redhat DO280

转自：https://www.cnblogs.com/smlile-you-me/p/16668899.html

DO280

目录

 

前奏

  # 安装命令补全
  yum -y install bash-completion
  source /etc/profile.d/bash_completion.sh
   
  # 搭建集群(不考)
  [root@foundation0 ~]# echo y | rht-vmctl reset all
  [root@foundation0 ~]# ssh student@workstation
  [student@workstation ~]$ lab install-prepare setup
  [student@workstation ~]$ cd /home/student/do280-ansible
  [student@workstation do280-ansible]$ ./install.sh
   
  # 部署成功执行以下脚本
  [student@workstation do280-ansible]$ lab install-metrics setup
  [root@foundation0 ~]# scp -r files/ root@master:~
   
  # 考试注意事项
  相关节点密码均考试说明中给出;
  考试环境中OpenShift已部署完成,建议进行检查
  1、每个节点openshift已经安装完成。
  2、考试时候大部分操作在master上完成,master节点存储空间有限,push image在node上完成
  3、物理机作为集群外节点直接访问ocp中route,练习环境中workstation可以直接访问ocp中route(查文档也是同样方法)
  4、考试时没有workstation虚拟机
  5、所有练习基本全在master节点中完成
  6、不用管理防火墙
  7、练习环境中部署后的网站必须到workstation节点中的浏览器上验证（考试时直接在宿机上验证）
SHELL 复制 全屏

1、Create OpenShift users

  # 原题
  Create additional OpenShift users with the following characteristics:
  创建具有以下特征的其他 OpenShift 用户：
  1、The regular user joe with password redhat;
  2、The regular user lene with password redhat;
  3、You must use the existing authentication file at /etc/origin/master/htpasswd while preserving its original content;
  必须使用/etc/origin/master/htpasswd已有的认证文件，同时保留其原始内容；
  4、Both users must be able to authenticate to the OpenShift instance via CLI and on the web console at https://master.lab.example.com:443;
  两个用户都必须能够通过CLI和Web控制台https://master.lab.example.com:443对OpenShift实例进行身份验证；
  5、Regular users must NOT be able to create projects themselves
  普通用户不能自己创建项目
   
  [root@foundation0 ~]# ssh root@master
  [root@master ~]# rpm -qa httpd-tools
  [root@master ~]# yum -y install httpd-tools
  [root@master ~]# oc get nodes
  [root@master ~]# oc get pods
  [root@master ~]# oc whoami
  system:admin
  ===============================================================
  提示:
  master节点的系统root用户默认是集群管理员
  给普通用户授权集群管理权限
  [root@master ~]# oc adm policy add-cluster-role-to-user [ system:admin | cluster-admin ] admin
  ===============================================================
  # 按要求确定密码文件存放位置
  [root@master ~]# grep -A3 prov /etc/origin/master/master-config.yaml
  # 确认用户及密码配置文件
  [root@master ~]# cat /etc/origin/master/htpasswd
  # 创建用户
  [root@master ~]# htpasswd -b /etc/origin/master/htpasswd joe redhat
  [root@master ~]# htpasswd -b /etc/origin/master/htpasswd lene redhat
   
  # 为组去除集群的角色(所有用户不允许创建项目)
  注：在帮助文档的6.2.2 Disabling Self-provisioning节中有示例
  [root@master ~]# oc adm policy remove-cluster-role-from-group \
  self-provisioner \
  system:authenticated \
  system:authenticated:oauth
  # 验证
  [root@master ~]# oc login -u joe -p redhat
  [root@master ~]# oc login -u lene -p redhat
  注：都是提示能成功登录但不能创建项目则为正确

2、Configure persistent storage for the local registry

  # 原题
  Configure nfs persistent storage on services.lab.example.com,that does thefollowing:
  1、Create and share /OCP_mysql;
  2、Create and share /OCP_wordpress;
  3、Create and share /OCP_registry;
  4、Create and share /OCP_metrics;
  5、Associate the share named /OCP_registry to the builtin registry running within your OpenShift Enterprise instance so that it will be used for permanent storage
  将共享/OCP_registry关联到 OpenShift Enterprise 实例的镜像仓库
  6、Use train-registry-volume for the volume name and train-registry-claim for the claim name
  使用train-registry-volume 作为 volume name, train-registry-claim 作为 claim name
  7、You can find sample YAML files on http://classroom.example.com/materials/exam/storage/
  可在http://classroom.example.com/materials/exam/storage/上找到示例YAML文件
  (Note: This task needs to be solved before any applications are created)
  8、所有共享均可以被172.25.250.0/255.255.255.0网段用户共享 (考试时查看hosts文件来定义具体的网段)
   
