前言
在上一篇教程中我们已经实现了使用ansible-playbook批量在远程主机上部署zabbix客户端并正常运行，现在我们再次通过ansible-playbook给客户端主机批量增加zabbix监控项目配置（创建监控项目示例：自动发现远程主机监听的TCP端口、监控远程主机的TCP连接数状态）。
Ansible-playbook 配置
在原有的基础目录上创建一个configure角色以及ansible的各个模块任务目录列表，通过ansible-playbook调用入口文件zabbix_configure.yml，使得configure能够调用各个模块功能来完成同步所有zabbix的配置
[root@ansible /etc/ansible/zabbix_rhel/zabbix_agent ]# tree roles/configure/ roles/configure/ ├── files │ └── zabbix_scripts │ ├── discovery_tcp_port.sh │ └── tcp_connect_status.sh ├── handlers │ └── main.yml ├── meta ├── tasks │ ├── 01-sync-clock.yml │ ├── 02-allow-sudo.yml │ ├── 03-sync-conf_files.yml │ └── main.yml ├── templates │ ├── Userparameter_script.conf │ └── zabbix_agentd.conf └── vars └── main.yml 7 directories, 10 files [root@ansible /etc/ansible/zabbix_rhel/zabbix_agent ]# ls roles zabbix_configure.yml zabbix_delete.yml zabbix_install.yml
1、定义ansible程序入口调用文件
>> zabbix_configure.yml
--- - hosts: testhosts remote_user: root gather_facts: True roles: - configure
2、定义tasks任务列表
>> 同步时钟并添加计划任务(01-sync-clock.yml )
---
- name: Install ntpdate software
yum: name=ntpdate state=present
- name: Synchronization clock cron job
cron:
name: Sync clock
job: /usr/sbin/ntpdate {{ ntpserver }} &>/dev/null && hwclock -w
minute: 30
hour: 7
- name: Restart crond
service: name=crond state=restarted
- name: Running ntpdate to synchronization time
shell: /usr/sbin/ntpdate {{ ntpserver }} && hwclock -w
>> 定义允许zabbix用户使用sudo执行命令的task任务(02-allow-sudo.yml)
--- - name: Add allow zabbix user to nopasswd exec sudo lineinfile: dest: /etc/sudoers regexp: "^zabbix" insertafter: "^root" line: "zabbix ALL=(ALL) NOPASSWD: ALL" - name: Add allow zabbix user to nopasswd exec sudo lineinfile: dest: /etc/sudoers regexp: "^Defaults:zabbix" line: "Defaults:zabbix !requiretty"
修改zabbix用户允许使用sudo执行命令，并且不需要输入密码，该task任务可以重复执行，并且不会增加重复的配置
>> 定义同步所有zabbix配置文件的task任务(03-sync-conf_files.yml)
---
- name: Create zabbix scripts directory
file: dest={{ zabbix_basedir }}/scripts state=directory owner=root group=root mode=0755 recurse=yes
- name: Copy zabbix monitor scripts
copy: src=zabbix_scripts/ dest={{ zabbix_basedir }}/scripts/ owner=root group=root mode=0755
- name: Copy zabbix_agentd.conf file
template: src=zabbix_agentd.conf dest={{ zabbix_basedir }}/etc/zabbix_agentd.conf
owner=root group=root mode=0644
notify: restart zabbix_agentd
- name: Copy Userparameter_script.conf file
template: src=Userparameter_script.conf dest={{ zabbix_basedir }}/etc/zabbix_agentd.conf.d/Userparameter_script.conf
owner=root group=root mode=0644
notify: restart zabbix_agentd
>> 定义tasks任务列表调用接口文件(main.yml)
--- - include: 01-sync-clock.yml - include: 02-allow-sudo.yml - include: 03-sync-conf_files.yml
3、定义vars变量文件
在这里定义zabbix服务端主机名或IP地址，ntp服务器地址等
# cat roles/configure/vars/main.yml ntpserver: 10.17.87.8 zabbix_basedir: /usr/local/zabbix zabbix_server_ip: 10.17.81.120
4、定义handlers任务文件
handlers任务在同步配置文件时重启zabbix_agentd服务使配置生效
# cat roles/configure/handlers/main.yml --- - name: restart zabbix_agentd service: name=zabbix_agentd state=restarted
5、将zabbix脚本统一放到files/zabbix_scripts/目录下
# ls roles/configure/files/zabbix_scripts/ discovery_tcp_port.sh tcp_connect_status.sh
>> discovery_tcp_port.sh为自动发现客户端主机TCP端口脚本
