第三节：安装Keepalived 实现自动主备切换

Posted 2021-02-18 nanxiang

 

 

 

 网上没找到合适的图片，自己画一张简单的架构图吧

想单独讲一个keepalived是因为不止可以实现mysql主备的切换，包括nginx、proxysql等需要高可用架构的场景，都可以实现。

 

这是部署完成之后的架构图，服务器A会多一个虚拟IP地址。192.168.150.100

应用程序都会访问这个虚拟IP，

同时两台服务器上的keepalived软件会保持心跳，检测对方是否存在。

下面是当服务器A或MySQL 实例A挂跳之后的架构。可以看到，虚拟IP 192.168.150.100 已经飘移到了服务器B，这对应用侧是透明的，应用侧不需要修改数据库连接地址，还能正常访问到数据库

如果服务器A 或MySQL实例A修复好之后，虚拟IP还用飘移到服务器A吗，如果没有特殊需求，是不用的，毕竟MySQL实例A和MySQL实例B数据是一样的。

 

 

 

 

 

 

两台机器都要安装

下载地址：

https://www.keepalived.org/download.html

Keepalived for Linux - Version 2.0.20 

 

wget https://www.keepalived.org/software/keepalived-2.0.20.tar.gz

tar xvf keepalived-2.0.20.tar.gz

cd keepalived-2.0.20/

yum install gcc openssl-devel libnl3-devel pcre-devel -y
./configure --prefix=/usr/local/keepalived
make && make install



mkdir /etc/keepalived/
cp /usr/local/keepalived/etc/keepalived/keepalived.conf /etc/keepalived/
cp /usr/local/keepalived/etc/sysconfig/keepalived /etc/sysconfig/
cp /root/keepalived-2.0.20/keepalived/keepalived.service /etc/systemd/system/
ln -s /usr/local/keepalived/sbin/keepalived /usr/sbin/
ll /usr/sbin/k*
cp /root/keepalived-2.0.20/keepalived/etc/init.d/keepalived /etc/init.d/
chmod 755 /etc/init.d/keepalived
systemctl enable keepalived.service

systemctl start keepalived.service

[root@localhost ~]# ps -ef|grep keep
gdm 8380 8099 0 06:45 ? 00:00:00 /usr/libexec/gsd-housekeeping
root 13572 1 0 07:33 ? 00:00:00 /usr/local/keepalived/sbin/keepalived -D
root 13573 13572 0 07:33 ? 00:00:00 /usr/local/keepalived/sbin/keepalived -D
root 13574 13572 0 07:33 ? 00:00:00 /usr/local/keepalived/sbin/keepalived -D
root 13593 8592 0 07:33 pts/0 00:00:00 grep --color=auto keep
[root@localhost ~]# tail -f /var/log/messages 
May 18 07:33:35 localhost Keepalived_vrrp[13574]: Sending gratuitous ARP on eth0 for 192.168.200.18
May 18 07:33:36 localhost Keepalived_healthcheckers[13573]: HTTP_CHECK on service [192.168.201.100]:tcp:443 failed after 3 retry.
May 18 07:33:36 localhost Keepalived_healthcheckers[13573]: Removing service [192.168.201.100]:tcp:443 to VS [192.168.200.100]:tcp:443
May 18 07:33:36 localhost Keepalived_healthcheckers[13573]: Lost quorum 1-0=1 > 0 for VS [192.168.200.100]:tcp:443
May 18 07:33:36 localhost Keepalived_healthcheckers[13573]: Remote SMTP server [192.168.200.1]:25 connected.
May 18 07:34:00 localhost Keepalived_healthcheckers[13573]: Timeout reading data to remote SMTP server [192.168.200.1]:25.
May 18 07:34:01 localhost Keepalived_healthcheckers[13573]: Timeout reading data to remote SMTP server [192.168.200.1]:25.
May 18 07:34:03 localhost Keepalived_healthcheckers[13573]: Timeout reading data to remote SMTP server [192.168.200.1]:25.
May 18 07:34:04 localhost Keepalived_healthcheckers[13573]: Timeout reading data to remote SMTP server [192.168.200.1]:25.
May 18 07:34:06 localhost Keepalived_healthcheckers[13573]: Timeout reading data to remote SMTP server [192.168.200.1]:25.
#Keepalived已经正常启动了，现在先停掉服务
[root@localhost ~]# systemctl stop keepalived.service

[root@localhost ~]# service mysql stop

 

 

继续在两台服务器做相同配置

mv /etc/keepalived/keepalived.conf /etc/keepalived/keepalived.conf.bak
vi /etc/keepalived/keepalived.conf



global_defs {
router_id MySQL-HA   #标识，只是一个名字
}

vrrp_script check_run {
script "/usr/local/mysql/mysql_check.sh"   #检测脚本地址
interval 60　　　　　　　　　　　　　　　　　　　　#健康检查周期
}


vrrp_instance VI_1 {
state BACKUP　　　　　　　　　　　　　　　　　　#状态为BACKUP
interface eth0　　　　　　　　　　　　　　　　　　#绑定的网卡
virtual_router_id 50　　　　　　　　　　　　　　#主备服务器必须在一个id内，
priority 100　　　　　　　　　　　　　　　　　　#权重设置
advert_int 1　　　　　　　　　　　　　　　　　　　　#心跳间隔
nopreempt　　　　　　　　　　　　　　　　　　　　#非抢占模式
authentication {
auth_type PASS　　　　　　　　　　　　　　　　　　#用户名 主备要配置为一样
auth_pass 1234　　　　　　　　　　　　　　　　　　#密码 主备要配置为一样
}

track_script {
check_run　　　　　　　　　　　　　　　　　　　　#执行的检测脚本
}

virtual_ipaddress {
192.168.150.100　　　　　　　　　　　　　　　　　　#虚拟IP
}
}

 

