Kubernetes中的基本概念和操作

前言

环境准备

使用minikube或者kubeadm搭建好的k8s集群实验环境。

关于在Centos7中搭建环境可以看我以前的一篇文章。

kubectl的基本使用

kubectl是访问k8s的一个重要的命令行工具。

命令自动补全

通过命令kubectl completion -h可以查看命令自动补全的配置。

接着在控制台执行该命令：需要注意的是不同的环境该命令是不同的，如果是在windows或者mac环境下搭建的k8s，都可以先通过kubectl completion -h查看下该命令在执行。

接着我们可以使用命令补全查看下k8s的节点：输入n后按table键会出现提示如下。

[vagrant@vagrant1 ~]$ kubectl get n
namespaces                         networkpolicies.networking.k8s.io  nodes

查看节点信息

[vagrant@vagrant1 ~]$ kubectl get node
NAME       STATUS   ROLES    AGE   VERSION
vagrant1   Ready    master   39h   v1.19.3
vagrant2   Ready    <none>   39h   v1.19.3

查看集群信息

创建k8s成功后，会在用户根目录创建一个.kube文件夹：

[vagrant@vagrant1 .kube]$ ll
total 8
drwxr-x---. 4 vagrant vagrant   35 Oct 20 04:38 cache
-rw-------. 1 vagrant vagrant 5557 Oct 20 09:20 config

config文件：我是用kubeadm搭建的k8s集群，所以改文件是kubectl和kubernetes通信的配置文件。

config文件如下：

config文件主要由三部分组成：

• clusters: clusters包含各集群端点数据，包含完整apiserver的url以及证书等 。

• contexts: contexts定义集群环境，包括使用的user以及进入的cluster，用name来标识。

• users: users用于向 kubernetes 集群进行身份验证的客户端凭据。

• 查看当前集群：

  [vagrant@vagrant1 .kube]$ kubectl config current-context
  kubernetes-admin@kubernetes
  

  查看config中的环境：

  [vagrant@vagrant1 .kube]$ kubectl config get-contexts
  CURRENT   NAME                          CLUSTER      AUTHINFO           NAMESPACE
  *         kubernetes-admin@kubernetes   kubernetes   kubernetes-admin
  

  设置一个新的集群环境，查看下当前config中的集群环境：

  [vagrant@vagrant1 .kube]$ kubectl config set-context new_context
  Context "new_context" created.
  [vagrant@vagrant1 .kube]$ kubectl config get-contexts
  CURRENT   NAME                          CLUSTER      AUTHINFO           NAMESPACE
  *         kubernetes-admin@kubernetes   kubernetes   kubernetes-admin
            new_context
  

  切换集群环境：

  [vagrant@vagrant1 .kube]$ kubectl config use-context kubernetes-admin@kubernetes
  Switched to context "kubernetes-admin@kubernetes".
  [vagrant@vagrant1 .kube]$ kubectl config get-contexts
  CURRENT   NAME                          CLUSTER      AUTHINFO           NAMESPACE
  *         kubernetes-admin@kubernetes   kubernetes   kubernetes-admin
            new_context
  

  删除集群环境：

  [vagrant@vagrant1 .kube]$ kubectl config delete-context new_context
  deleted context new_context from /home/vagrant/.kube/config
  [vagrant@vagrant1 .kube]$ kubectl config get-contexts
  CURRENT   NAME                          CLUSTER      AUTHINFO           NAMESPACE
  *         kubernetes-admin@kubernetes   kubernetes   kubernetes-admin
  

k8s的节点和标签

• 查看节点信息

  [vagrant@vagrant1 .kube]$ kubectl get node
  NAME       STATUS   ROLES    AGE   VERSION
  vagrant1   Ready    master   41h   v1.19.3
  vagrant2   Ready    <none>   40h   v1.19.3
  [vagrant@vagrant1 .kube]$ kubectl get node vagrant1
  NAME       STATUS   ROLES    AGE   VERSION
  vagrant1   Ready    master   41h   v1.19.3
  

