最近公司新上架了50台服务器,系统包括centos7.3和centos6.8,现在需要在这50台服务器中安装zabbix客户端,由于本人比较懒,所以就想到了自动化工具ansible。

具体是实现方法如下:

1、playbook代码如下:

[root@deploy playbook-test]# cat zabbix-agent.yml 

- hosts: web
  remote_user: root
  gather_facts: yes

  vars:

     IPDD: "{{ ansible_default_ipv4['address'] }}"

  tasks:
  - name: 开始拷贝源码包
    copy:
      #src: "/tmp/zabbix-2.2.23.tar.gz"
      #dest: "/tmp/zabbix-2.2.23.tar.gz"
      src: "{{ item.src }}"
      dest: "{{ item.dest }}"
    with_items:
      - { src: "/tmp/zabbix-2.2.23.tar.gz", dest: "/tmp/zabbix-2.2.23.tar.gz" }
      #- { src: "/tmp/zabbix-2.2.23.tar.gz", dest: "/tmp/zabbix-2.2.23.tar.gz" }
      #- { src: "/PATH/to/FILE", dest: "/tmp/zabbix-2.2.23.tar.gz" }

  - name: 开始解压源码包
    unarchive:
      src: "/tmp/zabbix-2.2.23.tar.gz"
      dest: "/tmp"
      copy: no

  - name: 创建zabbix组
    group:
      name: zabbix
      gid: 201
      system: yes
      state: present

  - name: 创建zabbix用户
    user:
      name: zabbix
      comment: Monitor User
      uid: 201
      group: zabbix
      shell: /sbin/nologin

  - name: 安装GCC-C++
    yum:
      name: gcc-c++
      state: latest

  - name: 开始安装zabbix客户端
    shell: cd /tmp/zabbix-2.2.23 && ./configure --sysconfdir=/etc/zabbix/ --sbindir=/usr/local/sbin/ --enable-agent && make && make install

  - name: 拷贝centos6启动脚本
    copy:
      src: "{{ item.src }}"
      dest: "{{ item.dest }}"
      mode: 0755
    with_items:
      - { src: "/tmp/zabbix-2.2.23/misc/init.d/fedora/core5/zabbix_agentd", dest: "/etc/init.d/zabbix_agentd" }
    when: ansible_distribution == 'CentOS' and ansible_distribution_major_version == "6"

  - name: 拷贝centos7启动脚本
    copy:
      src: "{{ item.src }}"
      dest: "{{ item.dest }}"
      mode: 0644
    with_items:
      - { src: "/tmp/zabbix_agentd.service", dest: "/lib/systemd/system/zabbix_agentd.service" }
    when: ansible_distribution == 'CentOS' and ansible_distribution_major_version == "7"

  - name: systemctl daemon-reload
    command: systemctl daemon-reload
    when: ansible_distribution == 'CentOS' and ansible_distribution_major_version == "7"

  - name: 拷贝配置文件
    template:
      src: /tmp/zabbix_agentd.conf.j2
      dest: /etc/zabbix/zabbix_agentd.conf
    tags:
      - change-conf
    notify:
      - restart zabbix-agent

  handlers:
  - name: restart zabbix-agent
    service: name=zabbix_agentd state=restarted enabled=yes

2、准备源码包及相关配置文件

root@deploy tmp]# ll

total 15160
drwxr-xr-x 13 1039 1039     4096 Jul 30  2018 zabbix-2.2.23
-rw-r--r--  1 root root 15504361 Jul  4 01:28 zabbix-2.2.23.tar.gz
-rw-r--r--  1 root root     7465 Jul  6 01:48 zabbix_agentd.conf.j2
-rw-r--r--  1 root root      412 Jul  6 01:13 zabbix_agentd.service

