群内被pod客户端来访问
[root@linuxea ~]# kubectl get services
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
kubernetes ClusterIP 10.96.0.1 <none> 443/TCP 22h
www ClusterIP 10.105.49.157 <none> 8580/TCP 2skube-dns
-
但是，kubernetes集群要在pod中访问，也不是用ip来访问，在此处可能用www服务名称进行访问的，如果要用www访问就需要使用集群的coredns,当然，也不能直接使用coredns服务的ip地址
[root@linuxea ~]# kubectl get pods -n kube-system -o wide|grep coredns coredns-78fcdf6894-gc64f 1/1 Running 0 22h 172.16.0.3 linuxea.master-1.com <none> coredns-78fcdf6894-l67zq 1/1 Running 0 22h 172.16.0.2 linuxea.master-1.com <none>一般使用kube-dns的ip地址10.96.0.10 即可
[root@linuxea ~]# kubectl get service -n kube-system NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE kube-dns ClusterIP 10.96.0.10 <none> 53/UDP,53/TCP 22h为了验证这一点，我们创建一个pod进行测试client-linuxea为pod的名称，images为alpine:3.9 ，replicas1个，-it进入交互式，restart=Never不重启
[root@linuxea ~]# kubectl run client-linuxea --image=busybox --replicas=1 -it --restart=Never此刻如果镜象没有被下载，会进行下载因此卡住，从而它的状态是ContainerCreating
[root@linuxea ~]# kubectl get pod NAME READY STATUS RESTARTS AGE client-linuxea 0/1 ContainerCreating 0 3s nginx-linuxea-5786698598-48xjg 1/1 Running 0 32m但下载完成，进入后可以看见他的nameserver已经在10.96.0.10
If you don't see a command prompt, try pressing enter. / # cat /etc/resolv.conf nameserver 10.96.0.10 search default.svc.cluster.local svc.cluster.local cluster.local options ndots:5而后使用wget www:8580即可访问的
/ # wget www:8580 Connecting to www:8580 (10.105.49.157:8580) index.html 100% |************************************************************************************************************************| 70 0:00:00 ETA / #或者这样
/ # wget -O - -q http://www:8580/ linuxea-nginx-linuxea-5786698598-48xjg.com-127.0.0.1/8 172.16.3.12/24回到pod外可使用dig加上pod内/etc/resolv.conf 提供的地址进行解析测试search default.svc.cluster.local svc.cluster.local cluster.local
[root@linuxea ~]# dig -t A www.default.svc.cluster.local @10.96.0.10 ; <<>> DiG 9.9.4-RedHat-9.9.4-61.el7 <<>> -t A www.default.svc.cluster.local @10.96.0.10 ;; global options: +cmd ;; Got answer: ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 64367 ;; flags: qr aa rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 0, ADDITIONAL: 1 ;; OPT PSEUDOSECTION: ; EDNS: version: 0, flags:; udp: 4096 ;; QUESTION SECTION: ;www.default.svc.cluster.local. IN A ;; ANSWER SECTION: www.default.svc.cluster.local. 5 IN A 10.105.49.157 ;; Query time: 1 msec ;; SERVER: 10.96.0.10#53(10.96.0.10) ;; WHEN: Thu Aug 23 15:48:47 BST 2018 ;; MSG SIZE rcvd: 103这里得到的地址是10.105.49.157，10.105.49.157就是www的service的ip那么现在我们访问的是位于www下的nginx-linuxea pod的nginx信息，如果此刻删除nginx pod，同时在进行访问