   
  # 解题
  先创建nfs，再创建pv，pv就和nfs关联上了, 再创建pvc,pvc就会自动和pv进行绑定。然后对pod进行pvc关联，这样pod的内容就写入到nfs的路径了
  nfs--pv
  |
  pvc <-- pod
   
  [root@foundation0 ~]# ssh root@services
  [root@services ~]# mkdir /OCP_{mysql,wordpress,registry,metrics}
  [root@services ~]# chown nfsnobody.nfsnobody /OCP_*
  [root@services ~]# chmod 700 /OCP_*
   
   
  [root@services ~]# man exports 查看设置示例
  [root@services ~]# vim /etc/exports
  /OCP_metrics 172.25.250.0/24(rw,async,all_squash)
  /OCP_mysql 172.25.250.0/24(rw,async,all_squash)
  /OCP_registry 172.25.250.0/24(rw,async,all_squash)
  /OCP_wordpress 172.25.250.0/24(rw,async,all_squash)
   
  [root@services ~]# exportfs -r
  [root@services ~]# showmount -e | grep OCP
   
   
  # 将共享关联到实例镜像仓库(需要切换到master节点)
  [root@master ~]# oc login -u system:admin
  [root@master ~]# oc project default
  [root@master ~]# cd files/2/
  注：考试时会提供相应的文件，只需将文件wget下来即可
  # 创建pv和pvc
  [root@master 2]# vim sample-pv.yml
  apiVersion: v1
  kind: PersistentVolume
  metadata:
  name: train-registry-volume # volume的名字
  spec:
  capacity:
  storage: 5Gi
  accessModes:
  - ReadWriteMany # 访问的模式
  nfs:
  path: /OCP_registry # 共享的目录
  server: services.lab.example.com # 共享的服务器是哪里
  persistentVolumeReclaimPolicy: Recycle
   
  [root@master 2]# vim sample-pvc.yml
  apiVersion: v1
  kind: PersistentVolumeClaim
  metadata:
  name: train-registry-claim # pvc的名字
  spec:
  accessModes:
  - ReadWriteMany # 访问模式
  resources:
  requests:
  storage: 5Gi # 请求资源的大小
   
  # 创建pv只能是管理员用户
  [root@master 2]# oc whoami
  system:admin
  [root@master 2]# oc create -f sample-pv.yml -n default
  # 因为这个pvc也隶属于集群，所以也由集群管理员来管理
  [root@master 2]# oc create -f sample-pvc.yml -n default
  [root@master 2]# oc get pods
  [root@master 2]# oc get pv
  [root@master 2]# oc get pvc
  注：关联成功并不代表能写到共享存储中，因为nfs如果配错了会导致写不进去和容器起不来等情况
   
  # 将pod和pvc关联
  [root@master 2]# oc get dc
  NAME REVISION DESIRED CURRENT TRIGGERED BY
  docker-registry 1 2 2 config
   
  [root@master 2]# oc describe dc/docker-registry | grep -A8 Vol
  Volumes:
  registry-storage: # 对这个Volumes进行设置，下一条命令中的--name所用的就是这个名字
  Type: PersistentVolumeClaim (a reference to a PersistentVolumeClaim in the same namespace)
  ClaimName: registry-claim
  ReadOnly: false
   
  [root@master 2]# oc set volume dc/docker-registry --add --overwrite --name=registry-storage -t pvc --claim-name=train-registry-claim
  # 再来验证
  [root@master 2]# oc describe dc/docker-registry | grep -A8 Vol
  [root@master 2]# oc get pods

3、Create OpenShift Enterprise projects

创建 OpenShift 企业项目

  # 原题
  On your OpenShift Enterprise instance create the following projects:
  1、raleigh
  2、lobster
  3、farm
  4、ditto
  5、samples
   
  Additionally, configure the projects as follows:
  1、For all of the projects, set the description to 'This is a DO280 project';
  2、Make joe the admin of project raleigh and ditto;
  3、The user lene must be able to view the project raleigh but not administer or delete it;
  4、Make lene the admin of projects farm,lobster and samples。
   
   
  # 解题
  [root@master 2]# oc whoami
  system:admin
  [root@master 2]# for i in raleigh lobster farm ditto samples;do oc new-project $i --description="This is a DO280 project";done
  [root@master 2]# for i in raleigh lobster farm ditto samples;do oc get project $i|grep -v ^NAME;done
  ## joe用户是以下2个项目的管理员
  [root@master 2]# oc adm policy add-role-to-user admin joe -n raleigh
  [root@master 2]# oc adm policy add-role-to-user admin joe -n ditto
  ## lene用户只能查看raleigh项目
  [root@master 2]# oc adm policy add-role-to-user view lene -n raleigh
  ## lene用户是以下项目的管理员
  [root@master 2]# oc adm policy add-role-to-user admin lene -n farm
  [root@master 2]# oc adm policy add-role-to-user admin lene -n lobster
  [root@master 2]# oc adm policy add-role-to-user admin lene -n samples
   