  kubectl describe node查看节点详情信息：

  [vagrant@vagrant1 ~]$ kubectl describe node vagrant1
  Name:               vagrant1
  Roles:              master
  Labels:             beta.kubernetes.io/arch=amd64
                      beta.kubernetes.io/os=linux
                      kubernetes.io/arch=amd64
                      kubernetes.io/hostname=vagrant1
                      kubernetes.io/os=linux
                      node-role.kubernetes.io/master=
  Annotations:        kubeadm.alpha.kubernetes.io/cri-socket: /var/run/dockershim.sock
                      node.alpha.kubernetes.io/ttl: 0
                      volumes.kubernetes.io/controller-managed-attach-detach: true
  CreationTimestamp:  Tue, 20 Oct 2020 09:09:31 +0000
  Taints:             node-role.kubernetes.io/master:NoSchedule
  Unschedulable:      false
  Lease:
    HolderIdentity:  vagrant1
    AcquireTime:     <unset>
    RenewTime:       Thu, 22 Oct 2020 02:43:10 +0000
  Conditions:
    Type                 Status  LastHeartbeatTime                 LastTransitionTime                Reason                       Message
    ----                 ------  -----------------                 ------------------                ------                       -------
    NetworkUnavailable   False   Tue, 20 Oct 2020 09:32:51 +0000   Tue, 20 Oct 2020 09:32:51 +0000   WeaveIsUp                    Weave pod has set this
    MemoryPressure       False   Thu, 22 Oct 2020 02:39:47 +0000   Tue, 20 Oct 2020 09:09:26 +0000   KubeletHasSufficientMemory   kubelet has sufficient memory available
    DiskPressure         False   Thu, 22 Oct 2020 02:39:47 +0000   Tue, 20 Oct 2020 09:09:26 +0000   KubeletHasNoDiskPressure     kubelet has no disk pressure
    PIDPressure          False   Thu, 22 Oct 2020 02:39:47 +0000   Tue, 20 Oct 2020 09:09:26 +0000   KubeletHasSufficientPID      kubelet has sufficient PID available
    Ready                True    Thu, 22 Oct 2020 02:39:47 +0000   Tue, 20 Oct 2020 09:33:02 +0000   KubeletReady                 kubelet is posting ready status
  Addresses:
    InternalIP:  10.0.2.15
    Hostname:    vagrant1
  Capacity:
    cpu:                2
    ephemeral-storage:  41921540Ki
    hugepages-2Mi:      0
    memory:             1881936Ki
    pods:               110
  Allocatable:
    cpu:                2
    ephemeral-storage:  38634891201
    hugepages-2Mi:      0
    memory:             1779536Ki
    pods:               110
  System Info:
    Machine ID:                 cbbaf9d37827a2459390178b0258c55f
    System UUID:                CBBAF9D3-7827-A245-9390-178B0258C55F
    Boot ID:                    425bb8b3-8659-4bdb-bf2b-331495fc64e4
    Kernel Version:             3.10.0-1127.19.1.el7.x86_64
    OS Image:                   CentOS Linux 7 (Core)
    Operating System:           linux
    Architecture:               amd64
    Container Runtime Version:  docker://19.3.13
    Kubelet Version:            v1.19.3
    Kube-Proxy Version:         v1.19.3
  PodCIDR:                      172.100.0.0/24
  PodCIDRs:                     172.100.0.0/24
  Non-terminated Pods:          (8 in total)
    Namespace                   Name                                CPU Requests  CPU Limits  Memory Requests  Memory Limits  AGE
    ---------                   ----                                ------------  ----------  ---------------  -------------  ---
    kube-system                 coredns-6d56c8448f-7kn5f            100m (5%)     0 (0%)      70Mi (4%)        170Mi (9%)     41h
    kube-system                 coredns-6d56c8448f-lqw9t            100m (5%)     0 (0%)      70Mi (4%)        170Mi (9%)     41h
    kube-system                 etcd-vagrant1                       0 (0%)        0 (0%)      0 (0%)           0 (0%)         41h
    kube-system                 kube-apiserver-vagrant1             250m (12%)    0 (0%)      0 (0%)           0 (0%)         41h
    kube-system                 kube-controller-manager-vagrant1    200m (10%)    0 (0%)      0 (0%)           0 (0%)         41h
    kube-system                 kube-proxy-5hwx8                    0 (0%)        0 (0%)      0 (0%)           0 (0%)         41h
    kube-system                 kube-scheduler-vagrant1             100m (5%)     0 (0%)      0 (0%)           0 (0%)         41h
    kube-system                 weave-net-bwgwf                     100m (5%)     0 (0%)      200Mi (11%)      0 (0%)         41h
  Allocated resources:
    (Total limits may be over 100 percent, i.e., overcommitted.)
    Resource           Requests     Limits
    --------           --------     ------
    cpu                850m (42%)   0 (0%)
    memory             340Mi (19%)  340Mi (19%)
    ephemeral-storage  0 (0%)       0 (0%)
    hugepages-2Mi      0 (0%)       0 (0%)
  Events:              <none>
  