[root@linuxea ~]# kubectl get pod -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE client-linuxea 1/1 Running 0 8m 172.16.2.252 linuxea.node-2.com <none> nginx-linuxea-5786698598-48xjg 1/1 Running 0 12m 172.16.3.12 linuxea.node-3.com <none>[root@linuxea ~]# kubectl delete pod nginx-linuxea-5786698598-48xjg pod "nginx-linuxea-5786698598-48xjg" deleted[root@linuxea ~]# kubectl get pod -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE client-linuxea 1/1 Running 0 9m 172.16.2.252 linuxea.node-2.com <none> nginx-linuxea-5786698598-n89tw 1/1 Running 0 36s 172.16.3.13 linuxea.node-3.com <none> [root@linuxea ~]#通过关联选择器来进行关联访问
/ # wget -O - -q http://www:8580/ linuxea-nginx-linuxea-5786698598-48xjg.com-127.0.0.1/8 172.16.3.12/24 / # wget -O - -q http://www:8580/ linuxea-nginx-linuxea-5786698598-n89tw.com-127.0.0.1/8 172.16.3.13/24 / # wget -O - -q http://www:8580/由此可见，server可以为pod提供端点访问的效果而server在其中要么是iptables规则，要么就是ipvs规则使用 kubectl describe service www可以看到Labels标签中run标签的所有nginx-linuxea的pod资源其中Endpoints: 的ip如果删除会随之改变并且在使用kubectl edit svc www修改后会更新地址解析等信息，因为这些信息会动态反应到coredns中并更新
[root@linuxea ~]# kubectl describe service www Name: www Namespace: default Labels: run=nginx-linuxea Annotations: <none> Selector: run=nginx-linuxea Type: ClusterIP IP: 10.105.49.157 Port: <unset> 8580/TCP TargetPort: 80/TCP Endpoints: 172.16.3.13:80 Session Affinity: None Events: <none>如下:nginx-linuxea-5786698598-n89tw后的labels是有run=nginx-linuxea标签的
[root@linuxea ~]# kubectl get pod -o wide --show-labels NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE LABELS client-linuxea 1/1 Running 0 10m 172.16.2.252 linuxea.node-2.com <none> run=client-linuxea nginx-linuxea-5786698598-n89tw 1/1 Running 0 1m 172.16.3.13 linuxea.node-3.com <none> pod-template-hash=1342254154,run=nginx-linuxea动态扩展Pod
scale --replicas=10,扩展数量为10个，扩展到deployment是nginx-linuxea
[root@linuxea ~]# kubectl scale --replicas=10 deployment nginx-linuxea deployment.extensions/nginx-linuxea scaledget pods查看
[root@linuxea ~]# kubectl get pods NAME READY STATUS RESTARTS AGE client-linuxea 1/1 Running 0 15m nginx-linuxea-5786698598-9qttt 1/1 Running 0 39s nginx-linuxea-5786698598-cl4zb 1/1 Running 0 39s nginx-linuxea-5786698598-hb57b 1/1 Running 0 39s nginx-linuxea-5786698598-n89tw 1/1 Running 0 6m nginx-linuxea-5786698598-rr9d5 1/1 Running 0 39s nginx-linuxea-5786698598-sx7vl 1/1 Running 0 39s nginx-linuxea-5786698598-tgdq2 1/1 Running 0 39s nginx-linuxea-5786698598-x2zwp 1/1 Running 0 39s nginx-linuxea-5786698598-x9xxv 1/1 Running 0 39s nginx-linuxea-5786698598-z78l2 1/1 Running 0 39s之前的busybox中循环观察这个结果