  删除格式：oc adm policy remove-role-from-user <角色> <用户名> -n <项目>
  删除示例：oc adm policy remove-role-from-user admin lene -n farm
   
  [root@master 2]# oc get rolebinding -n {farm | lobster...}

4、Create an application from a Git repository

从 Git 存储库创建应用程序

  Use the S2I functionality of your OpenShift instance to build an application in the raleigh project
  使用 OpenShift 实例的 S2I 功能在 raleigh 项目中构建应用程序
  Use the Git repository at http://services.lab.example.com/php-helloworld for theapplication source
  使用位于 http://services.lab.example.com/php-helloworld 的 Git 存储库作为应用程序源
  1、Use the Docker image labeled openshift/php:5.6;
  使用标记为 openshift/php:5.6 的 Docker 镜像；
  2、Once deployed, the application must be reachable(and browsable)at the following address: http://helloworld.raleigh.apps.lab.example.com;
  部署后，应用程序必须可以通过以下地址访问http://...
  3、Update the original repository so that the index.php file contains the text "This is a DO280 exam" instead of the word 'Hello, World! php version is';
  更新原始repository中的index.php文件内容,使其内容成为"This is a DO280 test",以替换原来的文字
  4、Trigger a rebuild so that when browsing http://helloworld.raleigh.apps.lab.example.com it will display the new text
  触发重建，以便在浏览 http://helloworld.raleigh.apps.lab.example.com 时显示新文本
   
   
  # 查看本地仓库是否有该镜像
  [root@services ~]# docker-registry-cli services.lab.example.com [list all] | <[search IMAGE_NAME]> ssl
  # 查看公共is（openshift项目是公共的）
  [root@master ~]# oc get is -n openshift | grep php
   
  # 开始创建应用
  因为joe用户是raleigh项目的管理员，所以要切换用户,需要在master节点上操作
  [root@master 2]# oc login -u joe
  [root@master 2]# oc project raleigh
  [root@master 2]# oc new-app openshift/php:5.6~http://services.lab.example.com/php-helloworld --name=hello # php:5.6是is地址
  [root@master 2]# oc get build
  [root@master 2]# oc logs -f bc/hello
  注：如果这里push出错,说明第二题registry永久卷配置错误,重置环境重做
   
  # push后对raleigh项目的is验证
  [root@master 2]# oc get is -n raleigh | grep hello
  hello docker-registry.default.svc:5000/raleigh/hello latest 20 minutes ago
   
  # 对push后的存储结果进行验证
  [root@services ~]# ll /OCP_registry/docker/registry/v2/repositories/raleigh
   
  # 创建路由
  [root@master 2]# oc get svc
  NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
  hello ClusterIP 172.30.196.16 <none> 8080/TCP,8443/TCP 26m
   
  [root@master 2]# oc expose svc hello --hostname=helloworld.raleigh.apps.lab.example.com
  注：helloworld.raleigh.apps.lab.example.com是题目中给定的
  [root@master 2]# oc get route
  NAME HOST/PORT PATH SERVICES PORT TERMINATION WILDCARD
  hello helloworld.raleigh.apps.lab.example.com hello 8080-tcp None
  [root@master 2]# curl helloworld.raleigh.apps.lab.example.com
  Hello, World! php version is 5.6.25
   
  # 更新源代码
  [root@master 2]# cd
  [root@master ~]# git clone http://services.lab.example.com/php-helloworld
  [root@master ~]# cd php-helloworld/
  [root@master php-helloworld]# vim index.php
  <?php
  print "This is a DO280 test\n";
  ?>
   
  [root@master php-helloworld]# git add .
  [root@master php-helloworld]# git commit -m "xxx"
  [root@master php-helloworld]# git push
   
  # 手动触发rebuild
  [root@master php-helloworld]# oc start-build hello
  [root@master php-helloworld]# oc logs -f bc/hello
  [root@master php-helloworld]# curl helloworld.raleigh.apps.lab.example.com
  This is a DO280 test
  [root@master php-helloworld]# cd

