Keepalived + Haproxy 实现高可用负载均衡

Posted 2020-08-08

keepalived是一个用于做双机热备（HA）的软件，常和haproxy联合起来做热备＋负载均衡，达到高可用

1. 实验环境

Master:10.20.2.233

Backup:10.20.2.235

虚拟IP:10.20.2.238

2. Keepalived安装

1) 通过yum安装软件依赖包

yum install -y wget make gcc openssl-devel popt-devel

2) Keepalived源码包下载安装

下载地址：http://www.keepalived.org/software/keepalived-1.2.23.tar.gz

tar -zxf keepalived-1.2.23.tar.gz -C /usr/src
cd /sru/src/keepalived-1.2.23/
./configure --prefix=/usr/local/keepalived/
Make
make install

3) 安装成功后做成服务模式

cp /usr/local/keepalived/sbin/keepalived /usr/sbin/
cp /usr/local/keepalived/etc/sysconfig/keepalived /etc/sysconfig/
cp /usr/local/keepalived/etc/rc.d/init.d/keepalived /etc/init.d/

3. 编辑配置文件

mkdir -p /etc/keepalived
cp /usr/local/keepalived/etc/keepalived/keepalived.conf /etc/keepalived/

 

1) Master配置文件内容如下：

! Configuration File for keepalived
 
global_defs {
   router_id HA_TEST
}
 
#监测haproxy进程状态，每2秒执行一次
vrrp_script chk_haproxy {
    script "/etc/keepalived/check_haproxy.sh"
    interval 2
    weight 2
}
 
vrrp_instance VI_1 {
    state MASTER # 从为BACKUP
    priority 200 # 从为100
    virtual_router_id 51 #路由ID，可通过#tcpdump vrrp查看。
    garp_master_delay 1 #主从切换时间，单位为秒。
    interface eth0
    advert_int 1
    mcast_src_ip 10.20.2.233    #MASTER服务器IP,从服务器写从服务器的IP
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {
        10.20.2.238    #虚拟IP
    }
    track_script {
        chk_haproxy    #监测haproxy进程状态
    }
}

2) BACKUP配置文件内容如下：

! Configuration File for keepalived
 
global_defs {
   router_id HA_TEST
}
 
#监测haproxy进程状态，每2秒执行一次
vrrp_script chk_haproxy {
    script "/etc/keepalived/check_haproxy.sh"
    interval 2
    weight 2
}
 
vrrp_instance VI_1 {
    state BACKUP # 主为MASTER
    priority 100 # 主为101
    virtual_router_id 51 #路由ID，可通过#tcpdump vrrp查看。
    garp_master_delay 1 #主从切换时间，单位为秒。
    interface eth0
    advert_int 1
    mcast_src_ip 10.20.2.235    #BACKUP服务器IP,主服务器写主服务器的IP
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {
        10.20.2.238    #虚拟IP
    }
    track_script {
        chk_haproxy    #监测haproxy进程状态
    }
}

3) haproxy检测脚本（check_haproxy.sh）内容如下

#!/bin/bash  
A=`ps -C haproxy --no-header |wc -l`  
if [ $A -eq 0 ];then  
/etc/init.d/haproxyd start
sleep 3  
if [ `ps -C haproxy --no-header |wc -l` -eq 0 ];then  
/etc/init.d/keepalived stop  
fi  
Fi

chmod 755 /etc/keepalived/check_haproxy.sh   #脚本编写完成后需要赋上权限

4) Keepalived的启停

service keepalived start     #停止
service keepalived stop      #启动
service keepalived restart    #重启

4. 测试：

1) 启动MASTER和BACKUP上的keepalived

service keepalived start （keepalived启动的同时会将haproxy启动）
[[email protected] keepalived]# service keepalived start
Starting keepalived:                                       [  OK  ]
[[email protected] keepalived]# service haproxyd status
HAProxy(pid 32077) is running!

MASTER绑定了VIP,BACKUP未绑定VIP

[[email protected] keepalived]# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN 
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
    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 qlen 1000
    link/ether 00:50:56:85:46:b2 brd ff:ff:ff:ff:ff:ff
    inet 10.20.2.233/24 brd 10.20.2.255 scope global eth0
    inet 10.20.2.238/32 scope global eth0
    inet6 fe80::250:56ff:fe85:46b2/64 scope link 
       valid_lft forever preferred_lft forever
[[email protected] keepalived]# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN 
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
    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 qlen 1000
    link/ether 00:50:56:85:73:ce brd ff:ff:ff:ff:ff:ff
    inet 10.20.2.235/24 brd 10.20.2.255 scope global eth0
    inet6 fe80::250:56ff:fe85:73ce/64 scope link 
       valid_lft forever preferred_lft forever

2) 停止主上的haproxy，3秒后，keepalived会将其启动

[[email protected] keepalived]# service haproxyd stop
STOP HAProxy Listen
STOP HAProxy process
[[email protected] keepalived]# service haproxyd status
HAProxy(pid 32368) is running!

3) 停止主上的keepalived，备机马上接管服务

BACKUP上绑定了VIP

[[email protected] keepalived]# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN 
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
    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 qlen 1000
    link/ether 00:50:56:85:73:ce brd ff:ff:ff:ff:ff:ff
    inet 10.20.2.235/24 brd 10.20.2.255 scope global eth0
    inet 10.20.2.238/32 scope global eth0
    inet6 fe80::250:56ff:fe85:73ce/64 scope link 
       valid_lft forever preferred_lft forever

4) 启动主上的keepalived，MASTER绑定VIP，BACKUP释放了VIP

[[email protected] keepalived]# service keepalived start
Starting keepalived:                                       [  OK  ]
[[email protected] keepalived]# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN 
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
    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 qlen 1000
    link/ether 00:50:56:85:73:ce brd ff:ff:ff:ff:ff:ff
    inet 10.20.2.235/24 brd 10.20.2.255 scope global eth0
    inet 10.20.2.238/32 scope global eth0
    inet6 fe80::250:56ff:fe85:73ce/64 scope link 
       valid_lft forever preferred_lft forever

 

本文出自 “11528244” 博客，请务必保留此出处http://11538244.blog.51cto.com/11528244/1851606