[root@deploy tmp]# grep -Ev "^#|^$" /etc/ansible/hosts 

[web]
192.168.131.10
192.168.131.15
192.168.131.129

[root@deploy tmp]# grep -Ev "^#|^$" zabbix_agentd.conf.j2 

PidFile=/tmp/zabbix_agentd.pid
LogFile=/tmp/zabbix_agentd.log
Server=192.168.131.100
ListenPort=10023
ServerActive=192.168.131.100
Hostname={{ IPDD }}
Timeout=30

[root@deploy tmp]# cat zabbix_agentd.service 

[Unit]
Description=Zabbix Agent
After=syslog.target
After=network.target

[Service]
Environment="CONFFILE=/etc/zabbix/zabbix_agentd.conf"
EnvironmentFile=-/etc/sysconfig/zabbix-agent
Type=forking
Restart=on-failure
PIDFile=/tmp/zabbix_agentd.pid
KillMode=control-group
ExecStart=/usr/local/sbin/zabbix_agentd -c $CONFFILE
ExecStop=/bin/kill -SIGTERM $MAINPID
RestartSec=10s

[Install]
WantedBy=multi-user.target

3、测试ansible主机与客户端通信是否正常(ansible主机需要和客户端主机做ssh互信)

[root@deploy playbook-test]# ansible web -m ping

192.168.131.15 | SUCCESS => {
    "ansible_facts": {
        "discovered_interpreter_python": "/usr/bin/python"
    }, 
    "changed": false, 
    "ping": "pong"
}
192.168.131.10 | SUCCESS => {
    "ansible_facts": {
        "discovered_interpreter_python": "/usr/bin/python"
    }, 
    "changed": false, 
    "ping": "pong"
}
192.168.131.129 | SUCCESS => {
    "ansible_facts": {
        "discovered_interpreter_python": "/usr/bin/python"
    }, 
    "changed": false, 
    "ping": "pong"
}

4、执行剧本(批量部署)

[root@deploy playbook-test]# ansible-playbook zabbix-agent.yml

5、测试客户端机器zabbix-agent是否成功安装并启动

[root@deploy playbook-test]# ansible web -m shell -a "ps -ef |grep zabbix_agentd|grep -v grep"

192.168.131.10 | CHANGED | rc=0 >>
zabbix    26025      1  0 13:49 ?        00:00:00 /usr/local/sbin/zabbix_agentd -c /etc/zabbix/zabbix_agentd.conf
zabbix    26026  26025  0 13:49 ?        00:00:00 /usr/local/sbin/zabbix_agentd: collector [idle 1 sec]
zabbix    26027  26025  0 13:49 ?        00:00:00 /usr/local/sbin/zabbix_agentd: listener #1 [waiting for connection]
zabbix    26028  26025  0 13:49 ?        00:00:00 /usr/local/sbin/zabbix_agentd: listener #2 [waiting for connection]
zabbix    26029  26025  0 13:49 ?        00:00:00 /usr/local/sbin/zabbix_agentd: listener #3 [waiting for connection]
zabbix    26030  26025  0 13:49 ?        00:00:00 /usr/local/sbin/zabbix_agentd: active checks #1 [idle 1 sec]

192.168.131.15 | CHANGED | rc=0 >>
zabbix    25956      1  0 13:49 ?        00:00:00 /usr/local/sbin/zabbix_agentd -c /etc/zabbix/zabbix_agentd.conf
zabbix    25957  25956  0 13:49 ?        00:00:00 /usr/local/sbin/zabbix_agentd: collector [idle 1 sec]
zabbix    25958  25956  0 13:49 ?        00:00:00 /usr/local/sbin/zabbix_agentd: listener #1 [waiting for connection]
zabbix    25959  25956  0 13:49 ?        00:00:00 /usr/local/sbin/zabbix_agentd: listener #2 [waiting for connection]
zabbix    25960  25956  0 13:49 ?        00:00:00 /usr/local/sbin/zabbix_agentd: listener #3 [waiting for connection]
zabbix    25961  25956  0 13:49 ?        00:00:00 /usr/local/sbin/zabbix_agentd: active checks #1 [idle 1 sec]