/ # while true;do wget -O - -q www:8580;sleep 1 ;done linuxea-nginx-linuxea-5786698598-n89tw.com-127.0.0.1/8 172.16.3.13/24 wget: can't connect to remote host (10.105.49.157): Connection refused linuxea-nginx-linuxea-5786698598-sx7vl.com-127.0.0.1/8 172.16.2.253/24 linuxea-nginx-linuxea-5786698598-hb57b.com-127.0.0.1/8 172.16.3.15/24 linuxea-nginx-linuxea-5786698598-rr9d5.com-127.0.0.1/8 172.16.2.254/24 linuxea-nginx-linuxea-5786698598-tgdq2.com-127.0.0.1/8 172.16.2.2/24 linuxea-nginx-linuxea-5786698598-rr9d5.com-127.0.0.1/8 172.16.2.254/24 linuxea-nginx-linuxea-5786698598-x2zwp.com-127.0.0.1/8 172.16.2.3/24 linuxea-nginx-linuxea-5786698598-9qttt.com-127.0.0.1/8 172.16.3.14/24 linuxea-nginx-linuxea-5786698598-rr9d5.com-127.0.0.1/8 172.16.2.254/24 linuxea-nginx-linuxea-5786698598-tgdq2.com-127.0.0.1/8 172.16.2.2/24 linuxea-nginx-linuxea-5786698598-x2zwp.com-127.0.0.1/8 172.16.2.3/24 linuxea-nginx-linuxea-5786698598-rr9d5.com-127.0.0.1/8 172.16.2.254/24 linuxea-nginx-linuxea-5786698598-rr9d5.com-127.0.0.1/8 172.16.2.254/24 linuxea-nginx-linuxea-5786698598-tgdq2.com-127.0.0.1/8 172.16.2.2/24 linuxea-nginx-linuxea-5786698598-rr9d5.com-127.0.0.1/8 172.16.2.254/24 linuxea-nginx-linuxea-5786698598-rr9d5.com-127.0.0.1/8 172.16.2.254/24 linuxea-nginx-linuxea-5786698598-hb57b.com-127.0.0.1/8 172.16.3.15/24 linuxea-nginx-linuxea-5786698598-x2zwp.com-127.0.0.1/8 172.16.2.3/24由此可见，但动态扩展是有一定的效果，每次访问都飘到各个节点
动态缩减
[root@linuxea ~]# kubectl scale --replicas=3 deployment nginx-linuxea deployment.extensions/nginx-linuxea scaled已经处于Terminating状态
[root@linuxea ~]# kubectl get pods NAME READY STATUS RESTARTS AGE client-linuxea 1/1 Running 0 18m nginx-linuxea-5786698598-9qttt 1/1 Running 0 3m nginx-linuxea-5786698598-cl4zb 0/1 Terminating 0 3m nginx-linuxea-5786698598-hb57b 1/1 Running 0 3m nginx-linuxea-5786698598-n89tw 1/1 Running 0 9m nginx-linuxea-5786698598-rr9d5 0/1 Terminating 0 3m nginx-linuxea-5786698598-sx7vl 0/1 Terminating 0 3m nginx-linuxea-5786698598-tgdq2 0/1 Terminating 0 3m nginx-linuxea-5786698598-x2zwp 0/1 Terminating 0 3m而后
[root@linuxea ~]# kubectl get pods -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE client-linuxea 1/1 Running 0 18m 172.16.2.252 linuxea.node-2.com <none> nginx-linuxea-5786698598-9qttt 1/1 Running 0 4m 172.16.3.14 linuxea.node-3.com <none> nginx-linuxea-5786698598-hb57b 1/1 Running 0 4m 172.16.3.15 linuxea.node-3.com <none> nginx-linuxea-5786698598-n89tw 1/1 Running 0 9m 172.16.3.13 linuxea.node-3.com <none> [root@linuxea ~]#在观察busybox已经回到172.16.3.13，172.16.3.15，172.16.3.14上进行调度
