Ansible的任务执行控制

目录

一.循环迭代

1.简单循环

 2.循环散列或者字典列表

 二.条件

 三.触发器

 四.处理失败任务

---

一.循环迭代

1.简单循环

{{item}}迭代变量名词   loop： 赋值列表

- name:
  hosts: westos
  tasks:
    - name: echo
      debug:
        msg: "{{item}}"

      loop:
        - luck
        - bad

 2.循环散列或者字典列表

- name:
  hosts: westos
  tasks:
    - name:
      service:
        name: "{{item.name}}"
        state: "{{item.state}}"
      loop:
        - name: httpd
          state: restarted
        - name: firewalld
          state: stopped

 二.条件

when
 - 条件1
 - 条件2

- name:
  hosts: all
  tasks:
    - name: check file
      shell: test -e /mnt/test
      register: CHECKOUT
      ignore_errors: yes
    - name: debug
      debug:
        msg:  "/mnt/test is not exit"
      when: 
        - inventory_hostname in groups.westos
        - CHECKOUT.rc !=0

   - name: debug
      debug:
        msg:  "/mnt/test is  exit"
      when:
        - inventory_hostname in groups.westos
        - CHECKOUT.rc == 0

=	value==“字符串”  value = = 数字
>	value < 数字
<	---
<=	---
>=	---
!=	---
is not defined	----
is defined	value is defined
in	value is in value
not in	value is not in value
true   /bool变量	value                            /       value值为true
flase   /bool变量	not value                /       value值为flase
	value in value2   value值在value2

when:>
  1
  2

- name:
  hosts: all
  tasks:
    - name: check file
      shell: test -e /mnt/test
      register: CHECKOUT
      ignore_errors: yes
    - name: debug
      debug:
        msg:  "/mnt/test is not exit"
      when: >
         inventory_hostname in groups.westos
         or
         CHECKOUT.rc !=0

 练习

建立lvm.yml

建立大小为1500M的名为exam_lvm的lvm在westos组中

如果westos不存在请输出

vg westos is not exit

如果 westos大小不足1500M  请输出

vg westos is less than 1500M

并建立800M的lvm

输出这一步并没有准确做到  应该在学了block,rescue,和always去做

- name: create exam_lvm
  hosts: westos
  tasks:
    - name: check /dev/vdb
      debug:
        msg: vg westos is not exit
      when: ansible_facts['lvm']['vgs']['westos'] is not defined

    - name: create 1500M lvm
      lvol:
        vg: westos
        lv: exam_lvm
        size: 1500m
      when: ansible_facts['lvm']['vgs']['westos'] is  defined
      ignore_errors: yes
      register: lvmstart

    - name: create 800M lvm
      lvol:
        vg: westos
        lv: exam_lvm
        size: 800m
      ignore_errors: yes
      register: lvmstop
      when:
        - ansible_facts['lvm']['vgs']['westos'] is defined
        - lvmstart['rc'] !=0
    - name:
      debug:
        msg: "vg westos less than 1500M"

 

 三.触发器

notify:触发器遇到更改触发handlers
handlers: 触发器触发后执行的动作

- name:
  hosts: westos
  tasks:

    - name: install apache
      dnf:
        name: httpd
        state: present

    - name: configure httpd
      lineinfile:
        regexp: "^Listen"
        path: /etc/httpd/conf/httpd.conf
        line: "Listen 8080"
      notify:
          restart httpd
  handlers:
    - name: restart httpd
      service:
        name: httpd
        state: restarted
        enabled: yes

要触发重启服务，需要触发器前面的模块动作

 四.处理失败任务

ignore_errors

 

force_handlers:    yes/no    yes时当任务失败后play中止仍会调用触发器

 将被控机httpd 卸载，删除/etc/httpd,失败的任务，不影响触发重启

changed_when:     控制任务在何时报告它已经更改   true 已经更改和触发器一起用一直触发

 failed_when： true无论任务是否成功，都会显示失败  可加条件

 五.block

block:

rescue:

always:

 测试

建立playbook ~/westos.yml

建立大小为1500M的/dev/vdb1设备

如果/dev/vdb1不存在请输入  /dev/vdb1 is not exit

如果大小不足2G请输入 /dev/vdb1 is less then 2G

并建立800M大小的/dev/vdb1

次设挂载在/westos上