ZABBIX冗余架构构筑（Centos6.4+pacemaker+corosync+drbd）

时间：2023-03-08 17:46:48
ZABBIX冗余架构构筑（Centos6.4+pacemaker+corosync+drbd）

基本构成：

  1. 用pacemaker+corosync控制心跳和资源迁移
  2. 用drbd同步zabbix配置文件和mysql数据库
  3. 所有软件都用yum安装至默认路径
  4. 主机的drbd领域挂载至/drbd，备机不挂载
  5. 主机在drbd领域中创建/drbd/zabbix，挂载至/etc/zabbix，备机不挂载
  6. 主机在drbd领域中创建/drbd/mysql，挂载至/var/lib/mysql，备机不挂载

    主机正常时如左图，主机故障时如右图

ZABBIX冗余架构构筑（Centos6.4+pacemaker+corosync+drbd）

IP构成：

 

ZAB01

ZAB02

VIP

192.168.0.8/24

主业务兼心跳IP

192.168.0.6/24

192.168.0.7/24

DRBD同步用IP

192.168.222.196/24

192.168.222.197/24

磁盘构成：

/dev/sda1

boot

0.5GB

/dev/sda2

lv_swap

8GB

lv_root

41.5GB

/dev/sda3

lv_drbd

450GB

*OS安装的时候只需要分出sda1和sda2，其他的空出来。等系统安装完毕后用fdisk分出sda3，并且先不要格式化。

正式作业前：

1，关闭selinux
2，关闭iptables
3，/etc/hosts 互相添加
4，NTP同步设定
5，在网卡配置文件中加入NOZEROCONF=yes，禁止一些不必要的路由信息自动生成
6，修改/etc/ssh/sshd_conf的 UseDNS参数为no，这样ssh登录的时候会快点
7，fdisk生成/dev/sda3，从sda3中取出剩余的450GB做成lv_drbd用作drbd领域，先不要格式化分区

正式作业：

1，drbd安装

安装elpl的repodata

rpm --import https://www.elrepo.org/RPM-GPG-KEY-elrepo.org

rpm -Uvh http://www.elrepo.org/elrepo-release-6-6.el6.elrepo.noarch.rpm

yum install drbd83-utils kmod-drbd83

这里要注意一点，kmod-drbd的版本要和内核版本匹配，不然drbd会报错无法启动。

但是我找遍了googlebaidu也没有找到列表，只能把我知道的写出来。

kmod-drbd83-8.3.13-1 对应kernel 2.6.32-358

kmod-drbd83-8.3.13-3 对应kernel 2.6.32- 573.22.1

关闭drbd的自启动

chkconfig drbd off

编辑配置文件

vi /etc/drbd.d/r0.res

resource r0 { 

protocol C; 

net { 

max-buffers ; 

max-epoch-size ; 

sndbuf-size ; 

} 

handlers { 

fence-peer "/usr/lib/drbd/crm-fence-peer.sh"; 

after-resync-target "/usr/lib/drbd/crm-unfence-peer.sh"; 

} 

syncer { 

rate 80M; 

use-rle; 

verify-alg md5; 

} 

disk{ 

fencing resource-only; 

on-io-error detach; 

} 

device /dev/drbd0; 

on ZAB01 { 

address ; 

disk /dev/vg_drbd/lv_drbd; 

meta-disk internal; 

} 

on ZAB02 { 

address ; 

disk /dev/vg_drbd/lv_drbd; 

meta-disk internal; 

} 

}

初始化盘

drbdadm create-md r0

如果报这个错

Device size would be truncated, which

would corrupt data and result in

'access beyond end of device' errors.

You need to either

* use external meta data (recommended)

* shrink that filesystem first

* zero out the device (destroy the filesystem)

Operation refused.