 

 

 

 

[root@localhost ~]#  vi /usr/local/mysql/mysql_check.sh 



#! /bin/bash

host=127.0.0.1
user=root
passwd=123456

/usr/local/mysql/bin/mysql -uroot -p123456 -h127.0.0.1 -P 3306 -e "select 1" >/dev/null 2>/dev/nul

if [ $? != 0 ] ; then
systemctl stop keepalived.service
fi



[root@localhost ~]#  chmod +x /usr/local/mysql/mysql_check.sh

 

 

配置完毕，

1、启动两个实例的mysql

[root@localhost ~]# service mysql start
Redirecting to /bin/systemctl start mysql.service
[root@localhost ~]# 

2、启动两个实例的keepalived

[root@localhost ~]# service keepalived start
Starting keepalived (via systemctl):                       [  OK  ]
[root@localhost ~]# 

3、先启动keepalived的服务器会先拿到虚拟IP，服务器A先启动的，在服务器A看下

[root@localhost ~]# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host 
       valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
    link/ether 00:0c:29:cd:3f:f3 brd ff:ff:ff:ff:ff:ff
    inet 192.168.150.101/24 brd 192.168.150.255 scope global noprefixroute eth0
       valid_lft forever preferred_lft forever
    inet 192.168.150.100/32 scope global eth0        ###可以看到已经有192.168.150.100这个IP了
       valid_lft forever preferred_lft forever
    inet6 fe80::9aea:2ec6:c21c:9e85/64 scope link tentative noprefixroute dadfailed 
       valid_lft forever preferred_lft forever
    inet6 fe80::6b3:ad64:3e31:9979/64 scope link noprefixroute 
       valid_lft forever preferred_lft forever
    inet6 fe80::e204:ab19:4112:354e/64 scope link noprefixroute 
       valid_lft forever preferred_lft forever
3: virbr0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue state DOWN group default qlen 1000
    link/ether 52:54:00:4a:de:09 brd ff:ff:ff:ff:ff:ff
    inet 192.168.122.1/24 brd 192.168.122.255 scope global virbr0
       valid_lft forever preferred_lft forever
4: virbr0-nic: <BROADCAST,MULTICAST> mtu 1500 qdisc pfifo_fast master virbr0 state DOWN group default qlen 1000
    link/ether 52:54:00:4a:de:09 brd ff:ff:ff:ff:ff:ff
[root@localhost ~]# 

 

模拟一下服务器A数据库宕机的场景，连接服务器A，执行

[root@host101 ~]# service mysql stop
Redirecting to /bin/systemctl stop mysql.service

keepalived的检测脚本会发现，mysql已经无法访问了，会触执行一条命令：

systemctl stop keepalived.service

 也就是将机的keepalived程序停掉，

1、服务器A的keepalived程序停掉后，无法与服务器B的keepalived程序保持心跳状态，同时也会删掉本机的虚拟IP

2、服务器B与服务器A的心跳中断之后，会认为服务器A已经宕机了，所以虚拟IP需要由自己来接管。

3、在服务器B执行，ip a  会发现IP已经在服务器B上边了

[root@host102 ~]# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host 
       valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
    link/ether 00:0c:29:80:76:e0 brd ff:ff:ff:ff:ff:ff
    inet 192.168.150.102/24 brd 192.168.150.255 scope global noprefixroute eth0
       valid_lft forever preferred_lft forever
    inet 192.168.150.100/32 scope global eth0    ###192.168.150.100已经在服务器B上了
       valid_lft forever preferred_lft forever
    inet6 fe80::9aea:2ec6:c21c:9e85/64 scope link noprefixroute 
       valid_lft forever preferred_lft forever
3: virbr0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc noqueue state DOWN group default qlen 1000
    link/ether 52:54:00:4a:de:09 brd ff:ff:ff:ff:ff:ff
    inet 192.168.122.1/24 brd 192.168.122.255 scope global virbr0
       valid_lft forever preferred_lft forever
4: virbr0-nic: <BROADCAST,MULTICAST> mtu 1500 qdisc pfifo_fast master virbr0 state DOWN group default qlen 1000
    link/ether 52:54:00:4a:de:09 brd ff:ff:ff:ff:ff:ff
[root@host102 ~]# 

如果此时在服务器A执行

service mysql start

service keepalived start
虚拟IP并不会再抢回来，根据之前keepalived.conf配置文件中，

state BACKUP
不会发生抢虚拟IP的行为，只有感知到另一方挂掉之后才会接管虚拟IP。

 如果服务器B的mysql进程挂掉后，192.168.150.100会瓢移到服务器A上边来。