Docker入门——Docker swarm集群

Doker集群：

• Docker swarm 是docker官方的三大项目之一，提供docker容器集群服务，是docker官方对容器云生态支持的核心方案，Docker Swarm项目开始于2014年，是Docker公司推出的第一个容器集群项目。使用docker swarm可以将多个docker主机封装为一个大型的虚拟docker主机，swarm集群提供给用户管理集群内所有容器的操作接口与使用一台docker主机几乎相同，从而可以快速打造一套容器云服务。

1. 集群搭建

1. 初始化集群

 [root@server1 ~]# systemctl start docker
 [root@server1 ~]# docker swarm init # 初始化集群
 Swarm initialized: current node (rai2qpd7sf6bugqc3ia7pd99t) is now a  manager.

 To add a worker to this swarm, run the following command:
 # 执行如下命令，为集群添加节点
    docker swarm join \
    --token SWMTKN-1-3uppnqtjkks8ktwys7nfvwbq3m3uqqrz7d3ojlietsgssgu4i7-9qa91w39e44uxt56hpbo0tpgi \
    172.25.21.1:2377

To add a manager to this swarm, run 'docker swarm join-token manager' and follow the instructions.

2. 添加节点

 [root@server2 ~]# docker swarm join \
 >     --token SWMTKN-1-3uppnqtjkks8ktwys7nfvwbq3m3uqqrz7d3ojlietsgssgu4i7-9qa91w39e44uxt56hpbo0tpgi \
 >     172.25.21.1:2377
 This node joined a swarm as a worker.

 [root@server3 ~]# docker swarm join \
 >     --token SWMTKN-1-3uppnqtjkks8ktwys7nfvwbq3m3uqqrz7d3ojlietsgssgu4i7-9qa91w39e44uxt56hpbo0tpgi \
 >     172.25.21.1:2377
 This node joined a swarm as a worker.

3. 在manager端查看集群节点信息

 [root@server1 ~]# docker node ls
 ID                           HOSTNAME  STATUS  AVAILABILITY  MANAGER STATUS
 ckqm9zm2pnmntnekwv2de2398    server2   Ready   Active        
 rai2qpd7sf6bugqc3ia7pd99t *  server1   Ready   Active        Leader
 s4es8qh4qant0zkfh1ciu9hpe    server3   Ready   Active    

2. 进行服务部署

 [root@server1 ~]# docker service create --name nginx --publish 80:80 --replicas 3 test.com/nginx:v1 
                 # --name 服务名称
                 # --publish 端口映射
                 # --replicas 副本个数
 v068oxv07bv60skpaorjrwb7x
 # 建立成功查看服务状态
 [root@server1 ~]# docker service ls
 ID            NAME   MODE        REPLICAS  IMAGE
 v068oxv07bv6  nginx  replicated  3/3       test.com/nginx:v1 
 # 查看服务的具体信息（分别在三个节点上运行）
 [root@server1 ~]# docker service ps nginx 
 ID            NAME     IMAGE              NODE     DESIRED STATE   CURRENT STATE          ERROR  PORTS
 vq468p2hyk92  nginx.1  test.com/nginx:v1  server3  Running         Running 4 seconds ago         
 q4fqojzhm3ty  nginx.2  test.com/nginx:v1  server1  Running         Running 4 seconds ago         
 k3vxbnr5dpqf  nginx.3  test.com/nginx:v1  server2  Running         Running 5 seconds ago 