5、Create an application using Docker images and definition files

使用 Docker 映像和定义文件创建应用程序

  Using the example files from the wordpress directory under http://classroom.example.com/materials/exam/wordpress
  使用 http://classroom.example.com/materials/exam/wordpress 下 wordpress 目录中的示例文件
  1、create a WordPress application in the farm project;
  在farm项目中创建一个WordPress应用；
  2、For permanent storage use the the NFS shares /OCP_wordpress and /OCP_mysql from services.lab.example.com
  对于永久存储，使用来自 services.lab.example.com 的 NFS 共享 /OCP_wordpress 和 /OCP_mysql
  Use the files from http://classroom.example.com/materials/exam/wordpress for the volumes.
  使用来自 http://classroom.example.com/materials/exam/wordpress 的文件作为卷
  3、For the WordPress pod,use the Docker image from http://classroom.example.com/materials/exam/wordpress/wordpress.tar;
  对于WordPress pod，使用来自http://classroom.example.com/materials/exam/wordpress/wordpress.tar的Docker镜像；
  (Note: It is normal if the WordPress pod initially restarts a couple of times due to permission issues)
  (注意：如果 WordPress pod 最初重启几次是正常的，因为权限问题）
  4、For the MySQL pod use the Docker image openshift3/mysql-55-rhel7;
  对于MySQL pod，使用Docker镜像openshift3/mysql-55-rhel7；
  5、Once deployed, the application must be reachable at the following address: http://shining.farm.apps.lab.example.com;
  部署后，应用程序必须可访问以下地址：http://shining.farm.apps.lab.example.com；
  6、Finally, complete the WordPress installation by setting lene as the admin user with password redhat and lene@master.lab.example.com for the email address;
  最后完成WordPress的安装，设置lene为admin用户，密码为redhat，邮箱为lene@master.lab.example.com；
  7、Set the blog name to do280 blog;
  设置博客名称为do280 blog；
  8、Create your first post with title "My first post"。The text in the post does not matter
  创建您的第一篇标题为"My first post"的文章，文章中的文字无关紧要
   
   
  farm项目的管理员是lene用户，所以要用lene用户并切换到farm中
  [root@master ~]# oc login -u lene -p redhat
  [root@master ~]# oc project farm
  [root@master ~]# cd files/5
   
  先创建2个pv--> 2个pvc--->加载WP镜像-->创建MySQL的pod-->MySQL service-->创建WP的pod-->WP的service-->指向WP的service-->|
  创建时需要是一个pv一个pvc，一个pv一个pvc的创建，如果统一把pv先创建完再统一创建pvc可能会匹配不上(混乱)
   
  [root@master 5]# mv pv-1.yaml pv-wp.yaml
  [root@master 5]# mv pv-2.yaml pv-mysql.yaml
  ===================================================
  [root@master 5]# vim pv-mysql.yaml
  apiVersion: v1
  kind: PersistentVolume
  metadata:
  name: pv-mysql # 名字可自定义
  spec:
  capacity:
  storage: 3Gi # 大小3G就够
  accessModes:
  - ReadWriteMany # 访问模式写成ReadWriteMany
  persistentVolumeReclaimPolicy: Recycle
  nfs:
  server: services.lab.example.com
  path: /OCP_mysql
   
   
  [root@master 5]# vim pvc-mysql.yaml
  kind: PersistentVolumeClaim
  apiVersion: v1
  metadata:
  name: claim-mysql
  spec:
  accessModes:
  - ReadWriteMany # 访问模式写成ReadWriteMany
  resources:
  requests:
  storage: 3Gi
  ===================================================
   
  [root@master 5]# vim pv-wp.yaml
  apiVersion: v1
  kind: PersistentVolume
  metadata:
  name: pv-wordpress # 名字可自定义
  spec:
  capacity:
  storage: 1Gi
  accessModes:
  - ReadWriteMany # 访问模式写成ReadWriteMany
  persistentVolumeReclaimPolicy: Recycle
  nfs:
  server: services.lab.example.com
  path: /OCP_wordpress
   
  [root@master 5]# vim pvc-wp.yaml
  kind: PersistentVolumeClaim
  apiVersion: v1
  metadata:
  name: claim-wp
  spec:
  accessModes:
  - ReadWriteMany # 访问模式写成ReadWriteMany
  resources:
  requests:
  storage: 1Gi
   
   
  # 创建2个pvc
  [root@master 5]# oc create -f pvc-mysql.yaml
  [root@master 5]# oc create -f pvc-wp.yaml
   
  # 创建2个pv(切到集群管理员用户)
  [root@master 5]# oc login -u system:admin
  [root@master 5]# oc project farm
  [root@master 5]# oc create -f pv-mysql.yaml
  [root@master 5]# oc create -f pv-wp.yaml
  # 检查
  [root@master 5]# oc get pvc
   
  # 把镜像导入进来并推送到内部镜像仓库中
  [root@master 5]# docker load -i wordpress.tar
  [root@master 5]# docker tag docker.io/wordpress:latest registry.lab.example.com/wordpress:latest
  [root@master 5]# docker push registry.lab.example.com/wordpress:latest # 第一次可查看hosts文件中的registry的地址
  [root@master 5]# docker rmi -f 4ad4
   