linuxea-nginx-linuxea-5786698598-n89tw.com-127.0.0.1/8 172.16.3.13/24 linuxea-nginx-linuxea-5786698598-n89tw.com-127.0.0.1/8 172.16.3.13/24 linuxea-nginx-linuxea-5786698598-9qttt.com-127.0.0.1/8 172.16.3.14/24 linuxea-nginx-linuxea-5786698598-9qttt.com-127.0.0.1/8 172.16.3.14/24 linuxea-nginx-linuxea-5786698598-hb57b.com-127.0.0.1/8 172.16.3.15/24 linuxea-nginx-linuxea-5786698598-n89tw.com-127.0.0.1/8 172.16.3.13/24 linuxea-nginx-linuxea-5786698598-hb57b.com-127.0.0.1/8 172.16.3.15/24 linuxea-nginx-linuxea-5786698598-9qttt.com-127.0.0.1/8 172.16.3.14/24 linuxea-nginx-linuxea-5786698598-hb57b.com-127.0.0.1/8 172.16.3.15/24 linuxea-nginx-linuxea-5786698598-n89tw.com-127.0.0.1/8 172.16.3.13/24 linuxea-nginx-linuxea-5786698598-9qttt.com-127.0.0.1/8 172.16.3.14/24 linuxea-nginx-linuxea-5786698598-9qttt.com-127.0.0.1/8 172.16.3.14/24 linuxea-nginx-linuxea-5786698598-9qttt.com-127.0.0.1/8 172.16.3.14/24 linuxea-nginx-linuxea-5786698598-9qttt.com-127.0.0.1/8 172.16.3.14/24 linuxea-nginx-linuxea-5786698598-9qttt.com-127.0.0.1/8 172.16.3.14/24 linuxea-nginx-linuxea-5786698598-9qttt.com-127.0.0.1/8 172.16.3.14/24动态升级
这里指的是容器，并不是pod,也就是pod中的容器试图改变镜像的版本，换成新的镜像查看image的版本关注
Containers: nginx-linuxea:和
Image: marksugar/nginx:1.14.a如下：
[root@linuxea ~]# kubectl get pods NAME READY STATUS RESTARTS AGE client-linuxea 1/1 Running 0 19m nginx-linuxea-5786698598-9qttt 1/1 Running 0 5m nginx-linuxea-5786698598-hb57b 1/1 Running 0 5m nginx-linuxea-5786698598-n89tw 1/1 Running 0 10m [root@linuxea ~]# kubectl describe pods nginx-linuxea-5786698598-9qttt Name: nginx-linuxea-5786698598-9qttt Namespace: default Priority: 0 PriorityClassName: <none> Node: linuxea.node-3.com/10.10.240.146 Start Time: Thu, 23 Aug 2018 16:01:40 +0100 Labels: pod-template-hash=1342254154 run=nginx-linuxea Annotations: <none> Status: Running IP: 172.16.3.14 Controlled By: ReplicaSet/nginx-linuxea-5786698598 Containers: nginx-linuxea: Container ID: docker://f70bf85a100d5fa901e49ecc321467e06c76dcfa7fa35840b316b66033922fa3 Image: marksugar/nginx:1.14.a Image ID: docker-pullable://marksugar/nginx@sha256:103dd97b01c3283b56a587e7d95135f8fc410be1df36477d2d477a41f00daa59 Port: 8787/TCP Host Port: 0/TCP State: Running Started: Thu, 23 Aug 2018 16:01:41 +0100 Ready: True Restart Count: 0 Environment: <none> Mounts: /var/run/secrets/kubernetes.io/serviceaccount from default-token-l2x78 (ro) Conditions: Type Status Initialized True Ready True ContainersReady True PodScheduled True Volumes: default-token-l2x78: Type: Secret (a volume populated by a Secret) SecretName: default-token-l2x78 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 6m default-scheduler Successfully assigned default/nginx-linuxea-5786698598-9qttt to linuxea.node-3.com Normal Pulled 6m kubelet, linuxea.node-3.com Container image "marksugar/nginx:1.14.a" already present on machine Normal Created 6m kubelet, linuxea.node-3.com Created container Normal Started 6m kubelet, linuxea.node-3.com Started container在升级前，为了以便于区别，打开一个html，www:8580/linuxea.html的version number 是 1.0
