zabbix安装

安装服务器端包，其中mysql，web可分开为单独的服务器
[root@node9 zabbix]# yum -y install zabbix-server-2.0.8-3.el6.x86_64.rpm zabbix-2.0.8-3.el6.x86_64.rpm zabbix-get-2.0.8-3.el6.x86_64.rpm zabbix-server-mysql-2.0.8-3.el6.x86_64.rpm zabbix-web-mysql-2.0.8-3.el6.noarch.rpm zabbix-web-2.0.8-3.el6.noarch.rpm php-bcmath-5.3.3-26.el6.x86_64.rpm php-mbstring-5.3.3-26.el6.x86_64.rpm

安装mysql
[root@node9 zabbix]# yum -y install mysql-server mysql
[root@node9 zabbix]# service mysqld start
mysql创建数据库和授权
mysql> create database zabbix;
mysql> grant all on zabbix.* to zabbix@127.0.0.1 identified by 'redhat';
mysql> flush privileges;

把zabbix-server-mysql生成的数据库导入msyql的zabbix库
[root@node9 create]# mysql zabbix < schema.sql
[root@node9 create]# mysql zabbix < p_w_picpaths.sql
[root@node9 create]# mysql zabbix < data.sql
[root@node9 create]# pwd
/usr/share/doc/zabbix-server-mysql-2.0.8/create

编辑配置文件，修改DBname,DBpass,DBuser
[root@node9 zabbix]# vim zabbix_server.conf
DBUser=zabbix
DBPassword=redhat
DBName=zabbix
DBHost=127.0.0.1
改php配置文件
date.timezone = Asia/Shanghai

启动服务
[root@node9 zabbix]# service httpd start
[root@node9 zabbix]# service zabbix-server start

在浏览器访问
http://192.168.100.9/zabbix

配置连接数据库，test ok,才能下一步

登录，默认用户密码 admin zabbix

启用中文支持，建议在server上安装中文的包，不然有的地方有？？？这样的

配置被监控端（agent）
安装包
[root@node2 tmp]# yum -y install zabbix-2.0.8-3.el6.x86_64.rpm zabbix-agent-2.0.8-3.el6.x86_64.rpm zabbix-sender-2.0.8-3.el6.x86_64.rpm

编辑配置文件
# vim zabbix_agentd.conf
Server=192.168.100.25
ServerActive=192.168.100.25

启动服务
# service zabbix-agent start

------------------
配置监控mysql主从复制

1,授权zabbix-agent帐号
mysql> grant replication client on *.* to zabbix@'127.0.0.1' identified by 'redhat';
flush privileges;
2，修改/etc/zabbix/.my.cnf
[mysql]
user = zabbix
host = localhsot
password = redhat
socket = /var/lib/mysql/mysql.sock

[mysqladmin]
user = zabbix
host = localhsot
password = redhat
socket = /var/lib/mysql/mysql.sock

3,修改/etc/zabbix/zabbix_agent.d/userparameter_mysql.conf
把HOME=/var/lib/zabbix改为HOME=/etc/zabbix

4，写一个userparameter
UserParameter=mysql.repli,echo "show slave status\G"|HOME=/etc/zabbix mysql|awk '/Slave/{print }'|grep -c 'Yes'

5，重启服务
service zabbix-agent restart



6，定义项目

7，定义触发器

8，定义一个示警媒介

9，定义用户
要定义一个受邮件的地址，不然会收不到信息



10，定义动作/条件/操作
动作：相当于邮件内容
条件：什么时候，比如要不再维护期
操作：从第几步到第几步给谁发信息

主要定义报警升级的

主要是邮件内容是什么

这个基本不用改，表示只要有问题就报警

测试

我们把mysql的io_thread停掉

mysql> stop slave io_thread;

Query OK, 0 rows affected (0.00 sec)

查看状态

我试了好几回所以状态比较多，可以看到正在作业中，就是在发邮件

看是否邮件

U 3 [e-mail] Thu Jun 12 00:14 23/1010 "PROBLEM: mysql replication fail"

U 4 [e-mail] Thu Jun 12 00:15 23/1001 "OK: mysql replication fail"

U 5 [e-mail] Thu Jun 12 00:15 23/1010 "PROBLEM: mysql replication fail"

Message 5:

From [e-mail] Wed Jun 11 23:49:51 2014

Return-Path: <[e-mail]>

X-Original-To: root@localhost

Delivered-To: [e-mail]

From: <[e-mail]>

To: <[e-mail]>

Date: Wed, 11 Jun 2014 23:49:51 +0800

Subject: PROBLEM: mysql replication fail

Content-Type: text/plain; charset="UTF-8"

Status: RO

----------------------------

其他的条目可以自己写，也可以用默认的挺好的

定义多个screen，两行两列，表示能显示四幅图

11，定义发现

先去配置一个新的agent
[root@node3 ~]# ls
anaconda-ks.cfg install.log.syslog zabbix-agent-2.0.8-3.el6.x86_64.rpm
install.log zabbix-2.0.8-3.el6.x86_64.rpm zabbix-sender-2.0.8-3.el6.x86_64.rpm
[root@node3 ~]# yum -y install *.rpm

# vim zabbix_agentd.conf
Server=192.168.100.25
ServerActive=192.168.100.25

[root@node3 ~]# ss -anl|grep 10050
LISTEN 0 128 :::10050 :::*
LISTEN 0 128 *:10050 *:*

先创建一个发现

 

定义一个发现动作

---------------------

web监控

配置一台web

yum -y install httpd

echo 'ok' >/var/www/html/test.html

service httpd start

1，新建一个应用合集

2，创建web方案

3，查看

4，为web监控创建触发器，创建web后会自动创建很多触发器

我们来保存测试一下

[root@localhost ~]# ab -c 100 -n 100000 http://192.168.100.103/web.html

[root@localhost ~]#mail

U 19 [e-mail] Thu Jun 12 03:33 25/1131 "PROBLEM: web response time to long"

U 20 [e-mail] Thu Jun 12 03:33 24/1117 "OK: web response time to long"

U 21 [e-mail] Thu Jun 12 03:33 24/1134 "PROBLEM: Processor load is too high on 192.168.100.103"

U 22 [e-mail] Thu Jun 12 03:33 24/1117 "OK: web response time to long"

>N 23 [e-mail] Thu Jun 12 03:34 23/1124 "PROBLEM: Processor load is too high on 192.168.100.103"

N 24 [e-mail] Thu Jun 12 03:35 23/1124 "PROBLEM: Processor load is too high on 192.168.100.103"

N 25 [e-mail] Thu Jun 12 03:36 23/1111 "OK: Processor load is too high on 192.168.100.103"

N 26 [e-mail] Thu Jun 12 03:39 23/1075 "PROBLEM: Disk I/O is overloaded on node102.xy.com"

N 27 [e-mail] Thu Jun 12 03:40 23/1075 "PROBLEM: Disk I/O is overloaded on node102.xy.com"

N 28 [e-mail] Thu Jun 12 03:40 23/1062 "OK: Disk I/O is overloaded on node102.xy.com"

因为用的虚拟机，所以全都报警了

 

本文出自 ““小雨” 博客，请务必保留此出处http://evildexy.blog.51cto.com/8607019/1425249