本文来自我的github pages博客http://galengao.github.io/ 即www.gaohuirong.cn

摘要:

• swarm是docker原生的集群管理软件，与kubernetes比起来比较简单

1、部署

系统时centos7上

关闭防火墙 systemctl stop firewalld.service 

关闭selinux vi /etc/selinux/comfig

192.168.10.140 swarm manager

192.168.10.141 swarm node

192.168.10.142 swarm mode

2、分别在manager节点和node节点上安装docker

安装方式参照我的另一篇文章docker安装

yum updatetee /etc/yum.repos.d/docker.repo<

3、分别在manager节点和node节点上配置docker

vi /lib/systemd/system/docker.service# 修改ExecStart项为如下：ExecStart=/usr/bin/docker daemon -H 0.0.0.0:2375 -H unix:///var/run/docker.sock# 重新载入配置，使修改生效。systemctl daemon-reload# 重启docker。systemctl restart docker

4、在manager节点和node节点上push swarm镜像

[root@swarm1 ~]# docker pull swarmUsing default tag: latestlatest: Pulling from library/swarm1e61bbec5d24: Pull complete 8c7b2f6b74da: Pull complete 245a8db4f1e1: Pull complete Digest: sha256:661f2e4c9470e7f6238cebf603bcf5700c8b948894ac9e35f2cf6f63dcda723aStatus: Downloaded newer image for swarm:latest

5、创建集群token，获取全球唯一的 token，作为集群唯一标识

# 在任何节点都可以，但是要记住该值，以后要用到[root@swarm1 ~]# docker run --rm swarm createeca9b4ab85feb53f8a9676c72dd01b1a

6、加入集群

# 在manager也就是节点node1[root@swarm1 ~]# docker run -d swarm join -addr=192.168.10.140:2375 token://eca9b4ab85feb53f8a9676c72dd01b1a109da11914295c588c6afe5f83ab731bd0d0012897c39c311de89534e2f5bc13# node2上[root@swarm1 ~]# docker run -d swarm join -addr=192.168.10.141:2375 token://eca9b4ab85feb53f8a9676c72dd01b1a1da02eb6a00a8860eefe965a0aded446aebff8b502962c717dd3f494b546841a# node3上[root@swarm1 ~]# docker run -d swarm join -addr=192.168.10.142:2375 token://eca9b4ab85feb53f8a9676c72dd01b1ab5483c91bff0ad21e19700af51990d631e991f9d67188c7419f147652d494972

7、启动管理机

# 在管理机上执行:[root@swarm1 ~]# docker run -d -p 2376:2375 swarm manage token://eca9b4ab85feb53f8a9676c72dd01b1a3073a3dd59a5782f706d6481cfd1a36e8090f21764dfec2532899450bd719456

8、查看节点信息

# 本机上查看节点信息[root@swarm1 ~]# docker ps -aCONTAINER ID        IMAGE               COMMAND                  CREATED             STATUS              PORTS               NAMES1da02eb6a00a        swarm               "/swarm join -addr=19"   27 minutes ago      Up 27 minutes       2375/tcp            sick_bose# 查看集群所有节点信息，在任何一台机器上执行[root@swarm1 ~]# docker run --rm swarm list token://eca9b4ab85feb53f8a9676c72dd01b1a192.168.10.142:2375192.168.10.141:2375192.168.10.140:2375# 查看集群详细信息。在任何一台机器上执行:# 该IP地址是manager的地址[root@swarm1 ~]# docker -H 192.168.10.140:2376 infoContainers: 0 Running: 0 Paused: 0 Stopped: 0Images: 0Server Version: swarm/1.2.3Role: primaryStrategy: spreadFilters: health, port, containerslots, dependency, affinity, constraintNodes: 3 (unknown): 192.168.10.142:2375  └ ID:   └ Status: Pending  └ Containers: 0  └ Reserved CPUs: 0 / 0  └ Reserved Memory: 0 B / 0 B  └ Labels:   └ UpdatedAt: 2016-07-28T07:54:39Z  └ ServerVersion:  (unknown): 192.168.10.141:2375  └ ID:   └ Status: Pending  └ Containers: 0  └ Reserved CPUs: 0 / 0  └ Reserved Memory: 0 B / 0 B  └ Labels:   └ UpdatedAt: 2016-07-28T07:54:39Z  └ ServerVersion:  (unknown): 192.168.10.140:2375  └ ID:   └ Status: Pending  └ Containers: 0  └ Reserved CPUs: 0 / 0  └ Reserved Memory: 0 B / 0 B  └ Labels:   └ UpdatedAt: 2016-07-28T07:54:39Z  └ ServerVersion: Plugins:  Volume:  Network: Kernel Version: 3.10.0-229.el7.x86_64Operating System: linuxArchitecture: amd64CPUs: 0Total Memory: 0 BName: 3073a3dd59a5Docker Root Dir: Debug mode (client): falseDebug mode (server): falseWARNING: No kernel memory limit support