打这个命令后再执行一次初始化就好了

dd if=/dev/zero bs=1M count=1 of=/dev/vg_drbd/lv_drbd

两边服务器上启动drbd

/etc/init.d/drbd start

把ZAB01V这边升为主机

drbdadm -- --overwrite-data-of-peer primary r0

主机这边格式化

mkfs.ext4 /dev/drbd0

两边都做个文件夹，用来挂载drbd的盘

mkdir /drbd

主机这边挂载

mount /dev/drbd0 /drbd/

2，mysql安装

安装mysql的repodata

wget http://dev.mysql.com/get/mysql57-community-release-el6-8.noarch.rpm

rpm -ivh mysql57-community-release-el6-8.noarch.rpm

默认会安装的是mysql5.7，如果需要5.6或者5.5的话可以修改repo文件，我装的是5.6

vi /etc/yum.repos.d/mysql-community.repo

把对应版本的

[mysql56-community]下面的

enabled=0改成enabled=1

[mysql57-community]下面的

enabled=1改成enabled=0

安装mysql

yum install mysql-server mysql-devel

关闭mysql的自启动

chkconfig mysqld off

创建文件夹用来放DRBD用的mysql数据

mkdir /drbd/mysql

chmod 700 /drbd/mysql

chown mysql.mysql /drbd/mysql

以下只需要在主机上配置，备机反正是用drbd里的数据，只要保证mysql能启动就可以了。

编辑配置文件

按主机的性能自行修改，以下是我4核8G内存主机配置

vi /etc/my.cnf

[mysqld] 

datadir=/var/lib/mysql 

socket=/var/lib/mysql/mysql.sock 

character-set-server = utf8 

collation-server = utf8_bin 

skip-character-set-client-handshake 

skip-external-locking 

symbolic-links= 

innodb_buffer_pool_size = 2048M 

innodb_log_file_size = 512M 

sort_buffer_size = 2M 

innodb_additional_mem_pool_size = 30M 

innodb_log_buffer_size = 8M 

key_buffer_size = 16M 

innodb_data_file_path = ibdata1:1G 

innodb_file_per_table 

log-error=/var/log/mysqld.log 

pid-file=/var/run/mysqld/mysqld.pid

挂载

mount -o bind /drbd/mysql /var/lib/mysql/

启动mysql

service mysqld start

创建zabbix用的DB和账号

>create database zabbix;

>grant all privileges on zabbix.* to zabbix@localhost identified by 'YOURPWD' ;

>flush privileges;

