LINUX学习------5.5 Linux中Anisble中的任务执行控制

5.5.1 循环迭代任务

1.简单循环

loop:        ##赋值列表
          - value1
          - value2
          - ...

{{item}}    ##迭代变量名称

vim test.yml
    ---
    - name: create file
      hosts: all
      tasks:
        - name: file
          file:
            path: /mnt/{{ item }}
            state: touch
            mode: '755'
            owner: westos

          loop:
            - westos_file1
            - westos_file2

2.循环散列或字典列表

vim test.yml
    ---
    - name: create file
      hosts: westos
      tasks:
        - name: file module
          service:
            name: "{{ item.name}}"
            state: "{{ item.state }}"
          loop:
            - name: httpd
              state: started
            - name: vsftpd
              state: stopped

5.5.2 条件

先执行在判断（shell 先判断再执行）

条件判断
    =			value == "字符串",value == 数字
    <			value <  数字
    >			value >  数字
    <=			value <= 数字
    >=			value >= 数字 
    !=			value != 数字
    is defined value	value is defined	            变量存在
    is not defined	value is not defined	    变量不存在
    in                  value is in value           变量为    
    not in              value is not in value       变量不为
    bool变量 为true      value			    value的值为true
    bool变量 false	not value		    value的值为false
			value in value2		    value的值在value2列表中

示例：检测westos清单主机中/mnt/file是否存在并显示提示内容

vim test.yml
    ---
    - name: test
      hosts: westos
      tasks:
        - name: test
          shell: test -e /mnt/file
          ignore_errors: yes		        #忽略任务失败使下面的任务继续运行
          register: OUTPUT			#取失败结果rc=0/1的值

        - name: show messages
          debug:
            msg: /mnt/file is not find
          when: OUTPUT.rc != 0
    
        - name: show messages
          debug:
            msg: /mnt/file is exit
          when: OUTPUT.rc == 0

多条条件组合

when：
          条件1 and 条件2

when:

  - 条件1
  - 条件2

---

when:
  条件1 or 条件2

when: >
  条件1
  or
  条件2

测试题：
        建立playbook  ~/ansibles/lvm.yml要求如下:
        *建立大小为1500M名为exam_lvm的lvm 在westos组中
        *如果westos不存在请输出:
         vg westos  is not exist
        *如果westos大小不足1500M请输出：
         vg westos is less then 1500M
         并建立800M大小的lvm

虚拟件加入不同大小硬盘并设置分区
fdisk /dev/vdb
pvcreate /dev/vdb1            ##添加pv 
vgcreate westos /dev/vdb1     ##添加vg

---
- name: test
  hosts: westos
  tasks:
    - name: create 1500M
      lvol:                    ##设置lvm逻辑卷
        lv: exam_lvm
        vg: westos
        size: 1500M
        pvs: /dev/vdb1
      when: ansible_facts['lvm']['vgs']['westos'] is defined    ##判定主机真实变量下参数是否存在
      ignore_errors: yes        ##错误跳过并执行以下命令
      register: OUTPUT          ##捕捉此命令下rc的值

    - name: message
      debug:
        msg: vg westos is less then 1500M
      when:
        - ansible_facts['lvm']['vgs']['westos'] is defined
        - OUTPUT.rc is defined    ##成功创建的主机rc值不一定存在需判定rc是否存在

    - name: create 800M
      lvol:
        lv: exam_lvm
        vg: westos
        size: 800M
        pvs: /dev/vdb1
      when:
        - ansible_facts['lvm']['vgs']['westos'] is defined
        - OUTPUT.rc is defined

    - name: show not exist
      debug:
        msg: vg westos is not exist
      when: ansible_facts['lvm']['vgs']['westos'] is not defined

5.5.3 触发器

notify:        触发器当遇到更改是触发handlers

handlers：    触发器触发后执行的动作

示例：安装并启动apache，当输入端口号时重启apache服务并设置火墙允许此端口

---
- name: test
  hosts: westos
  tasks:
    - name: install apache
      dnf:
        name: httpd
        state: latest

    - name: start service
      service:
        name: httpd
        state: restarted
        enabled: yes

    - name: firewalld
      firewalld:
        service: http
        permanent: yes
        state: enabled
        immediate: yes

    - name: configure port
      lineinfile:
        path: /etc/httpd/conf/httpd.conf
        line: Listen "{{port}}"
        regexp: "^Listen"
        backrefs: yes
      notify:
        - firewalld
        - service
  handlers:
    - name: firewalld
      firewalld:
        port: "{{port}}/tcp"
        permanent: yes
        state: enabled
        immediate: yes
    - name: service
      service:
        name: httpd
        state: restarted
        enabled: yes

5.5.4 处理任务失败

1. ignore_errors

当play遇到任务失败不会终止playbook

ignore_errors: yes     ##将会忽略当前任务失败继续运行下面的任务，默认为no，与play片段中的name同级

2.  force_handlers

 当任务失败后play被终止也会调用触发器进程

 force_handlers: yes    ##与playbook中的hosts、tasks同级

3. changed_when

强制控制所在play片段是否报告changed

changed_when: true | false           

 4. failed_when

当符合条件时强制任务失败

failed_when: true         ##强制play任务失败

failed_when: false         ##强制play任务成功

failed_when: 条件         ##满足条件时，强制play任务失败

5. block、rescue、always 关键字

block	定义首要要运行的任务，与rescue成对出现
rescue	定义当block中出现失败任务后运行的任务，与block成对出现
always	定义最终独立运行的任务