  # 搭建MySQL数据库的pod
  [root@master 5]# vim pod-mysql.yaml
  apiVersion: v1
  kind: Pod
  metadata:
  name: mysql
  labels:
  name: mysql # 这个labels下面会用到
  spec:
  containers:
  - resources:
  limits :
  cpu: 0.5
  image: openshift3/mysql-55-rhel7 # 修改成提供的镜像
  ....
  ....
  volumes:
  - name: mysql-persistent-storage
  persistentVolumeClaim:
  claimName: claim-mysql # 已对,如不对就按oc get pvc的名字修改(MySQL的pvc的名字)
   
  [root@master 5]# oc create -f pod-mysql.yaml
  [root@master 5]# oc get pods
   
   
  # 创建和MySQL对应的service
  [root@master 5]# vim service-mysql.yaml
  ....
  ....
  selector:
  name: mysql # 需要和上面的labels匹配(pod名)
   
  [root@master 5]# oc create -f service-mysql.yaml
  [root@master 5]# oc get svc
   
  # 创建WordPress的pod
  [root@master 5]# vim pod-wordpress.yaml
  apiVersion: v1
  kind: Pod
  metadata:
  name: wordpress
  labels:
  name: wordpress
  spec:
  containers:
  - image: wordpress # 默认是到内部镜像仓库中找镜像，该镜像是用docker load导入进来后push到内部仓库的所以这里不用修改
  ....
  ....
  - name: WORDPRESS_DB_HOST
  # this is the name of the mysql service fronting the mysql pod in the same namespace
  # expands to mysql.<namespace>.svc.cluster.local - where <namespace> is the current namespace
  value: mysql.farm.svc.cluster.local # 上面有说明
   
  # 创建scc
  [root@master 5]# oc login -u system:admin # 很重要的一步
  [root@master 5]# oc adm policy add-scc-to-user anyuid -z default -n farm # 很重要的一步
  释义：
  给farm项目中的default默认的用户，这个默认的用户不管是什么名字就是执行程序的serviceaccount
  换言之，有时要自己起一个serviceaccount名字，例如执行程序的时候不是root用户，我就给它起个名字，
  例如叫xiaoming，我可以用oc adm policy add-scc-to-user anyuid -z xiaoming -n farm
  但是现在我不想知道这个程序中他到底起了什么名字，也不想去查了，你不管用什么名字，我就用default来代替来统称
   
   
  [root@master 5]# oc login -u lene
  [root@master 5]# oc create -f pod-wordpress.yaml
  [root@master 5]# oc get pods
  NAME READY STATUS RESTARTS AGE
  mysql 1/1 Running 0 19m
  wordpress 1/1 Running 2 3m
   
   
  # 给wp创建一个服务
  [root@master 5]# vim service-wp.yaml # 不用修改，默认即可
  [root@master 5]# oc create -f service-wp.yaml
  [root@master 5]# oc get svc
  NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
  mysql ClusterIP 172.30.139.14 <none> 3306/TCP 12m
  wpfrontend LoadBalancer 172.30.140.86 172.29.227.62,172.29.227.62 80:32247/TCP 2s
   
  # 给wp的service创建路由
  [root@master 5]# oc expose svc wpfrontend --hostname=shining.farm.apps.lab.example.com
  注：wpfrontend是svc的名字，--hostname是题目中给定的
  [root@master 5]# oc get route
  NAME HOST/PORT PATH SERVICES PORT TERMINATION WILDCARD
  wpfrontend shining.farm.apps.lab.example.com wpfrontend 80 None
   
  去workstations图形化中操作....

6、Create an application with a secure edgeterminated route

创建具有 secure edge-terminate路由的应用程序

  Create an application greeter in the project samples, which uses the Docker image
  registry.lab.example.com/openshift/hello-openshift so that it is reachable at the following address only: https://greeter.samples.apps.lab.example.com
  (Note you can use the script http://classroom.example.com/materials/exam/cert/gencert.sh to generate the necessary certificate files)
  # 译文
  在samples项目中创建一个应用greeter,它使用registry.lab.example.com/openshift/hello-openshift镜像,应用可以通过: https://greeter.samples.apps.lab.example.com访问
  (请注意, 请使用/files/6/gencert.sh 创建必要的证书文件)
  lene用户是samples项目的管理员
  [root@master 5]# cd ../6/
  [root@master 6]# oc login -u lene
  [root@master 6]# oc project samples
  [root@master 6]# oc new-app --docker-image=registry.lab.example.com/openshift/hello-openshift --name=greeter
  [root@master 6]# oc get pods
  [root@master 6]# oc get svc
  ## 创建证书
  [root@master 6]# bash gencert.sh greeter
  ## 创建路由
  [root@master 6]# oc create route edge --help
  [root@master 6]# oc create route edge \
  --cert=greeter.crt \
  --key=greeter.key \
  --service=greeter \
  --hostname=greeter.samples.apps.lab.example.com
   
  [root@master 6]# oc get route
  NAME HOST/PORT PATH SERVICES PORT TERMINATION WILDCARD
  greeter greeter.samples.apps.lab.example.com greeter 8080-tcp edge None
   
  [root@master 6]# cd
   
  注：在workstation访问HOST/PORT列的地址，https://greeter.samples.apps.lab.example.com
  然后点击Advanced---Add Exception---Confirm Security Exception---看到最终的结果是Hello OpenShift!