>exit

3，zabbix安装配置

安装zabbix2.4.1的repodata

wget http://repo.zabbix.com/zabbix/2.4/rhel/6/x86_64/zabbix-release-2.4-

1.el6.noarch.rpm

rpm -ivh zabbix-release-2.4-1.el6.noarch.rpm

安装zabbix

yum install zabbix-server-mysql zabbix-web-mysql zabbix-get zabbix-agent

因为agent不需要冗余，设置开机启动

chkconfig zabbix-agent on

创建文件夹用来放DRBD用的zabbix配置文件

mkdir /drbd/zabbix

把http的配置也改了

vi /etc/httpd/conf.d/zabbix.conf

php_value date.timezone Asia/Shanghai

以下同样只需要在主机上配置，备机只要保证zabbix能启动就可以了。

导入zabbix的模板

cd /usr/share/doc/zabbix-server-mysql-2.4.7/create

mysql zabbix < schema.sql

mysql zabbix < images.sql

mysql zabbix < data.sql

配置zabbix_server.conf

vi /etc/zabbix/zabbix_server.conf

#必须的参数 

DBPassword=YOURPWD 

#以下参数是性能优化参数 

StartPollers= 

StartPollersUnreachable= 

StartTrappers= 

StartPingers= 

StartDiscoverers= 

MaxHousekeeperDelete= 

CacheSize=64M 

StartDBSyncers= 

HistoryCacheSize=128M 

ValueCacheSize=128M 

TrendCacheSize=4M 

HistoryTextCacheSize=16M

如果需要的话，可以把mysql的patitioning也做了。

参考我的另一篇博文

http://www.cnblogs.com/justbio/p/5379010.html

开启zabbix

service zabbix-server start

service zabbix-agent start

service httpd start

打开网页配置zabbix的页面

http://192.168.0.6/zabbix

配置完成后关闭zabbix

service zabbix-server stop

service zabbix-agent stop

service httpd stop

service mysqld stop

因为最终要同步的文件夹是/drbd，

所以把zabbix的配置文件拷贝到DRBD的文件夹中。

加上-p是为了保持权限不变

cp -rp /etc/zabbix/* /drbd/zabbix/

写个文件到/drbd/mysql和/drbd/zabbix里面去

touch /drbd/zabbix/1.txt

挂载看看有没有问题

mount -o bind /drbd/zabbix/ /etc/zabbix/

ls /etc/zabbix/

1.txt web zabbix_agentd.conf zabbix_agentd.d zabbix_server.conf

里面都能看到有1.txt

记住umount掉

umount /var/lib/mysql/

umount /etc/zabbix/

关闭drbd

/etc/init.d/drbd stop

4，pacemaker安装配置

首先确认没有进程在启动着（zabbix-agent没关系）

ps -ef |grep zabbix |grep -v grep

ps -ef |grep drbd |grep -v grep

ps -ef |grep http |grep -v grep

ps -ef |grep mysql |grep -v grep

安装pacemaker的源

因为clusterlabs.org的pacemaker源默认安装的是cman，我还是喜欢用corosync和heartbeat

，

所以用的是linux-ha.jp的源，墙内下载源可能要FQ。

wget https://osdn.jp/frs/redir.php?m=iij&f=%2Flinux-ha%2F63919%2Fpacemaker-repo-

1.1.13-1.1.el6.x86_64.rpm

rpm -ivh pacemaker-repo-1.1.13-1.1.el6.x86_64.rpm

安装pacemaker

yum install pacemaker corosync crmsh glib2

配置corosync

vi /etc/corosync/corosync.conf

service { 

name: pacemaker 

ver:  

use_mgmtd: yes 

} 

totem { 

version:  

token:  

consensus:  

rrp_problem_count_timeout:  

rrp_mode: active 

interface { 

member { 

memberaddr: 192.168.222.196 

} 

member { 

memberaddr: 192.168.222.197 

} 

ringnumber:  

bindnetaddr: 192.168.222.0 

mcastport:  

ttl:  

} 

transport: udpu 

} 

logging { 

fileline: off 

to_stderr: no 

to_logfile: yes 

logfile: /var/log/cluster/corosync.log 

to_syslog: yes 

debug: off 

timestamp: on 

logger_subsys { 

subsys: QUORUM 

debug: off 

} 

} 

quorum { 

provider: corosync_votequorum 

expected_votes:  

}

启动pacemaker

service corosync start

service pacemaker start

编辑crm配置文件

crm

>configure

>edit

node : ZAB01 

node : ZAB02 

primitive res_drbd_r0 ocf:linbit:drbd \ 

params drbd_resource=r0 \ 

op start interval= timeout= \ 

op stop interval= timeout= 

primitive res_httpd apache \ 

params configfile="/etc/httpd/conf/httpd.conf" \ 

op start interval= timeout= \ 

op stop interval= timeout= \ 

op monitor interval= timeout= 

primitive res_ip IPaddr2 \ 

 \ 

op monitor interval= timeout= 

primitive res_mount Filesystem \ 

params device="/dev/drbd0" fstype=ext4 directory="/drbd" options=noatime \ 

op monitor interval= timeout= \ 

op start interval= timeout= \ 

op stop interval= timeout= 

primitive res_mount_mysql Filesystem \ 

params device="/drbd/mysql" fstype=ext4 directory="/var/lib/mysql" 

options="bind,noatime" \ 

op monitor interval= timeout= \ 

op start interval= timeout= \ 

    op stop interval= timeout= 

primitive res_mount_zabbix Filesystem \ 

params device="/drbd/zabbix" fstype=ext4 directory="/etc/zabbix" options="bind,noatime" \ 

op monitor interval= timeout= \ 

op start interval= timeout= \ 

op stop interval= timeout= 

primitive res_mysqld mysql \ 

op start interval= timeout= \ 

op stop interval= timeout= \ 

op monitor interval= timeout= 

primitive res_srcaddr IPsrcaddr \ 

params ipaddress=192.168.0.8 \ 

op monitor interval= timeout= 

primitive res_zabbix_server lsb:zabbix-server \ 

op start interval= timeout= \ 

op stop interval= timeout= \ 

op monitor interval= timeout= 

group zabbix res_mount res_mount_mysql res_mount_zabbix res_ip res_srcaddr \     

res_httpd res_mysqld res_zabbix_server 

ms ms_drbd_r0 res_drbd_r0 \ 

meta master-max= master-node-max= \ 

clone-max= clone-node-max= notify=true 

location l_zabbix zabbix : VCCZAB01V 

colocation c_zabbix inf: zabbix ms_drbd_r0:Master 

order o_zabbix inf: ms_drbd_r0:promote zabbix:start 

property cib-bootstrap-options: \ 

have-watchdog=false \ 

dc-version=-6052cd1 \ 

cluster-infrastructure=corosync \ 

 \ 

no-quorum-policy=ignore \ 

stonith-enabled=false

配置完成后

>commit

退出crm后打crm_mon查看

crm_mon

…… 

Online: [ ZAB01 ZAB02 ] 