/ # while true;do wget -O - -q www:8580/linuxea.html;sleep 1 ;done linuxea-nginx-linuxea-5786698598-hb57b.com ▍ fd4dee2e257d3 ▍version number 1.0 linuxea-nginx-linuxea-5786698598-n89tw.com ▍ 7044e9119475c ▍version number 1.0 linuxea-nginx-linuxea-5786698598-9qttt.com ▍ ea7f85d42da44 ▍version number 1.0 linuxea-nginx-linuxea-5786698598-n89tw.com ▍ 7044e9119475c ▍version number 1.0 linuxea-nginx-linuxea-5786698598-9qttt.com ▍ ea7f85d42da44 ▍version number 1.0升级到marksugar/nginx:1.14.b
[root@linuxea ~]# kubectl set image deployment nginx-linuxea nginx-linuxea=marksugar/nginx:1.14.b deployment.extensions/nginx-linuxea image updated而后kubectl rollout status deployment nginx-linuxea查看升级状态，这个过程快慢取决于镜像的大小，因为本地没有的话，是要去下载
[root@linuxea ~]# kubectl rollout status deployment nginx-linuxea Waiting for deployment "nginx-linuxea" rollout to finish: 1 out of 3 new replicas have been updated... Waiting for deployment "nginx-linuxea" rollout to finish: 1 out of 3 new replicas have been updated... Waiting for deployment "nginx-linuxea" rollout to finish: 1 out of 3 new replicas have been updated... Waiting for deployment "nginx-linuxea" rollout to finish: 2 out of 3 new replicas have been updated... Waiting for deployment "nginx-linuxea" rollout to finish: 2 out of 3 new replicas have been updated... Waiting for deployment "nginx-linuxea" rollout to finish: 1 old replicas are pending termination... Waiting for deployment "nginx-linuxea" rollout to finish: 1 old replicas are pending termination... Waiting for deployment "nginx-linuxea" rollout to finish: 1 old replicas are pending termination... deployment "nginx-linuxea" successfully rolled out在get pods
[root@linuxea ~]# kubectl get pods NAME READY STATUS RESTARTS AGE client-linuxea 1/1 Running 0 28m nginx-linuxea-795745985-fxvx2 1/1 Running 0 1m nginx-linuxea-795745985-tzj25 1/1 Running 0 1m nginx-linuxea-795745985-vp4nl 1/1 Running 0 1m验证更新的版本
[root@linuxea ~]# kubectl describe pods nginx-linuxea-795745985-fxvx2|grep image Normal Pulled 2m kubelet, linuxea.node-2.com Container image "marksugar/nginx:1.14.b" already present on machine此外，这里仍然可以观察，这个过程并没有就绪检测，会随意调度到旧的镜像和新的镜像，但新的升级完成，才会全部调度至升级好的镜像，在这里表现为version number 2.0
