Ansible自动化运维笔记3(playbook)
1.基本语法###
playbook文件格式为yaml语法.示例如下:
1.1 nginx.yaml
---
- hosts: all
tasks:
- name: Install Nginx Package
yum: name=nginx state=present
- name: Copy Nginx.conf
template: src=./nginx.conf.j2 dest=/etc/nginx/nginx.conf owner=root group=root mode=0644 validate='nginx -t -c %s'
notify:
- Restart Nginx Service
handlers:
- name: Restart Nginx Service
service: name=nginx state=restarted
---第一行表示该文件是yaml文件,非必须,建议写上
- hosts:all 定义该playbook针对的目标主机,all表示针对所有主机,这个参数支持Ad-Hoc模式的所有参数
tasks: 定义该playbook所有的tasks集合- name: Install Nginx Package定义一个task的名称,建议根据task实际任务命名
yum: name=nginx state=present 定义一个状态的action,这里使用yum模块,实现nginx软件包的安装
第6行-第9行使用template模板去管理/etc/nginx/nginx.conf文件,owner,group定义该文件的属主及属组,使用validate参数指文件生成后使用nginx -t -c 检测配置文件语法,notify是触发handlers,如果同步后,文件md5值有变化的话会触发handler
第10-12行定一个一个handler状态让Nginx去重启,
1.2 主机清单文件
cat /tmp/hosts
[nginx]
192.168.1.1
192.168.1.2
[nginx:vars]
ansible_python_interpreter=/usr/bin/python2.6
1.3 nginx.conf.j2
user admin admin;
worker_processes 8;
worker_cpu_affinity {{ ansible_processor_cores }};
error_log /export/servers/nginx/logs/nginx_error.log warn;
pid /export/servers/nginx/run/nginx.pid;
worker_rlimit_nofile 65535;
events
{
use epoll;
worker_connections 65535;
}
http
{
include mime.types;
default_type application/octet-stream;
server_tokens on;
log_format main '$remote_addr - $remote_user [$time_local] "$http_x_forwarded_for" "$http_j_forwarded_for" '
'"$request" $status $bytes_sent '
'"$http_referer" "$http_user_agent" '
'"$gzip_ratio"';
#charset utf-8;
server_names_hash_bucket_size 128;
client_header_buffer_size 32k;
large_client_header_buffers 4 32k;
client_max_body_size 300m;
sendfile on;
tcp_nopush on;
keepalive_timeout 0;
tcp_nodelay on;
client_body_buffer_size 512k;
fastcgi_intercept_errors on;
proxy_connect_timeout 90;
proxy_read_timeout 180;
proxy_send_timeout 180;
proxy_buffer_size 256k;
proxy_buffers 4 256k;
proxy_busy_buffers_size 256k;
proxy_temp_file_write_size 256k;
proxy_intercept_errors on;
server_name_in_redirect off;
proxy_hide_header X-Powered-By;
gzip on;
gzip_min_length 100;
gzip_buffers 4 16k;
gzip_http_version 1.0;
gzip_comp_level 9;
gzip_types text/plain application/x-javascript text/css application/xml;
gzip_vary on;
gzip_proxied any;
include domains/*;
###########status#########
# server
# {
# listen 80;
# server_name status.360buy.com;
# location / {
# stub_status on;
# access_log off;
# }
# }
}
1.hosts文件里面定义了一个nginx组,里面有2个IP
2.组变量ansible_python_interpreter是ansible自带的影藏变量,是facts套件提供的;如果目标机器上python版本多,指定一个版本来运行
3.nginx.conf.j2是nginx.conf的模板文件,里面只针对worker_processes参数通过facts套件中的cpu核心数生成,其他配置都是默认的
1.4 检查yaml文件的语法是否正确
$ ansible-playbook -i /tmp/hosts nginx.yaml --syntax-check