#!/bin/bash #Author: HMLinux Email: [email protected] #port_array=`netstat -tnlp | sed -e '1,2d' -e '/-/d' | awk '{print $4}' | awk -F':' '{if($NF~/^[0-9]*$/) print $NF}' | sort -n | u niq` port_array=`netstat -ntlp | sed -e '1,2d' -e '/-/d' | awk '{print $4" "$NF}' | awk -F'[:/ ]+' '($NF !~ /^[0-9]*$/) && ($2>18) {pri nt $2" "$NF}' |sort -g|uniq` tcp_ports=(`echo "$port_array"|cut -d" " -f1`) proc_name=(`echo "$port_array"|cut -d" " -f2`) length=${#tcp_ports[@]} printf "{\n" printf '\t'"\"data\":[" for ((i=0;i<$length;i++)) do printf '\n\t\t{' printf '\n\t\t\t' printf "\"{#TCP_PORT}\":\"${tcp_ports[$i]}\"," printf '\n\t\t\t' printf "\"{#TCP_NAME}\":\"${proc_name[$i]}\"}" if [ $i -lt $[$length-1] ];then printf ',' fi done printf "\n\t]\n" printf "}\n"
>> tcp_connect_status.sh为监控客户端主机TCP连接状态脚本
#!/bin/bash #Author: HMLinux Email: [email protected] parameter_l=$1 parameter_u=$(echo $parameter_l | tr '[:lower:]' '[:upper:]') ptcp_status=$(/bin/netstat -an|awk '/^tcp/{++S[$NF]}END{for(a in S) print a,S[a]}' | awk '/'''$parameter_u'''/{print $2}') case $parameter_l in listen) if [ "$ptcp_status" == "" ];then echo 0 else echo $ptcp_status fi ;; established) if [ "$ptcp_status" == "" ];then echo 0 else echo $ptcp_status fi ;; time_wait) if [ "$ptcp_status" == "" ];then echo 0 else echo $ptcp_status fi ;; syn_sent) if [ "$ptcp_status" == "" ];then echo 0 else echo $ptcp_status fi ;; syn_recv) if [ "$ptcp_status" == "" ];then echo 0 else echo $ptcp_status fi ;; closed) if [ "$ptcp_status" == "" ];then echo 0 else echo $ptcp_status fi ;; closing) if [ "$ptcp_status" == "" ];then echo 0 else echo $ptcp_status fi ;; close_wait) if [ "$ptcp_status" == "" ];then echo 0 else echo $ptcp_status fi ;; fin_wait1) if [ "$ptcp_status" == "" ];then echo 0 else echo $ptcp_status fi ;; fin_wait2) if [ "$ptcp_status" == "" ];then echo 0 else echo $ptcp_status fi ;; lastack) if [ "$ptcp_status" == "" ];then echo 0 else echo $ptcp_status fi ;; *) echo -e "\E[33mUsage: sh $0 [closed|closing|close_wait|syn_recv|syn_sent|fin_wait1|fin_wait2|listen|established|lastack|ti me_wait]\E[0m" esac
6、将zabbix相关的配置文件放到templates/目录下
# ls roles/configure/templates/ Userparameter_script.conf zabbix_agentd.con
>> Userparameter_script.conf配置文件，该配置文件定义zabbix监控项目键值与脚本路径
# cat Userparameter_script.conf
UserParameter=tcp.listen.port,sudo {{ zabbix_basedir }}/scripts/discovery_tcp_port.sh
UserParameter=tcp.connect.status[*],sudo {{ zabbix_basedir }}/scripts/tcp_connect_status.sh $1
>> zabbix_agentd.conf配置文件，该配置文件为zabbix客户端的主配置文件
Server={{ zabbix_server_ip }}
ServerActive={{ zabbix_server_ip }}:10051
Hostname={{ ansible_default_ipv4.address }}
Include={{ zabbix_basedir }}/etc/zabbix_agentd.conf.d
UnsafeUserParameters=1
7、执行zabbix_configure.yml同步zabbix配置