/ # while true;do wget -O - -q www:8580/linuxea.html;sleep 1 ;done linuxea-nginx-linuxea-5786698598-7kxf9.com ▍ 043a36f15a1ce ▍version number 1.0 linuxea-nginx-linuxea-5786698598-89kg9.com ▍ 86c44dfe63414 ▍version number 1.0 linuxea-nginx-linuxea-5786698598-96z6z.com ▍ 31d6e347eeb91 ▍version number 1.0 linuxea-nginx-linuxea-5786698598-7kxf9.com ▍ 043a36f15a1ce ▍version number 1.0 linuxea-nginx-linuxea-5786698598-7kxf9.com ▍ 043a36f15a1ce ▍version number 1.0 linuxea-nginx-linuxea-5786698598-89kg9.com ▍ 86c44dfe63414 ▍version number 1.0 linuxea-nginx-linuxea-5786698598-96z6z.com ▍ 31d6e347eeb91 ▍version number 1.0 linuxea-nginx-linuxea-5786698598-89kg9.com ▍ 86c44dfe63414 ▍version number 1.0 linuxea-nginx-linuxea-5786698598-96z6z.com ▍ 31d6e347eeb91 ▍version number 1.0 linuxea-nginx-linuxea-5786698598-89kg9.com ▍ 86c44dfe63414 ▍version number 1.0 linuxea-nginx-linuxea-5786698598-7kxf9.com ▍ 043a36f15a1ce ▍version number 1.0 linuxea-nginx-linuxea-795745985-fxvx2.com ▍ d1d0bb1b7f381 ▍version number 2.0 linuxea-nginx-linuxea-795745985-fxvx2.com ▍ d1d0bb1b7f381 ▍version number 2.0 linuxea-nginx-linuxea-795745985-fxvx2.com ▍ d1d0bb1b7f381 ▍version number 2.0 linuxea-nginx-linuxea-5786698598-89kg9.com ▍ 86c44dfe63414 ▍version number 1.0 linuxea-nginx-linuxea-5786698598-89kg9.com ▍ 86c44dfe63414 ▍version number 1.0 linuxea-nginx-linuxea-795745985-vp4nl.com ▍ c7d7ff91d4178 ▍version number 2.0 linuxea-nginx-linuxea-795745985-fxvx2.com ▍ d1d0bb1b7f381 ▍version number 2.0 linuxea-nginx-linuxea-5786698598-89kg9.com ▍ 86c44dfe63414 ▍version number 1.0 linuxea-nginx-linuxea-5786698598-89kg9.com ▍ 86c44dfe63414 ▍version number 1.0 linuxea-nginx-linuxea-5786698598-89kg9.com ▍ 86c44dfe63414 ▍version number 1.0 linuxea-nginx-linuxea-795745985-vp4nl.com ▍ c7d7ff91d4178 ▍version number 2.0 linuxea-nginx-linuxea-795745985-vp4nl.com ▍ c7d7ff91d4178 ▍version number 2.0 linuxea-nginx-linuxea-795745985-vp4nl.com ▍ c7d7ff91d4178 ▍version number 2.0 linuxea-nginx-linuxea-795745985-fxvx2.com ▍ d1d0bb1b7f381 ▍version number 2.0 linuxea-nginx-linuxea-795745985-fxvx2.com ▍ d1d0bb1b7f381 ▍version number 2.0 linuxea-nginx-linuxea-795745985-fxvx2.com ▍ d1d0bb1b7f381 ▍version number 2.0 linuxea-nginx-linuxea-795745985-fxvx2.com ▍ d1d0bb1b7f381 ▍version number 2.0 linuxea-nginx-linuxea-5786698598-89kg9.com ▍ 86c44dfe63414 ▍version number 1.0 linuxea-nginx-linuxea-5786698598-89kg9.com ▍ 86c44dfe63414 ▍version number 1.0 linuxea-nginx-linuxea-795745985-fxvx2.com ▍ d1d0bb1b7f381 ▍version number 2.0 linuxea-nginx-linuxea-795745985-fxvx2.com ▍ d1d0bb1b7f381 ▍version number 2.0 linuxea-nginx-linuxea-795745985-tzj25.com ▍ 129ec7e0fde65 ▍version number 2.0 linuxea-nginx-linuxea-795745985-tzj25.com ▍ 129ec7e0fde65 ▍version number 2.0 linuxea-nginx-linuxea-795745985-vp4nl.com ▍ c7d7ff91d4178 ▍version number 2.0 linuxea-nginx-linuxea-795745985-fxvx2.com ▍ d1d0bb1b7f381 ▍version number 2.0 linuxea-nginx-linuxea-795745985-tzj25.com ▍ 129ec7e0fde65 ▍version number 2.0已经更新完成
动态回滚
回滚到上一个版本。当然，此刻仍然可以就更新的版本进行重新升级到旧的版本，也可以使用rollout undo deployment。加入不指定版本，就回滚到上一个版本kubectl rollout undo deployment nginx-linuxea