  也可以通过kubectl get node加上-o指令显示节点信息：

  • kubectl get node -o wide查看节点较多（相比于 kubectl get node）信息:

    [vagrant@vagrant1 ~]$ kubectl get node -o wide
    NAME       STATUS   ROLES    AGE   VERSION   INTERNAL-IP   EXTERNAL-IP   OS-IMAGE                KERNEL-VERSION                CONTAINER-RUNTIME
    vagrant1   Ready    master   41h   v1.19.3   10.0.2.15     <none>        CentOS Linux 7 (Core)   3.10.0-1127.19.1.el7.x86_64   docker://19.3.13
    vagrant2   Ready    <none>   40h   v1.19.3   10.0.2.15     <none>        CentOS Linux 7 (Core)   3.10.0-1127.19.1.el7.x86_64   docker://19.3.13
    

  • kubectl get node -o yaml以yaml文件形式查看节点详情:

  • kubectl get node vagrant1 -o json以json形式查看节点详情:
    

• 查看节点标签信息

  通过kubectl describe node我们可以看到最上边有键值对的label信息，这些就是节点的标签，当我们查询节点的时候可以通过标签来过滤。

  查看标签信息：

  [vagrant@vagrant1 ~]$ kubectl get nodes --show-labels
  NAME       STATUS   ROLES    AGE   VERSION   LABELS
  vagrant1   Ready    master   41h   v1.19.3   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=vagrant1,kubernetes.io/os=linux,node-role.kubernetes.io/master=
  vagrant2   Ready    <none>   41h   v1.19.3   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=vagrant2,kubernetes.io/os=linux
  

  更新标签

  kubectl label node vagrant1 role=master给vagrant1节点设置标签role=master, 在查看下vagrant1的label信息发现多了一个role=master的键值对：

  [vagrant@vagrant1 ~]$ kubectl label node vagrant1 role=master
  node/vagrant1 labeled
  
  [vagrant@vagrant1 ~]$ kubectl get node vagrant1 --show-labels
  NAME       STATUS   ROLES    AGE   VERSION   LABELS
  vagrant1   Ready    master   41h   v1.19.3   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,kubernetes.io/arch=amd64,kubernetes.io/hostname=vagrant1,kubernetes.io/os=linux,node-role.kubernetes.io/master=,role=master
  [vagrant@vagrant1 ~]$
  

  查看节点信息的时候可以看到把vagrant2节点是没有role的：

  [vagrant@vagrant1 ~]$ kubectl get node
  NAME       STATUS   ROLES    AGE   VERSION
  vagrant1   Ready    master   42h   v1.19.3
  vagrant2   Ready    <none>   41h   v1.19.3
  

  可以通过给vagrant2增加一个role的label显示角色信息：

  [vagrant@vagrant1 ~]$ kubectl label node vagrant2 node-role.kubernetes.io/worker1=
  node/vagrant2 labeled
  
  [vagrant@vagrant1 ~]$ kubectl get node
  NAME       STATUS   ROLES     AGE   VERSION
  vagrant1   Ready    master    42h   v1.19.3
  vagrant2   Ready    worker1   41h   v1.19.3
  

  删除标签：

  [vagrant@vagrant1 ~]$ kubectl label node vagrant1 role-
  node/vagrant1 labeled
  

k8s调度的最小单位pod

kubelet、Docker、节点、Pod、容器之间的关系图如下：

kueblet服务运行在Docker服务上，一个Node节点包含多个Pod，一个Pod包含多个容器和volume。

• Pod定义

  • 一个或者一组应用容器，它们分享资源（比如volume）

  • 分享相同命名空间（如网络空间）的容器

  • Pod是k8s中最小的调度单位

• 创建Pod

  可以通过yaml文件创建Pod，比如说创建一个包含nginx和busybox的Pod：

  [vagrant@vagrant1 vagrant]$ cat nginx_busybox.yml
  apiVersion: v1
  kind: Pod
  metadata:
    name: nginx-busybox
  spec:
    containers:
    - name: nginx
      image: nginx
      ports:
      - containerPort: 80
    - name: busybox
      image: busybox
      command: ["/bin/sh"]
      args: ["-c", "while true; do echo hello; sleep 10; done"]
  