1. 进行服务的扩容：

 # 将服务扩容成 6 个
 [root@server1 ~]# docker service scale nginx=6
 nginx scaled to 6
 # 扩容后查看服务状态（此刻为starting）
 [root@server1 ~]# docker service ps nginx 
 ID            NAME     IMAGE              NODE     DESIRED STATE   CURRENT STATE                    ERROR  PORTS
 vq468p2hyk92  nginx.1  test.com/nginx:v1  server3  Running         Running 9 minutes ago                   
 q4fqojzhm3ty  nginx.2  test.com/nginx:v1  server1  Running         Running 9 minutes ago                   
 k3vxbnr5dpqf  nginx.3  test.com/nginx:v1  server2  Running         Running 9 minutes ago                   
 sidqa0ps5wnn  nginx.4  test.com/nginx:v1  server3  Running         Starting less than a second ago         
 fyula2qmi3j1  nginx.5  test.com/nginx:v1  server1  Running         Starting 1 second ago                   
 nivx3r11b0aq  nginx.6  test.com/nginx:v1  server2  Running         Starting 2 seconds ago             
 # 再次观察，可看到服务启动成功 
 [root@server1 ~]# docker service ps nginx 
 ID            NAME     IMAGE              NODE     DESIRED STATE   CURRENT STATE           ERROR  PORTS
 vq468p2hyk92  nginx.1  test.com/nginx:v1  server3  Running         Running 9 minutes ago          
 q4fqojzhm3ty  nginx.2  test.com/nginx:v1  server1  Running         Running 9 minutes ago           
 k3vxbnr5dpqf  nginx.3  test.com/nginx:v1  server2  Running         Running 9 minutes ago          
 sidqa0ps5wnn  nginx.4  test.com/nginx:v1  server3  Running         Running 10 seconds ago         
 fyula2qmi3j1  nginx.5  test.com/nginx:v1  server1  Running         Running 11 seconds ago         
 nivx3r11b0aq  nginx.6  test.com/nginx:v1  server2  Running         Running 12 seconds ago         

2. 部署监控容器

 # 从私有仓库中获取镜像
 [root@server1 ~]# docker pull test.com/visualizer
 Using default tag: latest
 latest: Pulling from visualizer
 88286f41530e: Pull complete 

 cea94055e43e: Pull complete 
 Digest:  sha256:ea603c00f25b18c9a3aa49703c47bd5ca6c105637fe6f4887b606043c66ab528
 Status: Downloaded newer image for test.com/visualizer:latest
 # 进行服务建立
 [root@server1 ~]# docker service create \
 > --name=viz \         # 名称
 > --publish=8080:8080/tcp \   # 端口映射
 > --constraint=node.role==manager \  # 
 > --mount=type=bind,src=/var/run/docker.sock,dst=/var/run/docker.sock \
 > test.com/visualizer  # 镜像
 rgizfcjvt4s1qijp57cmmr80w
 # 建立成功，查看容器基本信息
 [root@server1 ~]# docker ps
 CONTAINER ID         IMAGE                                                                                         COMMAND                  CREATED              STATUS                            PORTS               NAMES
 a801ae7086cf        test.com/visualizer@sha256:ea603c00f25b18c9a3aa49703c47bd5ca6c105637fe6f4887b606043c66ab528   "npm start"              6 seconds ago       Up 4 seconds (health: starting)   8080/tcp            viz.1.p2x0cl1ep910bpp1giu6trxp6

在浏览器中进行监测：

3. 进行服务缩容

 # 将服务的容器数缩减为3个
 [root@server1 ~]# docker service scale nginx=3
 nginx scaled to 3
 # 查看服务信息，副本数变成3
 [root@server1 ~]# docker service ls
 ID            NAME   MODE        REPLICAS  IMAGE
 rgizfcjvt4s1  viz    replicated  1/1       test.com/visualizer:latest
 v068oxv07bv6  nginx  replicated  3/3       test.com/nginx:v1 
 # 查看服务具体信息
 [root@server1 ~]# docker service ps nginx 
 ID            NAME     IMAGE              NODE     DESIRED STATE  CURRENT STATE           ERROR  PORTS
 q4fqojzhm3ty  nginx.2  test.com/nginx:v1  server1  Running        Running 26 minutes ago         
 k3vxbnr5dpqf  nginx.3  test.com/nginx:v1  server2  Running        Running 26 minutes ago         
 sidqa0ps5wnn  nginx.4  test.com/nginx:v1  server3  Running        Running 16 minutes ago   

在监控端进行查看

4. 进行集群信息监测

为各个节点创建默认发布文件

 # 编辑默发布页面
 [root@server1 ~]# echo '
server1
' > index.html
 [root@server1 ~]# cat index.html 
 
server1

 查看容器信息
 [root@server1 ~]# docker ps
 CONTAINER ID         IMAGE                                                                                         COMMAND                  CREATED              STATUS                   PORTS               NAMES
 a801ae7086cf        test.com/visualizer@sha256:ea603c00f25b18c9a3aa49703c47bd5ca6c105637fe6f4887b606043c66ab528   "npm start"              8 minutes ago       Up 8 minutes (healthy)   8080/tcp            viz.1.p2x0cl1ep910bpp1giu6trxp6
 82aa12ca3102        test.com/nginx@sha256:2de9d5fc6585b3f330ff5f2c323d2a4006a49a476729bbc0910b695771526e3f        "nginx -g 'daemon ..."   29 minutes ago      Up 29 minutes            80/tcp              nginx.2.q4fqojzhm3tyi90q3slx6ciy2
 # 将默认发布页面传到容器中的指定位置
 [root@server1 ~]# docker container cp index.html nginx.2.q4fqojzhm3tyi90q3slx6ciy2:/usr/share/nginx/html