7、Configure OpenShift quotas for a project

为openshift的一个项目配置quotas

  Configure quotas and limits for project lobster so that:
  为lobster项目配置quotas 和 limits:
  The ResourceQuota resource is named do280-quota
  ResourceQuota资源名为do280-quota
   
  1、The amount of memory consumed across all containers may not exceed 1Gi;
  所有容器内存消耗不能超过1Gi
  2、The total amount of CPU usage consumed across all containers may not exceed 2 Kubernetes compute units;
  所有容器消耗的CPU使用总量不得超过2个Kubernetes计算单元；
  3、The maximum number of replication controllers does not exceed 3 The maximum number of pods does not exceed 3;
  ReplicationControllers和Pod数量都不能超过3个；
  4、The maximum number of services does not exceed 6
  Services不能超过6个
   
  The LimitRange resource is named do280-limits
  LimitRange资源名: do280-limits
  1、The amount of memory consumed by a single pod is between 5Mi and 300Mi;
  单个pod内存用量在5Mi-300Mi
  2、The amount of memory consumed by a single container is between 5Mi and 300Mi with a default request value of 100Mi;
  单个容器消耗的内存量在5Mi-300Mi之间，默认请求值为100Mi；
  3、The amount of cpu consumed by a single pod is between 10m and 500m;
  单个pod消耗的cpu量在10m-500m之间；
  4、The amount of cpu consumed by a single container is between 10m and 500m with a default request value of 100m
  单个容器消耗的cpu量在10m-500m之间，默认请求值为100m
   
   
  注：配置quota必须要有集群管理员，项目管理员无权限做
  [root@master 6]# cd
  [root@master ~]# oc login -u system:admin
  [root@master ~]# oc project lobster
  # 为lobster项目创建quota
  [root@master ~]# vim do280-quota.yaml
  apiVersion: v1
  kind: ResourceQuota
  metadata:
  name: do280-quota # ResourceQuota资源名
  spec:
  hard:
  services: "6" # Services不能超过6个
  pods: "3" # Pods不能超过3个
  limits.cpu: "2" # CPU用量最多2Kubernetes units
  limits.memory: 1Gi # 所有容器内存消耗不能超过1Gi
  replicationcontrollers: "3" # RC最大不能超过3个
   
  [root@master ~]# oc create -f do280-quota.yaml
  [root@master ~]# oc describe quota -n lobster
   
   
  # 为lobster项目创建limits
  [root@master ~]# vim do280-limits.yaml
  kind: "LimitRange"
  metadata:
  name: "do280-limits"
  spec:
  limits:
  - type: "Pod"
  max:
  cpu: "500m"
  memory: "300Mi"
  min:
  cpu: "10m"
  memory: "5Mi"
  - type: "Container"
  max:
  cpu: "500m"
  memory: "300Mi"
  min:
  cpu: "10m"
  memory: "5Mi"
  defaultRequest:
  cpu: "100m"
  memory: "100Mi"
   
  [root@master ~]# oc create -f do280-limits.yaml
  [root@master ~]# oc describe limitrange

8、Create an application from a third party template

从第三方模板创建应用程序

  On master.lab.example.com using the template file in http://classroom.example.com/materials/exam/gogs as a basis, install an application in the ditto project according to the following requirements:
  在master.lab.example.com上使用http://classroom.example.com/materials/exam/gogs中的模板文件(gogs-template.yaml),在ditto项目中安装一个应用,基于如下要求:
  ⚫ All of the registry entries must point to your local registry at registry.lab.example.com。The version in the ImageStream line for the postgresql image must be changed from postgresql:9.2 to postgresql:9.5;
  所有registry都必须指向您在registry.lab.example.com的本地registry,ImageStream中的postgresql image版本必须从 postgresql:9.2更改为postgresql:9.5
  ⚫ for the Gogs pod,use the Docker image from http://classroom.example.com/materials/exam/gogs/gogs.tar and make sure it is tagged as registry.lab.example.com/openshiftdemos/gogs:0.9.97 and pushed to your local registry vm;
  gogs pod的镜像来自于http://classroom.example.com/materials/exam/gogs.tar的Docker镜像,并确保它的tag改为registry.lab.example.com/openshiftdemos/gogs:0.9.97,并推送到registry中
  ⚫ Make the template gogs available across all projects and for all users;
  使模板gogs可用于所有项目和所有用户(就是导入到openshift项目中才可以)
  ⚫ Deploy the application using the template, setting the parameter HOSTNAME to gogs.ditto.apps.lab.example.com;
  使用模板部署应用，设置参数HOSTNAME为gogs.ditto.apps.lab.example.com；
  ⚫ Create a user joe with password redhat and email address joe@master.lab.example.com on the application frontend (use the Register link on the top right of the page at http://gogs.ditto.apps.lab.example.com) and, as this user, create a Git repository named do280;
  在应用程序前端创建用户joe使用密码redhat和电子邮件地址joe@master.lab.example.com,(使用页面右上角的注册链接http://gogs.ditto.apps.lab.example.com)并以此用户身份创建一个名为do280的Git存储库；
  ⚫ If there isn't one already, create a file named README.md in the repository do280 and put the line "do280 is fun" in it and commit it;
  如果还没有，请在存储库do280中创建一个名为README.md的文件，并将"do280 is fun"行放入其中并提交；
  ⚫The repository must be visible and accessible
  存储库必须可见且可访问
   
