其他分享
首页 > 其他分享> > 02-Kubenetes资源

02-Kubenetes资源

作者：互联网

目录

Kubenetes资源

常用资源对象

标签labels

labels 与 资源之间是多对多的关系

标签的定义一般从以下几个角度定义

标签格式：

key=value  
key: 字母 数字 _ .   
value：只能以字母数字开头及结尾

通过标签过滤

kubectl get pods -l <labels>

查看所有标签

kubectl get pods --show-labels

打标签

 kubectl label [--overwrite] (-f FILENAME | TYPE NAME) KEY_1=VAL_1 ... KEY_N=VAL_N
[--resource-version=version] [options]

标签选择器

许多资源支持内嵌字段

创建资源的方式

apiserver仅接受JSON格式的资源定义；

yaml格式提供配置清单， apiserver可自动将其转为json格式，然后提交

大部分的资源的配置清单，主要5个一级资源

使用explain 查看定义

例如：

kubectl explain pods.metadata
kubectl explain pods.spec.containers

Pod

k8s管理的最小单位，一个pod中可以有多个contaiers 例如

apiVersion: v1
kind: Pod
metadata:
  name: nginx
spec:
  containers:
  - name: nginx
    image: nginx:1.7.9
    ports:
    - containerPort: 80
    readinessProbe:
      httpGet:
        port: 80
      initialDelaySeconds: 2
      periodSeconds: 3
    livenessProbe:
      httpGet:
        port: 80
      initialDelaySeconds: 2
      periodSeconds: 3

  - name: busybox
    image: busybox:latest
    imagePullPolicy: IfNotPresent
    command: ['/bin/sh','-c','ping','www.baidu.com']
  nodeSelector:
    kubernetes.io/hostname: 192.168.0.165

pods.spec.containers 必须

- name <string>
  image <string>
  imagePullPolicy     <string>  Always, Never, IfNotPresent. 
  * Defaults to Always if :latest tag is specified, or IfNotPresent otherwise. Cannot be updated. (优化点)
 
  ports    <[]Object> 
  * 仅仅是说明性的
  - containerPort <integer> -required-
    hostIP   0.0.0.0
    hostPort  必须与containerPort 相同，大部分不需要定义该项
    name   名称
    protocol 默认TCP
command      <[]string>
args         <[]string>

- command 会覆盖镜像中的Entrypoint 与 command
- args 会覆盖镜像中的 command
  https://kubernetes.io/docs/tasks/inject-data-application/define-command-argument-container/

nodeSelector <map [string]string>

节点选择器， 限定pod运行在哪些节点上。

使用标签选择器

nodeName<map [string]string>

直接选择节点

annotations

注解，仅用于提供”元数据“并不提供 资源兑现选择。没有大小限制。

restartPolicy

Always, OnFailure, Never Default to Always

hostNetwork

Host networking requested for this pod. Use the host's network namespace.If this option is set, the ports that will be used must be specified. Default to false.

pod直接使用主机的网络名称空间。有用但不常用，默认false。

pod的生命周期

状态：

创建Pod：

apiServer etcd scheduler controller kubelet

容器重启策略：

restartPolicy

健康监测

健康监测主要针对容器，所以在 pod.spec.containers 层级下

监测类型

三种探针类型

ExecAction (exec)、TCPSocketAction (tcpSocket)、HTTPGetAction (httpGet)

健康监测主要参数

lifecycle

容器启动后 或者 停止前钩子。

注意：lifecycle的postStart执行在容器command 之后。

FIELDS:

env环境变量获取

env不仅可以传递key value 的数据，还可以从其他地方传值传递。

pods.spec.containers.env.valueFrom

pod 案例

apiVersion: v1
kind: Pod
metadata:
  name: nginx
spec:
  containers:
  - name: nginx
    image: nginx:1.7.9
    ports:
    - containerPort: 80
    readinessProbe:
      httpGet:
        port: 80
      initialDelaySeconds: 2
      periodSeconds: 3
    livenessProbe:
      httpGet:
        port: 80
      initialDelaySeconds: 2
      periodSeconds: 3

  - name: busybox
    image: busybox:latest
    imagePullPolicy: IfNotPresent
    command: [ping, www.baidu.com]
  nodeSelector:
    kubernetes.io/hostname: 192.168.0.165

