标签：haproxy双主
1、 使用LVS负载均衡用户请求到后端web服务器，并且实现健康状态检查
2、 使用keepalived高可用LVS，避免LVS单点故障
3、 集群中分别在LK-01和LK-02运行一个VIP地址，实现LVS双主
4、 用户通过DNS轮训的方式实现访问集群的负载均衡（不演示）
环境介绍:
IP地址 | 功能描述 | |
HK-01 | 172.16.4.100 | 调度用户请求到后端web服务器，并且和LK-02互为备份 |
HK-02 | 172.16.4.101 | 调度用户请求到后端web服务器，并且和LK-01互为备份 |
WEB-01 | 172.16.4.102 | 提供web服务 |
WEB-02 | 172.16.4.103 | 提供web服务 |
DNS | 172.16.4.10 | 实现DNS轮训解析用户请求域名地址 |
VIP1 | 172.16.4.1 | 用户访问集群的入口地址，可能在LK-01，也可能在LK-02 |
VIP2 | 172.16.4.2 | 用户访问集群的入口地址，可能在LK-01，也可能在LK-02 |
Web服务器的配置极其简单，只需要提供测试页面启动web服务即可，配置如下：
Web-01配置
[root@WEB-01 ~]# echo "web-01" >/var/www/html/index.html [root@WEB-01 ~]# service httpd start Web-02配置 [root@WEB-02 ~]# echo "web-02" >/var/www/html/index.html [root@WEB-02 ~]# service httpd start
LVS访问后端web服务器，验证web服务提供成功
[root@HK-01 ~]# curl 172.16.4.102 web-01 [root@HK-01 ~]# curl 172.16.4.103 web-02
出现设置的页面，就说明web服务是正常
两个HK节点都安装haproxy和keepalived
[root@HK-01 ~]# yum -y install haproxy [root@HK-02 ~]# yum -y install haproxy [root@HK-01 ~]# yum -y install keepalived [root@HK-02 ~]# yum -y install keepalived
修改内核参数设置，设置haproxy启动的时候不管有没有vip地址都可以启动
[root@HK-01 ~]# echo"net.ipv4.ip_nonlocal_bind = 1" >> /etc/sysctl.conf [root@HK-01 ~]# sysctl –p [root@HK-02 ~]# echo "net.ipv4.ip_nonlocal_bind= 1" >> /etc/sysctl.conf [root@HK-02 ~]# sysctl -p
两个haproxy节点的配置文件一模一样，所以只放出一个
[root@HK-01 ~]# vim /etc/haproxy/haproxy.cfg global log 127.0.0.1 local2 chroot /var/lib/haproxy pidfile /var/run/haproxy.pid maxconn 4000 user haproxy group haproxy daemon statssocket /var/lib/haproxy/stats defaults mode http log global option httplog option dontlognull optionhttp-server-close optionforwardfor except 127.0.0.0/8 option redispatch retries 3 timeouthttp-request 10s timeoutqueue 1m timeoutconnect 10s timeoutclient 1m timeoutserver 1m timeouthttp-keep-alive 10s timeoutcheck 10s maxconn 3000 statsenable statsuri /admin?stats statsauth proxy:proxy listen www1 bind172.16.4.1:80 modehttp option forwardfor server www01 172.16.4.102:80 check server www02 172.16.4.103:80 check listen www2 bind172.16.4.2:80 modetcp option forwardfor server www01 172.16.4.102:80 check server www01 172.16.4.103:80 check
HK-01配置
[root@HK-01 ~]# vim /etc/keepalived/keepalived.conf
global_defs {
router_idLVS_DEVEL
}
vrrp_script chk_mt_down {
script"[[ -f /etc/keepalived/down ]] && exit 1 || exit 0"
interval 1
weight -5
}
vrrp_instance VI_1 {
stateMASTER
interfaceeth0
virtual_router_id 51
priority100
advert_int 1
authentication {
auth_type PASS
auth_pass asdfgh
}
virtual_ipaddress {
172.16.4.1/32 brd 172.16.4.1 dev eth0 label eth0:0
}
track_script {
chk_mt_down
}
}
vrrp_instance VI_2 {
stateBACKUP
interfaceeth0
virtual_router_id 52
priority99
advert_int 1
authentication {
auth_type PASS
auth_pass qwerty
}
virtual_ipaddress {
172.16.4.2
}
track_script {
chk_mt_down
}
}HK-02配置
