说明:
系统为CentOS6.5
manager 192.168.10.1
web1 192.168.10.2
web2 192.168.10.3
安装ansible
wget http://mirrors.zju.edu.cn/epel/6/x86_64/epel-release-6-8.noarch.rpm rpm -ivh epel-release-6-8.noarch.rpm #双机互信 [root@manager ~]# ssh-keygen -t rsa -f ~/.ssh/id_rsa -P '' for i in web1 web2;do ssh-copy-id -i root@$i;done #安装ansible # yum install ansible # ansible --version ansible 1.8.2
ansible帮助文档:https://galaxy.ansible.com/explore#/ http://www.ansible.cn/docs/
#ansible 怎么用?测试几个常用的命令
ansible ansible-galaxy ansible-pull ansible-vault
ansible-doc ansible-playbook ansible-shell
# ansible -h #ansible的用法,后面可加的参数,比如connection、sudo、module-name、timeout
# ansible-doc -l #ansible的工具书,查看支持的modules,可以说是命令的目录
# ansible-doc ping #查看ping模块的使用方法,同理查看file的用法# ansible-doc file
> PING
A trivial test module, this module always returns `pong' on
successful contact. It does not make sense in playbooks, but it is
useful from `/usr/bin/ansible'
# Test 'webservers' status
cd /etc/ansible/ #修改ansible的hosts文件
vim hosts #注释所有,添加以下
[web1]
192.168.10.2
[web2]
192.168.10.3
ansible web1 -a uptime #等同于ansible web1 -m command -a uptime
ansible web1 -m yum -a "name=dstat state=latest" #执行yum install dstat
ansible web1 -m raw -a "rpm -qa|grep dstat" #执行rpm -qa |grep dstat
ansible web1 -m shell -a "service mysqld restart" #执行service mysqld restart
ansible web1 -m service -a "name=mysqld state=stopped" #执行service mysqld stop
ansible web1 -m setup #查看web1的综合信息
#小测试一下,看看结果
[root@manager ~]# ansible web1 -m ping [WARNING]: The version of gmp you have installed has a known issue regarding timing vulnerabilities when used with pycrypto. If possible, you should update it (i.e. yum update gmp). 192.168.10.2 | success >> { "changed": false, "ping": "pong" }
#那么ansible到底怎么用呢,还是上面已经写了的过程(绿色标记的),ansible默认执行的是command,可以省略,如果不是command,那就ansible-doc -l里面查找要用的模块命令,比如file。
那么ansible web1 -m file 后面怎么写?
# ansible-doc file 查看file模块的帮助,里面有详细的介绍,例:
file: path=/etc/foo.conf owner=foo group=foo mode=0644
#下面举个栗子,ansible file的用法,前提是web1的/tmp下有file这个文件,并且web1有mysql用户和组
# ansible web1 -m file -a "path=/tmp/file owner=mysql group=mysql mode=0755" 192.168.10.2 | success >> { "changed": true, "gid": 500, "group": "mysql", "mode": "0755", "owner": "mysql", "path": "/tmp/path", "size": 20, "state": "file", "uid": 500 }
#ansible可用的命令很多,所有的命令都可以这样查看具体的用法
#下面只是一个简单的发布例子,或者说是思路,具体的根据公司自己情况做改动
如果要做的比较完善,那么需加入对输入的路径和文件做检测,并且对线上正式机的备份做更详细些,这个不支持回滚,如果更新出错,需要到web服务器备份的文件夹backup再次覆盖线上文件
[root@manager ~]# cat publish.sh #!/bin/bash # echo "please choose server web1 or web2: " echo "1:web1 2:web2 3:QUIT" read -p "\>" SERVER case $SERVER in 1) read -p "PATH:" DIR NEWDIR=`echo $DIR|sed 's/var/home/'` #发布机的目录是/home/www而不是/var/www,根据自己环境 ansible web1 -a "cp $NEWDIR /backup" #把web1上文件做一个备份,复制到/backup ansible web1 -m copy -a "src=$DIR dest=$NEWDIR" #把当前路径的文件copy到web1上 ;; 2) read -p "PATH:" DIR ansible web2 -a "cp $DIR /backup" ansible web2 -m copy -a "src=$DIR dest=$DIR" ;; 3) exit ;; *) echo "you are wrong" esac