[root@linuxea ~]# kubectl rollout undo deployment nginx-linuxea deployment.extensions/nginx-linuxea可以观察到ContainerCreating和Terminating状态
[root@linuxea ~]# kubectl get pods NAME READY STATUS RESTARTS AGE client-linuxea 1/1 Running 0 35m nginx-linuxea-5786698598-8qt5n 0/1 ContainerCreating 0 1s nginx-linuxea-5786698598-9w5z7 1/1 Running 0 2s nginx-linuxea-5786698598-xl9xp 1/1 Running 0 3s nginx-linuxea-795745985-fxvx2 1/1 Running 0 7m nginx-linuxea-795745985-tzj25 1/1 Terminating 0 7m nginx-linuxea-795745985-vp4nl 1/1 Terminating 0 7m状态切换
[root@linuxea ~]# kubectl get pods NAME READY STATUS RESTARTS AGE client-linuxea 1/1 Running 0 35m nginx-linuxea-5786698598-8qt5n 1/1 Running 0 6s nginx-linuxea-5786698598-9w5z7 1/1 Running 0 7s nginx-linuxea-5786698598-xl9xp 1/1 Running 0 8s nginx-linuxea-795745985-fxvx2 1/1 Terminating 0 7m nginx-linuxea-795745985-tzj25 1/1 Terminating 0 7m nginx-linuxea-795745985-vp4nl 1/1 Terminating 0 7m一旦完成就 恢复running状态
[root@linuxea ~]# kubectl get pods NAME READY STATUS RESTARTS AGE client-linuxea 1/1 Running 0 36m nginx-linuxea-5786698598-8qt5n 1/1 Running 0 1m nginx-linuxea-5786698598-9w5z7 1/1 Running 0 1m nginx-linuxea-5786698598-xl9xp 1/1 Running 0 1m在busybox中已经可以看到完成了回滚
/ # while true;do wget -O - -q www:8580/linuxea.html;sleep 1 ;done linuxea-nginx-linuxea-5786698598-8qt5n.com ▍ 06a064230a427 ▍version number 1.0 linuxea-nginx-linuxea-5786698598-8qt5n.com ▍ 06a064230a427 ▍version number 1.0 linuxea-nginx-linuxea-5786698598-8qt5n.com ▍ 06a064230a427 ▍version number 1.0 linuxea-nginx-linuxea-5786698598-9w5z7.com ▍ c68eabeb23dfd ▍version number 1.0 linuxea-nginx-linuxea-5786698598-8qt5n.com ▍ 06a064230a427 ▍version number 1.0 linuxea-nginx-linuxea-5786698598-8qt5n.com ▍ 06a064230a427 ▍version number 1.0那么这些kubernetes的网络上在iptbales或者ipvs中做的规则，kubernetes外是不能访问的。可以使用iptables -vnL -t nat可以进行查看这些iptables的规则
集群外访问
修改 type: ClusterIP为NodePort
apiVersion: v1 kind: Service metadata: creationTimestamp: 2018-08-23T00:57:53Z labels: run: nginx-linuxea name: www namespace: default resourceVersion: "43877" selfLink: /api/v1/namespaces/default/services/www uid: 90b89d89-a66f-11e8-9c95-88882fbd1028 spec: clusterIP: 10.102.11.107 ports: - port: 8580 protocol: TCP targetPort: 80 selector: run: nginx-linuxea sessionAffinity: None type: ClusterIP status: loadBalancer: {}修改完成
[root@linuxea ~]# kubectl edit svc www service/www edited修改完成后在进行get svc（svc是service的简写）可见8580的集群内的pod访问的server端口被暴露在31886端口