Master/Slave Set: ms_drbd_r0 [res_drbd_r0] 

Masters: [ ZAB01 ] 

Slaves: [ ZAB02 ] 

Resource Group: zabbix 

res_mount (ocf::heartbeat:Filesystem): Started ZAB01 

res_mount_mysql (ocf::heartbeat:Filesystem): Started ZAB01 

res_mount_zabbix (ocf::heartbeat:Filesystem): Started ZAB01 

res_ip (ocf::heartbeat:IPaddr2): Started ZAB01 

res_srcaddr (ocf::heartbeat:IPsrcaddr): Started ZAB01 

res_httpd (ocf::heartbeat:apache): Started ZAB01 

res_mysqld (ocf::heartbeat:mysql): Started ZAB01 

res_zabbix_server (lsb:zabbix-server): Started ZAB01

查看drbd状态

service drbd status

刚开始会显示同步的进度条，同步完成后，两边的状态分别如下

主机上

drbd driver loaded OK; device status: 

version:  (api:/proto:-) 

GIT-hash: a798fa7e274428a357657fb52f0ecf40192c1985 build by phil@Build64R6, -- :: 

m:res cs ro ds p mounted fstype 

:r0 Connected Primary/Secondary UpToDate/UpToDate C /drbd ext4 

:r0 Connected Primary/Secondary UpToDate/UpToDate C /etc/zabbix ext4 

:r0 Connected Primary/Secondary UpToDate/UpToDate C /var/lib/mysql ext4

备机上

drbd driver loaded OK; device status: 

version:  (api:/proto:-) 

GIT-hash: 83ca112086600faacab2f157bc5a9324f7bd7f77 build by dag@Build64R6, -- :: 

m:res cs ro ds p mounted fstype 

:r0 Connected Secondary/Primary UpToDate/UpToDate C

系统构筑完成

测试：

关闭主机的pacemaker和corosync

service pacemaker stop

service corosync stop

然后在备机上查看状态

crm_mon

…… 

Online: [ZAB02 ] 

Offline: [ZAB01 ] 

Master/Slave Set: ms_drbd_r0 [res_drbd_r0] 

Masters: [ ZAB02 ] 

Resource Group: zabbix 

res_mount (ocf::heartbeat:Filesystem): Started ZAB02 

res_mount_mysql (ocf::heartbeat:Filesystem): Started ZAB02 

res_mount_zabbix (ocf::heartbeat:Filesystem): Started ZAB02 

res_ip (ocf::heartbeat:IPaddr2): Started ZAB02 

res_srcaddr (ocf::heartbeat:IPsrcaddr): Started ZAB02 

res_httpd (ocf::heartbeat:apache): Started ZAB02 

res_mysqld (ocf::heartbeat:mysql): Started ZAB02 

res_zabbix_server (lsb:zabbix-server): Started ZAB02

service drbd status

drbd driver loaded OK; device status: 

version:  (api:/proto:-) 

GIT-hash: a798fa7e274428a357657fb52f0ecf40192c1985 build by phil@Build64R6, -- :: 

m:res cs ro ds p mounted fstype 

:r0 Connected Primary/Unknow UpToDate/DUnknown C /drbd ext4 

:r0 Connected Primary/Unknow UpToDate/DUnknown C /etc/zabbix ext4 

:r0 Connected Primary/Unknow UpToDate/DUnknown C /var/lib/mysql ext4

自动电话，自动微信，自动邮件报警可以参考我的其他博文

自动电话报障 ： http://www.cnblogs.com/justbio/p/5378777.html

自动微信报障 ： http://www.cnblogs.com/justbio/p/5360160.html

自动邮件报障 ： http://www.cnblogs.com/justbio/p/5359918.html