  执行kubectl create -f nginx_busybox.yml创建Pod, -f 指令为指定yaml文件：

  [vagrant@vagrant1 vagrant]$ kubectl  create -f nginx_busybox.yml
  pod/nginx-busybox created
  

• 查看Pod

  [vagrant@vagrant1 vagrant]$ kubectl get pod
  NAME            READY   STATUS    RESTARTS   AGE
  nginx-busybox   2/2     Running   0          106s
  

  • NAME: 为在yaml文件中指定的name。

  • READY: 分母2表示在这个Pod中有两个容器，分子2表示这两个容器都运行正常。

  • STATUS：表示Pod的运行状态，Running表示该Pod是正常运行的。

  • AGE：Pod运行时间。

  • 查看Pod详情：

    [vagrant@vagrant1 vagrant]$ kubectl describe pod nginx-busybox
    Name:         nginx-busybox
    Namespace:    default
    Priority:     0
    Node:         vagrant2/10.0.2.15
    Start Time:   Thu, 22 Oct 2020 03:45:07 +0000
    Labels:       <none>
    Annotations:  <none>
    Status:       Running
    IP:           10.32.0.2
    IPs:
      IP:  10.32.0.2
    Containers:
      nginx:
        Container ID:   docker://78137f996dc9b2a003b6f65502c5b6f3c58390d3cf99055161fe06b8ed9010ab
        Image:          nginx
        Image ID:       docker-pullable://nginx@sha256:ed7f815851b5299f616220a63edac69a4cc200e7f536a56e421988da82e44ed8
        Port:           80/TCP
        Host Port:      0/TCP
        State:          Running
          Started:      Thu, 22 Oct 2020 03:45:25 +0000
        Ready:          True
        Restart Count:  0
        Environment:    <none>
        Mounts:
          /var/run/secrets/kubernetes.io/serviceaccount from default-token-mqsmr (ro)
      busybox:
        Container ID:  docker://53ad2c72c74fe4ddb7c605743a7885066dbbe618fe13c0099624ec6bb252be5e
        Image:         busybox
        Image ID:      docker-pullable://busybox@sha256:a9286defaba7b3a519d585ba0e37d0b2cbee74ebfe590960b0b1d6a5e97d1e1d
        Port:          <none>
        Host Port:     <none>
        Command:
          /bin/sh
        Args:
          -c
          while true; do echo hello; sleep 10; done
        State:          Running
          Started:      Thu, 22 Oct 2020 03:45:35 +0000
        Ready:          True
        Restart Count:  0
        Environment:    <none>
        Mounts:
          /var/run/secrets/kubernetes.io/serviceaccount from default-token-mqsmr (ro)
    Conditions:
      Type              Status
      Initialized       True
      Ready             True
      ContainersReady   True
      PodScheduled      True
    Volumes:
      default-token-mqsmr:
        Type:        Secret (a volume populated by a Secret)
        SecretName:  default-token-mqsmr
        Optional:    false
    QoS Class:       BestEffort
    Node-Selectors:  <none>
    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  8m4s   default-scheduler  Successfully assigned default/nginx-busybox to vagrant2
      Normal  Pulling    8m4s   kubelet            Pulling image "nginx"
      Normal  Pulled     7m47s  kubelet            Successfully pulled image "nginx" in 16.874039181s
      Normal  Created    7m47s  kubelet            Created container nginx
      Normal  Started    7m47s  kubelet            Started container nginx
      Normal  Pulling    7m47s  kubelet            Pulling image "busybox"
      Normal  Pulled     7m37s  kubelet            Successfully pulled image "busybox" in 9.398186417s
      Normal  Created    7m37s  kubelet            Created container busybox
      Normal  Started    7m37s  kubelet            Started container busybox
    