Pod控制器

  1. 用户应该直接操作Deployment。
  2. 最好不要将有状态的服务部署在k8s上

deployment

更新策略

deployment.spec.strategy

deployment.spec.revisionHistoryLimit

rc历史保存数量

案例：

apiVersion: extensions/v1beta1
kind: Deployment
metadata:
  annotations:
    author: huruizhi
    department: opreation
    usage: Java programs k8s template
  labels:
    module_name: pyfinance2v2-register-pro
    env: pro
    kind: deploy
  name: pyfinance2v2-register-pro
  namespace: default
spec:
  replicas: 4
  strategy:
    type: RollingUpdate
    rollingUpdate:    
      maxSurge: 2
      maxUnavailable: 2
  selector:
    matchLabels:
      module_name: pyfinance2v2-register-pro
      env: pro
      kind: pod
  template:
    metadata:
      creationTimestamp: null
      labels:
        module_name: pyfinance2v2-register-pro
        env: pro
        kind: pod
    spec:
      containers:
      - name: pyfinance2v2-register-pro
        image: harbor.pycf.com/pyfinance2v2/register:pro
        imagePullPolicy: Always
        ports:
        - containerPort: 5000 
        command: ['java','-jar','-Xms128m','-Xmx256m','/java8/app.jar','--server.port=5000']
        resources:
          limits:
            memory: 512Mi
          requests:
            memory: 128Mi
        env:
        - name: TZ
          value: Asia/Shanghai
        livenessProbe:
          tcpSocket:
            port: 5000
          initialDelaySeconds: 40
          periodSeconds: 3
        readinessProbe:
          tcpSocket:
            port: 5000
          initialDelaySeconds: 40
          periodSeconds: 3
          
      imagePullSecrets:
      - name: harborkey1
      restartPolicy: Always

DaemSet

在每个节点上部署一个pod

支持滚动更新，支持两种更新模式。可以使用kubectl explain daemonset.spec.updateStrategy 查看。

手动更新 kubectl set image daemonset abc *=nginx:1.9.1

案例：

apiVersion: extensions/v1beta1
kind: DaemonSet
metadata:
  name: filefeat-ds
  namespace: default
  labels:
        app: filebeat
spec:
  selector:
    matchLabels:
      app: filebeat
      release: stable
  template:
    metadata:
      labels:
        app: filebeat
        release: stable
    spec:
      containers:
      - name: filefeat
        image: ikubenetes/filebeat:5.6.5-alpine
        env:
        - name: REDIS_HOST
          value: redis.default.svc.cluster.local
        - name: REDIS_LOG_LEVEL
          value: info

Service

Service的名称解析依赖于dns 附件，网络依赖于第三方网络方案。

Service网络是一个虚拟网络，由kube-proxy维护。

工作模式：

ipvs没有被激活的情况下自动使用iptables

iptables 查看：

iptables -L -n -t nat

svc.spec的重要字段

域名后缀

默认为svc_name.namespace_name.svc.cluster.local.

案例：

apiVersion: v1
kind: Service
metadata:
  annotations:
    kompose.cmd: kompose convert -f docker-compose-pro.yml
    kompose.version: 1.7.0 (HEAD)
  creationTimestamp: null
  labels:
    io.kompose.service: pyfinance2v2-amc-pro
  name: pyfinance2v2-amc-pro
  namespace: pyfinance2v2-pro
spec:
  type: NodePort
  ports:
  - name: "7562"
    port: 7562
    targetPort: 5000
    nodePort: 7562
  selector:
    io.kompose.service: pyfinance2v2-amc-pro
status:
  loadBalancer: {}

Ingress Controller

外部路由引入，7层负载均衡，可以进行https 卸载。

案例：

apiVersion: extensions/v1beta1
kind: Ingress
metadata:
  name: my-nginx-ingress
  namespace: default
spec:
  rules:
  - host: my-nginx.com
    http:
      paths:
      - path: /main
        backend:
          serviceName: my-nginx
          servicePort: 80
      - path: /busybox
        backend:
          serviceName: busybox-demo
          servicePort: 80

path: Path is an extended POSIX regex as defined by IEEE Std 1003.1, (i.e this follows the egrep/unix syntax, not the perl syntax) matched against the path of an incoming request. Currently it can contain characters disallowed from the conventional "path" part of a URL as defined by RFC 3986. Paths must begin with a '/'. If unspecified, the path defaults to a catch all sending traffic to the backend.