# ansible-playbook zabbix_configure.yml
[root@ansible /etc/ansible/zabbix_rhel/zabbix_agent ]# ansible-playbook zabbix_configure.yml PLAY [testhosts] *************************************************************** TASK [setup] ******************************************************************* ok: [10.17.83.33] ok: [10.17.83.34] TASK [configure : Install ntpdate software] ************************************ ok: [10.17.83.33] ok: [10.17.83.34] TASK [configure : Synchronization clock cron job] ****************************** ok: [10.17.83.33] ok: [10.17.83.34] TASK [configure : Restart crond] *********************************************** changed: [10.17.83.33] changed: [10.17.83.34] TASK [configure : Running ntpdate to synchronization time] ********************* changed: [10.17.83.33] changed: [10.17.83.34] TASK [configure : Add allow zabbix user to nopasswd exec sudo] ***************** ok: [10.17.83.33] ok: [10.17.83.34] TASK [configure : Add allow zabbix user to nopasswd exec sudo] ***************** ok: [10.17.83.33] ok: [10.17.83.34] TASK [configure : Create zabbix scripts directory] ***************************** ok: [10.17.83.33] ok: [10.17.83.34] TASK [configure : Copy zabbix monitor scripts] ********************************* ok: [10.17.83.33] ok: [10.17.83.34] TASK [configure : Copy zabbix_agentd.conf file] ******************************** changed: [10.17.83.33] changed: [10.17.83.34] TASK [configure : Copy Userparameter_script.conf file] ************************* ok: [10.17.83.33] ok: [10.17.83.34] RUNNING HANDLER [configure : restart zabbix_agentd] **************************** changed: [10.17.83.33] changed: [10.17.83.34] PLAY RECAP ********************************************************************* 10.17.83.33 : ok=12 changed=4 unreachable=0 failed=0 10.17.83.34 : ok=12 changed=4 unreachable=0 failed=0
8、查看同步情况
在ansible服务器上执行
>> 脚本同步
[root@ansible ~ ]# ansible testhosts -m shell -a "ls -l /usr/local/zabbix/scripts/" 10.17.83.33 | SUCCESS | rc=0 >> total 8 -rwxr-xr-x. 1 root root 837 Jul 5 17:56 discovery_tcp_port.sh -rwxr-xr-x. 1 root root 1937 Jul 6 18:07 tcp_connect_status.sh 10.17.83.34 | SUCCESS | rc=0 >> total 8 -rwxr-xr-x. 1 root root 837 Jul 6 16:33 discovery_tcp_port.sh -rwxr-xr-x. 1 root root 1803 Jul 5 17:21 tcp_connect_status.sh
>> 配置文件
[root@ansible ~ ]# ansible testhosts -m shell -a "cat /usr/local/zabbix/etc/zabbix_agentd.conf.d/Userparameter_script.conf" 10.17.83.34 | SUCCESS | rc=0 >> UserParameter=tcp.listen.port,sudo /usr/local/zabbix/scripts/discovery_tcp_port.sh UserParameter=tcp.connect.status[*],sudo /usr/local/zabbix/scripts/tcp_connect_status.sh $1 10.17.83.33 | SUCCESS | rc=0 >> UserParameter=tcp.listen.port,sudo /usr/local/zabbix/scripts/discovery_tcp_port.sh UserParameter=tcp.connect.status[*],sudo /usr/local/zabbix/scripts/tcp_connect_status.sh $1
>> 执行同步过去的监控脚本
配置zabbix监控模版
1、创建监控TCP连接状态模板
Configuration-->Templates--> Create template-->
Ansible Linux Templates-TCP connect status-->Items-->Create Item
2、创建自动发现TCP监听端口模板
Configuration-->Templates--> Create template-->
Ansible Linux Templates-TCP listen ports-->Discovery rules--> Create discovery rule-->Discovery rules -->Item prototypes --> Create Item prototype
配置Zabbix自动发现规则
Configuration-->Actions-->Create action(Discovery)
>> 创建一个名字为Ansible-testhosts的自动发现规则(Action)
>> 添加自动发现IP地址范围
>> 添加链接的主机、监控模版(之前创建的两个监控模版)
完成zabbix配置
在配置完上面所有步骤之后，zabbix服务端就会根据自动发现规则自动添加监控主机与监控项目
>> 查看监控数据
>> 查看监控数据
>> 查看监控数据
END
此后我们仍然可以继续使用ansible管理zabbix远程客户端主机的配置文件、脚本同步、监控项目的添加等。