    以更精简的方式查看详情：

    [vagrant@vagrant1 vagrant]$ kubectl get pod nginx-busybox -o wide
    NAME            READY   STATUS    RESTARTS   AGE     IP          NODE       NOMINATED NODE   READINESS GATES
    nginx-busybox   2/2     Running   0          9m28s   10.32.0.2   vagrant2   <none>           <none>
    

  • 使用exec进入Pod内部：

    [vagrant@vagrant1 vagrant]$ kubectl exec nginx-busybox  -- date
    Defaulting container name to nginx.
    Use 'kubectl describe pod/nginx-busybox -n default' to see all of the containers in this pod.
    Thu Oct 22 05:41:40 UTC 2020
    

    如果不指定容器默认进入的是第一个容器，也就是yaml文件的第一个容器。

    使用-c 指定容器：

    [vagrant@vagrant1 vagrant]$ kubectl exec nginx-busybox -c busybox -- date
    Thu Oct 22 05:43:21 UTC 2020
    

    使用-it 进入容器后台：

    [vagrant@vagrant1 vagrant]$ kubectl exec -it nginx-busybox -c busybox -- sh
    / #
    

• 删除Pod

  [vagrant@vagrant1 vagrant]$ kubectl delete -f nginx_busybox.yml
  pod "nginx-busybox" deleted
  

Namespace命名空间

• Namespace命名空间用于不同team, 不同project之间的隔离。
• 在不同的命名空间中，各种资源的名字是相互独立，比如可以具有相同名称的pod存在。

查看命名空间

[vagrant@vagrant1 vagrant]$ kubectl get namespaces
NAME              STATUS   AGE
default           Active   44h
kube-node-lease   Active   44h
kube-public       Active   44h
kube-system       Active   44h

查看命名空间下的Pod

[vagrant@vagrant1 ~]$ kubectl get pod --namespace kube-system
NAME                               READY   STATUS    RESTARTS   AGE
coredns-6d56c8448f-7kn5f           1/1     Running   0          2d15h
coredns-6d56c8448f-lqw9t           1/1     Running   0          2d15h
etcd-vagrant1                      1/1     Running   0          2d15h
kube-apiserver-vagrant1            1/1     Running   2          2d15h
kube-controller-manager-vagrant1   0/1     Running   6          2d15h
kube-proxy-5hwx8                   1/1     Running   0          2d15h
kube-proxy-7b852                   1/1     Running   0          2d14h
kube-scheduler-vagrant1            0/1     Running   4          2d15h
weave-net-89zh8                    2/2     Running   0          2d14h
weave-net-bwgwf                    2/2     Running   0          2d14h

创建命名空间

[vagrant@vagrant1 ~]$ kubectl create namespace demo
namespace/demo created
[vagrant@vagrant1 ~]$ kubectl get namespaces | grep demo
demo              Active   18s

创建两个nginx Pod一个位于默认的命名空间，一个位于新创建的demo命名空间：

• 位于默认命名空间的yaml文件：

  apiVersion: v1
  kind: Pod
  metadata:
    name: nginx
  spec:
    containers:
    - name: nginx
      image: nginx
    ports:
    - containerPort: 80
  

• 位于demo命名空间的yaml文件：

  apiVersion: v1
  kind: Pod
  metadata:
    name: nginx
    namespace: demo
  spec:
    containers:
    - name: nginx
      image: nginx
      ports:
      - containerPort: 80
  

• 执行创建命令：

  [vagrant@vagrant1 vagrant]$ kubectl create -f nginx.yml
  pod/nginx created
  [vagrant@vagrant1 vagrant]$ kubectl create -f nginx_demo.yml
  pod/nginx created
  

• 查看所有命名空间的Pod:

  [vagrant@vagrant1 vagrant]$ kubectl get pod --all-namespaces
  NAMESPACE     NAME                               READY   STATUS    RESTARTS   AGE
  default       nginx                              1/1     Running   0          68s
  demo          nginx                              1/1     Running   0          60s
  kube-system   coredns-6d56c8448f-7kn5f           1/1     Running   0          2d15h
  kube-system   coredns-6d56c8448f-lqw9t           1/1     Running   0          2d15h
  kube-system   etcd-vagrant1                      1/1     Running   0          2d15h
  kube-system   kube-apiserver-vagrant1            1/1     Running   2          2d15h
  kube-system   kube-controller-manager-vagrant1   1/1     Running   6          2d15h
  kube-system   kube-proxy-5hwx8                   1/1     Running   0          2d15h
  kube-system   kube-proxy-7b852                   1/1     Running   0          2d14h
  kube-system   kube-scheduler-vagrant1            1/1     Running   4          2d15h
  kube-system   weave-net-89zh8                    2/2     Running   0          2d14h
  kube-system   weave-net-bwgwf                    2/2     Running   0          2d15h
  