[root@linuxea ~]# kubectl get svc NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE kubernetes ClusterIP 10.96.0.1 <none> 443/TCP 23h www NodePort 10.105.49.157 <none> 8580:31886/TCP 49m [root@linuxea ~]#[root@linuxea ~]# kubectl get pod -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE client-linuxea 1/1 Running 0 51m 172.16.2.252 linuxea.node-2.com <none> nginx-linuxea-5786698598-8qt5n 1/1 Running 0 16m 172.16.1.14 linuxea.node-1.com <none> nginx-linuxea-5786698598-9w5z7 1/1 Running 0 16m 172.16.2.6 linuxea.node-2.com <none> nginx-linuxea-5786698598-xl9xp 1/1 Running 0 16m 172.16.3.19 linuxea.node-3.com <none>在expose中是动态的绑定的，这就以为这，可以在集群外部访问通过IP:31886访问node的每个节点，并且还自带负载均衡效果访问10.10.240.202
[ot@linuxea-vm-Node_10_0_1_61 ~/nginx]# while true;do curl 10.10.240.202:31886;sleep 1 ;done linuxea-nginx-linuxea-5786698598-8qt5n.com-127.0.0.1/8 172.16.1.14/24 linuxea-nginx-linuxea-5786698598-8qt5n.com-127.0.0.1/8 172.16.1.14/24 linuxea-nginx-linuxea-5786698598-xl9xp.com-127.0.0.1/8 172.16.3.19/24 linuxea-nginx-linuxea-5786698598-xl9xp.com-127.0.0.1/8 172.16.3.19/24 linuxea-nginx-linuxea-5786698598-8qt5n.com-127.0.0.1/8 172.16.1.14/24 linuxea-nginx-linuxea-5786698598-9w5z7.com-127.0.0.1/8 172.16.2.6/24[root@linuxea-vm-Node_10_0_1_61 ~/nginx]# while true;do curl 10.10.240.202:31886/linuxea.html;sleep 1 ;done linuxea-nginx-linuxea-5786698598-xl9xp.com ▍ ee508d36114eb ▍version number 1.0 linuxea-nginx-linuxea-5786698598-9w5z7.com ▍ c68eabeb23dfd ▍version number 1.0 linuxea-nginx-linuxea-5786698598-8qt5n.com ▍ 06a064230a427 ▍version number 1.0访问10.10.240.203
[root@linuxea-vm-Node_10_0_1_61 ~/nginx]# while true;do curl 10.10.240.203:31886;sleep 1 ;done linuxea-nginx-linuxea-5786698598-xl9xp.com-127.0.0.1/8 172.16.3.19/24 linuxea-nginx-linuxea-5786698598-xl9xp.com-127.0.0.1/8 172.16.3.19/24 linuxea-nginx-linuxea-5786698598-9w5z7.com-127.0.0.1/8 172.16.2.6/24 linuxea-nginx-linuxea-5786698598-8qt5n.com-127.0.0.1/8 172.16.1.14/24 linuxea-nginx-linuxea-5786698598-8qt5n.com-127.0.0.1/8 172.16.1.14/24 linuxea-nginx-linuxea-5786698598-8qt5n.com-127.0.0.1/8 172.16.1.14/24 linuxea-nginx-linuxea-5786698598-8qt5n.com-127.0.0.1/8 172.16.1.14/24[root@linuxea-vm-Node_10_0_1_61 ~/nginx]# while true;do curl 10.10.240.203:31886/linuxea.html;sleep 1 ;done linuxea-nginx-linuxea-5786698598-xl9xp.com ▍ ee508d36114eb ▍version number 1.0 linuxea-nginx-linuxea-5786698598-8qt5n.com ▍ 06a064230a427 ▍version number 1.0 linuxea-nginx-linuxea-5786698598-xl9xp.com ▍ ee508d36114eb ▍version number 1.0 linuxea-nginx-linuxea-5786698598-9w5z7.com ▍ c68eabeb23dfd ▍version number 1.0 linuxea-nginx-linuxea-5786698598-8qt5n.com ▍ 06a064230a427 ▍version number 1.0 linuxea-nginx-linuxea-5786698598-8qt5n.com ▍ 06a064230a427 ▍version number 1.0 linuxea-nginx-linuxea-5786698598-xl9xp.com ▍ ee508d36114eb ▍version number 1.0那么现在，如果在前面加上负载均衡，它的样子大概是这样