   
  # 授权 SCC
  [root@master ~]# oc login -u system:admin
  [root@master ~]# oc adm policy add-scc-to-user anyuid -z default -n ditto
  # 导入镜像
  [root@master ~]# docker load -i ~/files/8/gogs.tar
  [root@master ~]# docker images
  [root@master ~]# docker tag docker.io/openshiftdemos/gogs:latest registry.lab.example.com/openshiftdemos/gogs:0.9.97
  [root@master ~]# docker push registry.lab.example.com/openshiftdemos/gogs:0.9.97
  [root@master ~]# docker rmi -f 3ca
   
  # 查看是否上传到了内部镜像仓库
  [root@services ~]# ll /var/lib/registry/docker/registry/v2/repositories/openshiftdemos
   
  # 在本地仓库验证
  [student@workstation ~]$ docker-registry-cli registry.lab.example.com search gogs ssl
  [student@workstation ~]$ docker-registry-cli registry.lab.example.com search postgresql ssl
   
  # 查看is中是否有，如有、在下面模板文件中可直接使用镜像名，如无、则需要写完整地址
  [root@master ~]# oc get is -n openshift| grep gogs
  [root@master ~]# oc get is -n openshift| grep pos
  postgresql docker-registry.default.svc:5000/openshift/postgresql 9.5,9.2,9.4 + 1 more... 2 weeks ago
   
   
  # 按题意修改模板(使用system:admin用户)
  [root@master ~]# oc project ditto
  [root@master ~]# vim files/8/gogs-template.yaml
  ....
  ....
  - kind: DeploymentConfig
  apiVersion: v1
  metadata:
  annotations:
  description: Defines how to deploy the database
  name: ${APPLICATION_NAME}-postgresql
  ....
  image: 'rhscl/postgresql-95-rhel7' # workstation上用docker-registry-cli搜索出来的
  ....
  triggers:
  - imageChangeParams:
  automatic: true
  containerNames:
  - postgresql
  from:
  kind: ImageStreamTag
  name: postgresql:9.5 # 考试中这里是9.2，需要改成9.5,练习中不用动
  ....
  ....
  - kind: DeploymentConfig
  apiVersion: v1
  metadata:
  labels:
  app: ${APPLICATION_NAME}
  ....
  spec:
  serviceAccountName: ${APPLICATION_NAME}
  containers:
  - image: "registry.lab.example.com/openshiftdemos/gogs:0.9.97" # 改镜像地址，因为is中无，所以写完整地址
  ....
  ....
  - kind: ImageStream
  apiVersion: v1
  metadata:
  labels:
  app: ${APPLICATION_NAME}
  name: ${APPLICATION_NAME}
  spec:
  tags:
  - name: "${GOGS_VERSION}"
  from:
  kind: DockerImage
  # 将自带的docker.io改为registry.lab.example.com
  name: registry.lab.example.com/openshiftdemos/gogs:${GOGS_VERSION}
  ....
  ....
  parameters:
  name: HOSTNAME
  required: true
  value: gogs.ditto.apps.lab.example.com # 就是待会创建好后的路由
   
   
  # 因为只有集群管理员才能往openshift项目中上传模板所以需要使用system:admin用户登录到集群(openshift项目中的所有模板所有用户都能使用)
  # 创建模板
  [root@master ~]# oc login -u system:admin
  [root@master ~]# oc create -f files/8/gogs-template.yaml -n openshift
  [root@master ~]# oc get template -n openshift | grep gogs
   
  [root@master ~]# oc login -u joe -p redhat
  [root@master ~]# oc project ditto
  [root@master ~]# oc process openshift//gogs | oc create -f - # 普通用户不能处理openshift中的模板,要加上//才能
  [root@master ~]# oc get pods
  [root@master ~]# oc get route
  NAME HOST/PORT PATH SERVICES PORT TERMINATION WILDCARD
  gogs gogs.ditto.apps.lab.example.com gogs <all> None
   