  可以发现有两个叫做nginx的Pod，但是运行在不同的命名空间中。

• 删除其中一个nginx Pod，需要指定命名空间删除，否则删除的是默认命名空间下的nginx Pod。

删除命名空间

连带命名空间下的所有内容都会删除。

[vagrant@vagrant1 vagrant]$ kubectl delete namespaces demo
namespace "demo" deleted

创建自己的context

• 如果是多人在同一个context下使用，可能会在不同的命名空间下创建相同名称得到Pod，这样不同人在操作这些Pod的时候都需要指定命名空间，否则可能会对别人创建的命名空间的相同名称的Pod进行修改。
• 可以通过创建自己的context，具有相同的集群和用户，但是命名空间不同的context，在自己的context下操作就不会影响别人。

创建context

[vagrant@vagrant1 vagrant]$ kubectl config set-context demo --user=kubernetes-admin --cluster=kubernetes --namespace=demo
Context "demo" created.

切换context至demo，在该context中创建Pod，默认的命名空间名称就是demo，也就是在创建context的时候指定的命名空间：

[vagrant@vagrant1 vagrant]$ kubectl config use-context demo
Switched to context "demo".

[vagrant@vagrant1 vagrant]$ kubectl create namespace demo
namespace/demo created

[vagrant@vagrant1 vagrant]$ kubectl create -f nginx.yml
pod/nginx created

[vagrant@vagrant1 vagrant]$ kubectl get pod -A
NAMESPACE     NAME                               READY   STATUS              RESTARTS   AGE
demo          nginx                              0/1     ContainerCreating   0          10s
kube-system   coredns-6d56c8448f-7kn5f           1/1     Running             0          2d16h
kube-system   coredns-6d56c8448f-lqw9t           1/1     Running             0          2d16h
kube-system   etcd-vagrant1                      1/1     Running             0          2d16h
kube-system   kube-apiserver-vagrant1            1/1     Running             2          2d16h
kube-system   kube-controller-manager-vagrant1   1/1     Running             6          2d16h
kube-system   kube-proxy-5hwx8                   1/1     Running             0          2d16h
kube-system   kube-proxy-7b852                   1/1     Running             0          2d15h
kube-system   kube-scheduler-vagrant1            1/1     Running             4          2d16h
kube-system   weave-net-89zh8                    2/2     Running             0          2d15h
kube-system   weave-net-bwgwf                    2/2     Running             0          2d15h

删除context

删除context前先切换context：

[vagrant@vagrant1 vagrant]$ kubectl config use-context kubernetes-admin@kubernetes
Switched to context "kubernetes-admin@kubernetes".

[vagrant@vagrant1 vagrant]$ kubectl config delete-context demo
deleted context demo from /home/vagrant/.kube/config

Controller和Deployment

k8s的架构图如下：

一个master节点管理多个worker节点，master节点主要由以下4个模块组成：

• etcd: 分布式键值对存储的数据库
• API Server: 提供访问k8s的接口服务
• Scheduler: 根据算法调度Node节点
• Controller Manager：使当前状态等于期望状态，比如说期望Node1上的一个服务正常运行，但是Node1节点宕机了，Controller就会在别的节点上运行该服务，或者重启Node1节点，已达到期望状态。