例如 path 设置为 /main 则可以访问 /main /main1 等。不能访问 / 、/aaa 等其他路径下资源

存储卷管理

支持的存储卷类型

kubectl explain pod.spec.volumes
kubectl explain persistentVolume.spec

定义一个简单的emptyDir, 包涵两个containers。两个容器公用存储卷。

apiVersion: v1
kind: Pod
metadata:
  name: busybox-demo
  labels:
    app: busybox
    role: volume_test
spec:
  containers:
  - name: httpd
    image: nginx:latest
    imagePullPolicy: IfNotPresent
    volumeMounts:
    - mountPath: /usr/share/nginx/html/
      name: tmp-volume
  - name: busybox
    image: busybox:latest
    imagePullPolicy: IfNotPresent
    command: ['/bin/sh','-c','while true;do echo $(date) > /data/index.html;sleep 3;done']
    volumeMounts:
    - mountPath: /data/
      name: tmp-volume
  volumes:
  - name: tmp-volume
    emptyDir:
      sizeLimit: 200M

PV 与 PVC 资源

pv与pvc

PV对象 及 主要参数

PV对象不属于名称空间

pv.Capacity

通过capacity给PV设置特定的大小。

pv.accessModes

k8s不会真正检查存储的访问模式或根据访问模式做访问限制，只是对真实存储的描述，最终的控制权在真实的存储端。目前支持三种访问模式：

* ReadWriteOnce – PV以 read-write 挂载到一个节点

* ReadOnlyMany – PV以read-only方式挂载到多个节点

* ReadWriteMany – PV以read-write方式挂载到多个节点

pv.spec.persistentVolumeReclaimPolicy

当前支持的回收策略:

* Retain – 允许用户手动回收

* Recycle – 删除PV上的数据 (“rm -rf /thevolume/*”)

* Delete – 删除PV

PVC对象 与重要参数

PVC 与PV对象 关联

pvc.spec.accessModes

同 pv对象

pvc.spec.resources

定义存储大小的需要

案例 Glusterfs：

apiVersion: v1
kind: Endpoints
metadata:
  name: gfs-endpoint
  labels:
    storage: gfs
subsets:
- addresses:
  - ip: 192.168.0.165
  ports:
  - port: 49158
    protocol: TCP
- addresses:
  - ip: 192.168.0.162
  - ip: 192.168.0.166
  ports:
  - port: 49157
    protocol: TCP
--- 
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: gfs-pvc
spec:
  accessModes: 
  - ReadWriteMany
  volumeName: gfs-pv
  resources:
    requests:
      storage: 20Gi
---    
apiVersion: v1
kind: PersistentVolume
metadata:
  name: gfs-pv
  labels:
    role: gfs-pv
spec:
  accessModes: 
  - ReadWriteMany
  glusterfs:  
    endpoints: gfs-endpoint
    path: gluster-test
  capacity:
    storage: 20Gi
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: gfs-pvc
spec:
  accessModes: 
  - ReadWriteMany
  volumeName: gfs-pv
  resources:
    requests:
      storage: 20Gi
---
apiVersion: v1
kind: Pod
metadata:
  name: busybox-demo
  labels:
    app: busybox
    role: volume_test
spec:
  containers:
  - name: httpd
    image: nginx:latest
    imagePullPolicy: IfNotPresent
    volumeMounts:
    - mountPath: /usr/share/nginx/html/busybox
      name: gfs-volume
  - name: busybox
    image: busybox:latest
    imagePullPolicy: IfNotPresent
    command: ['/bin/sh','-c','while true;do echo $(date) >> /data/index.html;sleep 3;done']
    volumeMounts:
    - mountPath: /data/
      name: gfs-volume
  volumes:
  - name: gfs-volume
    persistentVolumeClaim:
      claimName: gfs-pvc

StorageClass 动态生成pv

容器配置管理 secret 与 configmap

可以使用环境变量以及 挂载的方式配置到pod当中。

注意：环境变量的方式只能在容器启动的时候注入，更新configmap 不会更新容器中环境变量的值。使用挂载的方式可以实时更新。

创建configMap 有多种方式

  # Create a new configmap named my-config based on folder bar
  kubectl create configmap my-config --from-file=path/to/bar
  
  # Create a new configmap named my-config with specified keys instead of file basenames on disk
  kubectl create configmap my-config --from-file=key1=/path/to/bar/file1.txt --from-file=key2=/path/to/bar/file2.txt
  
  # Create a new configmap named my-config with key1=config1 and key2=config2
  kubectl create configmap my-config --from-literal=key1=config1 --from-literal=key2=config2
  
  # Create a new configmap named my-config from the key=value pairs in the file
  kubectl create configmap my-config --from-file=path/to/bar
  
  # Create a new configmap named my-config from an env file
  kubectl create configmap my-config --from-env-file=path/to/bar.env
apiVersion: v1
kind: ConfigMap
metadata:
  name: test-cfg
  namespace: default
data:
  cache_host: memcached-gcxt
  cache_port: "11211"
  cache_prefix: gcxt
  my.cnf: |
    [mysqld]
    log-bin = mysql-bin
  app.properties: |
    property.1 = value-1
    property.2 = value-2
    property.3 = value-3