192.168.131.129 | CHANGED | rc=0 >>
zabbix   11258     1  0 06:49 ?        00:00:00 /usr/local/sbin/zabbix_agentd
zabbix   11260 11258  0 06:49 ?        00:00:00 /usr/local/sbin/zabbix_agentd: collector [idle 1 sec]
zabbix   11261 11258  0 06:49 ?        00:00:00 /usr/local/sbin/zabbix_agentd: listener #1 [waiting for connection]
zabbix   11262 11258  0 06:49 ?        00:00:00 /usr/local/sbin/zabbix_agentd: listener #2 [waiting for connection]
zabbix   11263 11258  0 06:49 ?        00:00:00 /usr/local/sbin/zabbix_agentd: listener #3 [waiting for connection]
zabbix   11264 11258  0 06:49 ?        00:00:00 /usr/local/sbin/zabbix_agentd: active checks #1 [getting list of active checks][object Object]

结果显示在客户端机器上已经有zabbix_agentd的进程在运行。


6、如果需要变更配置内容,比如更改zabbix客户端的监听端口为10023,此时不需要将整个playbook都执行一遍可以在playbook相应的任务下加入tag,执行时使用“--tags”参数,指定要执行的playbook片段。

[root@deploy playbook-test]# grep -Ev "^#|^$" /tmp/zabbix_agentd.conf.j2

PidFile=/tmp/zabbix_agentd.pid
LogFile=/tmp/zabbix_agentd.log
Server=192.168.131.100
ListenPort=10023
ServerActive=192.168.131.100
Hostname={{ IPDD }}
Timeout=30

[root@deploy playbook-test]# ansible-playbook zabbix-agent.yml --tags change-conf

PLAY [web] **************************************************************************************************************************************************

TASK [Gathering Facts] **************************************************************************************************************************************
ok: [192.168.131.15]
ok: [192.168.131.10]
ok: [192.168.131.129]

TASK [拷贝配置文件] ***********************************************************************************************************************************************
changed: [192.168.131.15]
changed: [192.168.131.10]
changed: [192.168.131.129]

RUNNING HANDLER [restart zabbix-agent] **********************************************************************************************************************
changed: [192.168.131.15]
changed: [192.168.131.10]
changed: [192.168.131.129]

PLAY RECAP **************************************************************************************************************************************************
192.168.131.10             : ok=3    changed=2    unreachable=0    failed=0    skipped=0    rescued=0    ignored=0   
192.168.131.129            : ok=3    changed=2    unreachable=0    failed=0    skipped=0    rescued=0    ignored=0   
192.168.131.15             : ok=3    changed=2    unreachable=0    failed=0    skipped=0    rescued=0    ignored=0[object Object]

结果显示配置文件已替换成功,并且跳过了其他步骤,只执行了拷贝配置文件的任务,配置文件改变,触发了重启zabbix agent的handler。

7、查看客户端监听的端口是否发生了变化

[root@deploy playbook-test]# ansible web -m shell -a "netstat -antlp | grep zabbix "

192.168.131.15 | CHANGED | rc=0 >>
tcp        0      0 0.0.0.0:10023           0.0.0.0:*               LISTEN      26358/zabbix_agentd 

192.168.131.10 | CHANGED | rc=0 >>
tcp        0      0 0.0.0.0:10023           0.0.0.0:*               LISTEN      26427/zabbix_agentd 

192.168.131.129 | CHANGED | rc=0 >>
tcp        0      0 0.0.0.0:10023               0.0.0.0:*                   LISTEN      11615/zabbix_agentd

ansible批量部署zabbix客户端到此结束,仅做记录,能力有限,存在很多不足,不喜勿喷。

后期更新通过ansible批量修改线上服务器用户密码。