Deployment: 在一个Deployment对象中描述一个期望状态，通过controller改变当前状态至期望状态。

deployment更新演示：

• 首先创建三个yaml文件如下，当kind指定为Deployment时，创建的Pod就会被Deployment管理使当前状态达到期望状态：

  nginx_deployment.yml:

  apiVersion: apps/v1
  kind: Deployment
  metadata:
    name: nginx-deployment
  spec:
    selector:
      matchLabels:
        app: nginx
    replicas: 2
    template:
      metadata:
        labels:
          app: nginx
      spec:
        containers:
        - name: nginx
          image: nginx:1.7.9
          ports:
          - containerPort: 80
  

  nginx_deployment_update.yml: nginx 版本从1.7.9升级到1.8

  apiVersion: apps/v1
  kind: Deployment
  metadata:
    name: nginx-deployment
  spec:
    selector:
      matchLabels:
        app: nginx
    replicas: 2
    template:
      metadata:
        labels:
          app: nginx
      spec:
        containers:
        - name: nginx
          image: nginx:1.8
          ports:
          - containerPort: 80
  

  nginx_deployment_scale.yml: nginx scale扩展为4个

  apiVersion: apps/v1
  kind: Deployment
  metadata:
    name: nginx-deployment
  spec:
    selector:
      matchLabels:
        app: nginx
    replicas: 4
    template:
      metadata:
        labels:
          app: nginx
      spec:
        containers:
        - name: nginx
          image: nginx:1.8
          ports:
          - containerPort: 80
  

• 接着使用nginx_deployment.yml文件创建Pod:

  [vagrant@vagrant1 vagrant]$ kubectl create -f nginx_deployment.yml
  deployment.apps/nginx-deployment created
  

  查看deployment，当前状态和期望状态：

  [vagrant@vagrant1 vagrant]$ kubectl get deployment -o wide
  NAME               READY   UP-TO-DATE   AVAILABLE   AGE     CONTAINERS   IMAGES        SELECTOR
  nginx-deployment   2/2     2            2           2m35s   nginx        nginx:1.7.9   app=nginx
  

  READY字段分母表示有两个当前状态运行着，分子表示两个期望状态运行着，也就是说当前状态和期望状态一致。

  查看pod详情：

  [vagrant@vagrant1 vagrant]$ kubectl get pod -l app=nginx
  NAME                                READY   STATUS    RESTARTS   AGE
  nginx-deployment-5d59d67564-4tjsj   1/1     Running   0          8m50s
  nginx-deployment-5d59d67564-q5gr9   1/1     Running   0          8m50s
  

  可以发现有两个nginx Pod运行着，接着删除其中一个, 删除后立即查看pod详情发现还是两个nginx-deployment运行着:

  [vagrant@vagrant1 vagrant]$ kubectl delete pod nginx-deployment-5d59d67564-4tjsj
  pod "nginx-deployment-5d59d67564-4tjsj" deleted
  
  [vagrant@vagrant1 vagrant]$ kubectl get pod -l app=nginx
  NAME                                READY   STATUS    RESTARTS   AGE
  nginx-deployment-5d59d67564-c85l6   1/1     Running   0          7s
  nginx-deployment-5d59d67564-q5gr9   1/1     Running   0          11m
  

  这就是deployment保证期望状态不变，又创建了一个相同的pod。

• 现在将部署好的nginx-deployment 服务的nginx更新至1.8版本：

  使用命令kubectl apply -f nginx_deployment_update.yml，apply 包括更新和创建，create只有创建：

  [vagrant@vagrant1 vagrant]$ kubectl get deployments.apps  nginx-deployment -o wide
  NAME               READY   UP-TO-DATE   AVAILABLE   AGE   CONTAINERS   IMAGES        SELECTOR
  nginx-deployment   2/2     2            2           18m   nginx        nginx:1.7.9   app=nginx
  
  [vagrant@vagrant1 vagrant]$ kubectl apply -f nginx_deployment_update.yml
  Warning: kubectl apply should be used on resource created by either kubectl create --save-config or kubectl apply
  deployment.apps/nginx-deployment configured
  
  [vagrant@vagrant1 vagrant]$ kubectl get deployments.apps nginx-deployment -o wide
  NAME               READY   UP-TO-DATE   AVAILABLE   AGE   CONTAINERS   IMAGES      SELECTOR
  nginx-deployment   2/2     2            2           22m   nginx        nginx:1.8   app=nginx
  

  可以发现nginx从1.7.9变更为了1.8

• 现在又想将nginx-deployment 服务扩展为4个：

  和上边一样执行命令kubectl apply -f nginx_deployment_scale.yml

  [vagrant@vagrant1 vagrant]$ kubectl apply -f nginx_deployment_scale.yml
  deployment.apps/nginx-deployment configured
  