  在 workstation 上进行网页访问,或者 添加 hosts 映射在 foundation0 上进行访问。http://gogs.ditto.apps.lab.example.com/





  [root@master ~]# mkdir test && cd test
  [root@master test]# echo "do280 is fun" >> README.md
  [root@master test]# git init
  [root@master test]# git add README.md
  [root@master test]# git commit -m "do280 is fun"
  [root@master test]# git remote add origin http://gogs.ditto.apps.lab.example.com/joe/do280.git
  [root@master test]# git push -u origin master
  Username for 'http://gogs.ditto.apps.lab.example.com': 用户名是joe
  Password for 'http://joe@gogs.ditto.apps.lab.example.com': 密码是redhat

9、Scale an application

  Scale the application greeter in the project samples to a total of 5 replicas
   
   
  [root@master test]# oc login -u lene # 如要输入密码则是redhat
  [root@master test]# oc project samples
  [root@master test]# oc get dc
  NAME REVISION DESIRED CURRENT TRIGGERED BY
  greeter 1 1 1 config,image(greeter:latest)
   
  [root@master test]# oc scale --replicas=5 dc/greeter
  [root@master test]# oc get pods

10、Install OpenShift metrics

  On workstation.lab.example.com install the OpenShift Metrics component with the following requirements:
  在workstation.lab.example.com上安装具有以下要求的OpenShift Metrics组件：
  ⚫ Use the storage /OCP_metrics for cassandra storage。You can use the files on http://classroom.example.com/materials/exam/metrics for the pv sample;
  cassandra存储使用存储/OCP_metrics，pv示例可使用http://classroom.example.com/materials/exam/metrics上的文件；(可参考files/10中pv样例文件)
  ⚫ Use the file /home/student/DO280/labs/installmetrics/host for the inventory。Use the playbook /usr/share/ansible/openshiftansible/playbooks/openshiftmetrics/config.yml for the installation;
  使用/home/student/DO280/labs/install-metrics/目录中的inventory文件。
  playbook在/usr/share/ansible/openshiftansible/playbooks/openshiftmetrics/config.yml；
  ⚫ Use the following environment variables:
  使用以下环境变量：
  openshift_metrics_image_version=v3.9
  openshift_metrics_heapster_requests_memory=300M
  openshift_metrics_hawkular_requests_memory=750M
  openshift_metrics_cassandra_requests_memory=750M
  openshift_metrics_cassandra_storage_type=pv
  openshift_metrics_cassandra_pvc_size=5Gi
  openshift_metrics_cassandra_pvc_prefix=metrics
  openshift_metrics_install_metrics=True
   
   
  [root@master ~]# oc login -u system:admin
  [root@master ~]# vim files/10/sample-pv.yml
  apiVersion: v1
  kind: PersistentVolume
  metadata:
  name: metrics # 名字要改
  spec:
  capacity:
  storage: 5Gi
  accessModes:
  - ReadWriteOnce # 只能是Once
  nfs:
  path: /OCP_metrics # 共享要改
  server: services.lab.example.com # 共享要改
  persistentVolumeReclaimPolicy: Recycle
   
  [root@master test]# oc create -f files/10/sample-pv.yml
  [root@master test]# oc get pv | grep metrics
   
  # 切换机器
  [root@foundation0 ~]# ssh student@workstation
  [student@workstation ~]# cd DO280/labs/install-metrics/
  [student@workstation install-metrics]# vim inventory # 追加到该文件中以下内容
  openshift_metrics_image_version=v3.9
  openshift_metrics_heapster_requests_memory=300M
  openshift_metrics_hawkular_requests_memory=750M
  openshift_metrics_cassandra_requests_memory=750M
  openshift_metrics_cassandra_storage_type=pv
  openshift_metrics_cassandra_pvc_size=5Gi
  openshift_metrics_cassandra_pvc_prefix=metrics
  openshift_metrics_install_metrics=True
  openshift_metrics_image_prefix=registry.lab.example.com/openshift3/ose- # 该文件中有示例，简单修改/复制即可
   
   
  [student@workstation install-metrics]# ansible-playbook -i inventory /usr/share/ansible/openshift-ansible/playbooks/openshift-metrics/config.yml
   
   
  # 切换到master节点
  [root@master ~]# oc project openshift-infra
  [root@master ~]# oc get pods
  [root@master ~]# oc get route
  NAME HOST/PORT PATH SERVICES PORT TERMINATION WILDCARD
  hawkular-metrics hawkular-metrics.apps.lab.example.com hawkular-metrics <all> reencrypt None
   
   
  在workstations上用浏览器访问 https://hawkular-metrics.apps.lab.example.com,看到老鹰头即算成功



posted @ 2023-03-06 13:54  呆瓜小贼66  阅读(340)  评论(0)    收藏  举报