1.5 检查yaml文件中的tasks任务
$ ansible-playbook -i /tmp/hosts nginx.yaml --list-task
1.6 检查yaml文件中的生效主机
$ ansible-playbook -i /tmp/hosts nginx.yaml --list-hosts
1.7 运行playbook
$ ansible-playbook -i /tmp/hosts nginx.yaml
1.8 运行playbook里面特定的某个task,从某个task开始运行
$ ansible-playbook -i /tmp/hosts nginx.yaml --start-at-task='Copy Nginx.conf'
2.变量与引用###
2.1 通过inventory文件定义主机以及主机组变量
太简单了,写了好多次了,这次不写了
2.2 通过playbook文件当前目录下新建group_vars和host_vars这2个目录(目录名字固定,不能改)
$ cat group_vars/nginx
---
key: NGINX
在当前目录下运行playbook的时候,会自动去找group_vars和host_vars这2个目录
2.3 通过ansible-playbook命令行参数传入
(1)命令行传递变量
$ ansible-playbook -i /tmp/hosts nginx.yaml -e "key=KEY"
(2)命令行传递变量文件
$ cat var.yaml
---
key: YAML
$ cat var.json
{"key":"JSON"}
$ ansible-playbook -i /tmp/hosts nginx.yaml -e "@var.json"
$ ansible-playbook -i /tmp/hosts nginx.yaml -e "@var.yaml"
2.4 在playbook文件内使用vars(用的很少)
---
- hosts: all
vars:
key: Ansible
tasks:
- name: xxx
deubg: msg="The {{ key}}"
2.5 在playbook文件内使用vars_files引用外部变量文件(也可以是json文件,后期开发api模式主要用这种方式)
---
- hosts: all
vars_files:
- var.yaml
- var.json
2.6 使用register内的变量
ansible的task之间还可以互相传递数据,把第一个task执行的结果register注册为变量然后传递给第二个task
---
- hosts: all
tasks:
- name: register variable
shell: hostname
register: info
- name: display variable
debug: msg="The msg is {{ info }}"
debug: msg="The msg is {{ info['stdout'] }}"
实际跑下看下结果(-l呢是在匹配主机inventory里面再进一步筛选,只跑192.168.1.118主机)
$ ansible-playbook -i /tmp/hosts variable.yaml -l 192.168.1.118
3.循环###
3.1 标准loops
分别打印one two这2个值
---
- hosts:all
tasks:
- name: debug loops
debug: msg = "name---> {{ item }}"
with_items:
- one
- two
with_items:值呢是python list数据结构,每个task会循环读取list里面的值,key的名称是item,当然支持列表里面嵌套字典,例子如下
---
- hosts:all
tasks:
- name: debug loops
debug: msg = "name-----> {{ item.key }} value---->{{ item.value }}"
with_items:
- {key:"one",value:"value1"}
- {key:"two",value:"value2"}
3.2 嵌套loops
实现一堆多或者多对多的合并
---
- hosts:all
tasks:
- name: debug loops
debug: msg="name ----->{{ item[0] }} value----->{{ item[1] }}"
with_nested:
- ['A']
- ['a','b','c']
3.3 字典循环
---
- hosts:all
tasks:
- name: debug loops
debug: msg="name ----->{{ item.key }} value----->{{ item.value }}"
with_dict: user
3.4 文件循环
---
- hosts:all
tasks:
- name: debug loops
debug: msg="{{ item }}"
with_fileglob:
- /tmp/*.yaml
3.5 随机循环
---
- hosts:all
tasks:
- name: debug loops
debug: msg="{{ item }}"
with_random_choice:
- "ansible1"
- "ansible2"
- "ansible3"
3.6 条件判断循环
---
- hosts:all
tasks:
- name: debug loops
shell: cat /root/ansible
register: host
until: host.stdout.startswith('Master')
retries:5
delay:5
5秒执行一次cat /root/ansible,将结果注册给变量host,判断host.stdout的内容是否以Master开头,条件成立,task运行完成,条件不成立,5秒后重试,5次还不成立,task运行失败
4.lookups###
从外部拉取信息,定义给一个变量的形式,lookups插件