在客户端进行测试：
进行测试：

 [root@foundation21 docker]# for i in {1..12}; do curl 172.25.21.1; done
 <h1>server2<h1/>
 <h1>server3<h1/>
 <h1>server1<h1/>
 <h1>server2<h1/>
 <h1>server3<h1/>
 <h1>server1<h1/>
 <h1>server2<h1/>
 <h1>server3<h1/>
 <h1>server1<h1/>
 <h1>server2<h1/>
 <h1>server3<h1/>
 <h1>server1<h1/>

自动进行负载均衡

5. 进行高可用测试

手动让节点3（server3）宕机，进行测试

 [root@server3 ~]# systemctl stop docker.service 

在swarm manager 端进行查看：

 查看服务信息，副本数没有变化
 [root@server1 ~]# docker service ls
 ID            NAME   MODE        REPLICAS  IMAGE
 rgizfcjvt4s1  viz    replicated  1/1       test.com/visualizer:latest
 v068oxv07bv6  nginx  replicated  3/3       test.com/nginx:v1
 # 通过ps 了看到server3上的容器宕掉了，server2 在多建立一个容器
 [root@server1 ~]# docker service ps nginx 
 ID            NAME         IMAGE              NODE     DESIRED STATE  CURRENT STATE           ERROR  PORTS
 q4fqojzhm3ty  nginx.2      test.com/nginx:v1  server1  Running        Running 53 minutes ago         
 k3vxbnr5dpqf  nginx.3      test.com/nginx:v1  server2  Running        Running 53 minutes ago         
 m9770t1100q2  nginx.4      test.com/nginx:v1  server2  Running        Running 9 seconds ago          
 sidqa0ps5wnn   \_ nginx.4  test.com/nginx:v1  server3  Shutdown       Running 26 seconds ago   

在监控页面进行查看

恢复server3后，服务是不会自动迁移的

6. 进行服务的滚动更新

 将服务的副本数扩展到30个
 [root@server1 ~]# docker service scale nginx=30
 nginx scaled to 30
 查看服务列表，nginx副本30全alive
 [root@server1 ~]# docker service ls
 ID            NAME   MODE        REPLICAS  IMAGE
 rgizfcjvt4s1  viz    replicated  1/1       test.com/visualizer:latest
 v068oxv07bv6  nginx  replicated  30/30     test.com/nginx:v1

在监控页面进行查看

 [root@server1 ~]# docker service update --image test.com/rhel7:v1 --update-parallelism 3 --update-delay 10s nginx 
 nginx

在监控页面进行查看

更新成功后：
更新成功后，镜像换取完成

 [root@server1 ~]# docker service ls
 ID            NAME   MODE        REPLICAS  IMAGE
 rgizfcjvt4s1  viz    replicated  1/1       test.com/visualizer:latest
 v068oxv07bv6  nginx  replicated  30/30     test.com/rhel7:v1
 [root@server1 ~]# docker service ps nginx 
 ID            NAME          IMAGE              NODE     DESIRED STATE  CURRENT STATE                ERROR  PORTS
 xqwutjl221o2  nginx.1       test.com/rhel7:v1  server3  Running        Running about a minute ago          
 v46h5o48oaui   \_ nginx.1   test.com/nginx:v1  server3  Shutdown       Shutdown about a minute ago         
 xdtwrrzsx1d5  nginx.2       test.com/rhel7:v1  server1  Running        Running 40 seconds ago              
 q4fqojzhm3ty   \_ nginx.2   test.com/nginx:v1  server1  Shutdown       Shutdown 43 seconds ago             
 uaefrls6ejda  nginx.3       test.com/rhel7:v1  server1  Running        Running 56 seconds ago              
 k3vxbnr5dpqf   \_ nginx.3   test.com/nginx:v1  server2  Shutdown       Shutdown 59 seconds ago             
 ......................省略...........................

可在监控界面进行查看