  查看下deployment详情和pod详情，发现扩展为了4个：

  [vagrant@vagrant1 vagrant]$ kubectl get deployments.apps nginx-deployment -o wide
  NAME               READY   UP-TO-DATE   AVAILABLE   AGE   CONTAINERS   IMAGES      SELECTOR
  nginx-deployment   4/4     4            4           26m   nginx        nginx:1.8   app=nginx
  
  [vagrant@vagrant1 vagrant]$ kubectl get pod -l app=nginx -o wide
  NAME                                READY   STATUS    RESTARTS   AGE     IP          NODE       NOMINATED NODE   READINESS GATES
  nginx-deployment-64c9d67564-4b4zv   1/1     Running   0          51s     10.32.0.2   vagrant2   <none>           <none>
  nginx-deployment-64c9d67564-hzv8q   1/1     Running   0          51s     10.32.0.5   vagrant2   <none>           <none>
  nginx-deployment-64c9d67564-m2x5f   1/1     Running   0          5m32s   10.32.0.4   vagrant2   <none>           <none>
  nginx-deployment-64c9d67564-nqm67   1/1     Running   0          4m44s   10.32.0.3   vagrant2   <none>           <none>
  

• 也可以不使用yaml文件更新，直接使用kubectl edit deployment命令进行更新, 将会进入一个vi编辑界面：

  [vagrant@vagrant1 vagrant]$ kubectl edit deployment nginx-deployment
  

退出后会有提示：

接着查看下nginx-deployment 服务的Dployment状态和Pod详情，发现都变为了3：

Replicaset在Deployment更新中的作用

当创建一个deployment类型的pod后，当服务进行了更新都会有历史事件记录保存。

使用命令kubectl describe deployments.apps nginx-deployment-test, 查看下刚创建的nginx-deployment-test服务的详情：

现在使用kubectl scale命令将服务的scale扩展更新为3个：

[vagrant@vagrant1 vagrant]$ kubectl scale --current-replicas=4 --replicas=3 deployment/nginx-deployment-test

在查看下nginx-deployment详情：

再将image更新到1.9.1，在查看下deployment详情：

历史回滚

在更新后如果想要回滚到更新前的版本可以使用命令kubectl rollout。

• 首先使用命令kubectl rollout history查看下版本历史：

  [vagrant@vagrant1 vagrant]$ kubectl rollout history deployment nginx-deployment-test
  deployment.apps/nginx-deployment-test
  REVISION  CHANGE-CAUSE
  1         <none>
  2         <none>
  

  使用命令kubectl rollout history deployment nginx-deployment-test --revision 1可以查看版本1的记录（也就是nginx版本为1.7.9），同样的可以查看版本2的记录（最新版本nginx为1.9.1），最多记录两个版本也就是当前版本和上次的版本，所以这里看不到scale为4的版本：

  [vagrant@vagrant1 vagrant]$ kubectl rollout history deployment nginx-deployment-test --revision 1
  deployment.apps/nginx-deployment-test with revision #1
  Pod Template:
    Labels:       app=nginx
          pod-template-hash=5d59d67564
    Containers:
     nginx:
      Image:      nginx:1.7.9
      Port:       80/TCP
      Host Port:  0/TCP
      Environment:        <none>
      Mounts:     <none>
    Volumes:      <none>
  
  [vagrant@vagrant1 vagrant]$ kubectl rollout history deployment nginx-deployment-test --revision 2
  deployment.apps/nginx-deployment-test with revision #2
  Pod Template:
    Labels:       app=nginx
          pod-template-hash=69c44dfb78
    Containers:
     nginx:
      Image:      nginx:1.9.1
      Port:       80/TCP
      Host Port:  0/TCP
      Environment:        <none>
      Mounts:     <none>
    Volumes:      <none>
  

• 使用kubectl rollout undo可以回滚到上个版本：

  [vagrant@vagrant1 vagrant]$ kubectl rollout undo deployment nginx-deployment-test
  deployment.apps/nginx-deployment-test rolled back
  

  查看下当前deployment详情，发现nginx版本变为了1.7.9：

  [vagrant@vagrant1 vagrant]$ kubectl get deployments.apps nginx-deployment-test -o wide
  NAME                    READY   UP-TO-DATE   AVAILABLE   AGE   CONTAINERS   IMAGES        SELECTOR
  nginx-deployment-test   3/3     3            3           57m   nginx        nginx:1.7.9   app=nginx