[root@HK-02 ~]# vim /etc/keepalived/keepalived.conf
global_defs {
router_idLVS_DEVEL
}
vrrp_script chk_mt_down {
script"[[ -f /etc/keepalived/down ]] && exit 1 || exit 0"
interval 1
weight -5
}
vrrp_instance VI_1 {
stateBACKUP
interfaceeth0
virtual_router_id 51
priority99
advert_int 1
authentication {
auth_type PASS
auth_pass asdfgh
}
virtual_ipaddress {
172.16.4.1/32 brd 172.16.4.1 dev eth0 label eth0:0
}
track_script {
chk_mt_down
}
}
vrrp_instance VI_2 {
stateMASTER
interfaceeth0
virtual_router_id 52
priority100
advert_int 1
authentication {
auth_type PASS
auth_pass qwerty
}
virtual_ipaddress {
172.16.4.2
}
track_script {
chk_mt_down
}
}设置完成之后两个节点分别启动haproxy和keepalived服务，集群就配置完成了
[root@HK-01 ~]# service haproxy start [root@HK-01 ~]# service keepalived start [root@HK-02 ~]# service haproxy start [root@HK-02 ~]# service keepalived start
访问状态页面，www1，和www2示例都显示正常
VIP地址也可以正常启动
[root@HK-01 ~]# ip addr show dev eth0 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP>mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 00:0c:29:22:c5:c2 brd ff:ff:ff:ff:ff:ff inet172.16.4.100/16 brd 172.16.255.255 scope global eth0 inet172.16.4.1/32 brd 172.16.4.1 scope global eth0:0 inet6fe80::20c:29ff:fe22:c5c2/64 scope link valid_lft forever preferred_lft forever [root@HK-02 ~]# ip addr show dev eth0 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP>mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 00:0c:29:f1:dd:b2 brd ff:ff:ff:ff:ff:ff inet172.16.4.101/16 brd 172.16.255.255 scope global eth0 inet172.16.4.2/32 scope global eth0 inet6fe80::20c:29ff:fef1:ddb2/64 scope link valid_lft forever preferred_lft forever
分别访问两个vip地址均实现了负载均衡的效果
[root@localhost ~]# curl 172.16.4.1 web-02 [root@localhost ~]# curl 172.16.4.1 web-01 [root@localhost ~]# curl 172.16.4.2 web-01 [root@localhost ~]# curl 172.16.4.2 web-02
手动关闭其中HK-02，验证vip地址是否会自动漂移到HK-01服务器
[root@HK-02 ~]# touch /etc/keepalived/down
两个vip地址均正常漂移到HK-02
[root@HK-01 ~]# ip addr show dev eth0 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP>mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 00:0c:29:22:c5:c2 brd ff:ff:ff:ff:ff:ff inet172.16.4.100/16 brd 172.16.255.255 scope global eth0 inet172.16.4.1/32 brd 172.16.4.1 scope global eth0:0 inet172.16.4.2/32 scope global eth0 inet6fe80::20c:29ff:fe22:c5c2/64 scope link valid_lft forever preferred_lft forever
手动关闭web-02，查看是否会自动下线web-02
[root@WEB-02 ~]# service httpd stop
查看web状态页面，WEB-02已经自动下线了
访问验证，调度请求不会转发到web-02
[root@localhost ~]# curl 172.16.4.1 web-01 [root@localhost ~]# curl 172.16.4.1 web-01 [root@localhost ~]# curl 172.16.4.2 web-01 [root@localhost ~]# curl 172.16.4.2 web-01
本文出自 “梅花香自苦寒来” 博客，请务必保留此出处http://ximenfeibing.blog.51cto.com/8809812/1663395
Haproxy+keepalived实现双主负载均衡高可用集群
标签：haproxy双主
原文地址：http://ximenfeibing.blog.51cto.com/8809812/1663395