使用命令行创建更灵活。

可以使用inotify监控配置文件实现重载

例如:

#!/bin/sh
oldcksum=`cksum /etc/nginx/conf.d/default.conf`

inotifywait -e modify,move,create,delete -mr --timefmt '%d/%m/%y %H:%M' --format '%T' \
/etc/nginx/conf.d/ | while read date time; do

    newcksum=`cksum /etc/nginx/conf.d/default.conf`
    if [ "$newcksum" != "$oldcksum" ]; then
        echo "At ${time} on ${date}, config file update detected."
        oldcksum=$newcksum
        nginx -s reload
    fi

done

关于configmap的详细总结： https://www.cnblogs.com/breezey/p/6582082.html

StatefuleSet

特点：

  1. 稳定且唯一的网络标识符；
  2. 稳定且持久的存储；
  3. 有序、平滑的部署和扩展；
  4. 有序、平滑的删除和终止；
  5. 有序的滚动更新；

三个主要组件：headless service 、 StatefulSet、 volumeClaimTemplate

名称解析：

pod_name,service_name.ns_name.svc.cluster.local

更新策略

sts.spec.updateStrategy.rollingUpdate

PV定义

apiVersion: v1
kind: Endpoints
metadata:
  name: gfs-endpoint
  labels:
    storage: gfs
subsets:
- addresses:
  - ip: 192.168.0.165
  ports:
  - port: 49158
    protocol: TCP
- addresses:
  - ip: 192.168.0.162
  - ip: 192.168.0.166
  ports:
  - port: 49157
    protocol: TCP

---

apiVersion: v1
kind: PersistentVolume
metadata:
  name: gfs-pv-01
  labels:
    role: gfs-pv-01
spec:
  accessModes: 
  - ReadWriteMany
  - ReadWriteOnce
  glusterfs:  
    endpoints: gfs-endpoint
    path: pv-01
  capacity:
    storage: 5Gi
---
apiVersion: v1
kind: PersistentVolume
metadata:
  name: gfs-pv-02
  labels:
    role: gfs-pv-02
spec:
  accessModes:
  - ReadWriteMany
  - ReadWriteOnce
  glusterfs:
    endpoints: gfs-endpoint
    path: pv-02
  capacity:
    storage: 5Gi
---
apiVersion: v1
kind: PersistentVolume
metadata:
  name: gfs-pv-03
  labels:
    role: gfs-pv-03
spec:
  accessModes:
  - ReadWriteMany
  - ReadWriteOnce
  glusterfs:
    endpoints: gfs-endpoint
    path: pv-03
  capacity:
    storage: 5Gi
--- 
apiVersion: v1
kind: PersistentVolume
metadata:
  name: gfs-pv-04
  labels:
    role: gfs-pv-04
spec:
  accessModes:
  - ReadWriteMany
  - ReadWriteOnce
  glusterfs:
    endpoints: gfs-endpoint
    path: pv-04
  capacity:
    storage: 5Gi
---
apiVersion: v1
kind: PersistentVolume
metadata:
  name: gfs-pv-05
  labels:
    role: gfs-pv-05
spec:
  accessModes:
  - ReadWriteMany
  - ReadWriteOnce
  glusterfs:
    endpoints: gfs-endpoint
    path: pv-05
  capacity:
    storage: 5Gi

StatefulSet定义

apiVersion: v1
kind: Service
metadata:
  name: myapp-svc
  labels:
    roles: myapp-svc-test
spec:
  clusterIP: None
  ports:
  - targetPort: 80
    port: 80
  selector:
    roles: myapp-pod
---
apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: myapp-sts
  labels:
    roles: myapp-sts-test
spec:
  replicas: 3
  serviceName: myapp-svc
  selector: 
    matchLabels:
      roles: myapp-pod
  template:
    metadata:
      labels:
        roles: myapp-pod
    spec:
      containers:
       - name: httpd
         image: nginx:latest
         imagePullPolicy: IfNotPresent
         volumeMounts:
         - mountPath: /usr/share/nginx/html/busybox
           name: gfs-volume
  volumeClaimTemplates:
  - metadata:
      name: gfs-volume
    spec:
      accessModes: [ "ReadWriteOnce" ]
      resources:
        requests:
          storage: 5Gi
  updateStrategy:
    rollingUpdate: 
      partition: 2

标签：02,kind,name,gfs,Kubenetes,pod,spec,资源,metadata
来源： https://www.cnblogs.com/pythonPath/p/11267275.html