Nginx负载均衡&高可用配置

最新推荐文章于 2025-05-21 20:52:28 发布

原创最新推荐文章于 2025-05-21 20:52:28 发布 · 320 阅读

1 ·

CC 4.0 BY-SA版权

文章标签：

#nginx #负载均衡 #运维

本文介绍了如何配置Nginx作为负载均衡器实现高可用性。通过部署RS01和RS02作为业务服务器，设置LB01为主负载均衡器，LB02为备用，并详细阐述了负载均衡的测试、HA部署以及使用监控脚本来实现半自动的主备切换过程。

Nginx负载均衡&高可用配置

环境说明：

各主机均已关闭防火墙与SELinux。

主机名	IP地址	应用服务	系统
LB01	192.168.92.130	keepalived nginx	Centos8
LB02	192.168.92.129	keepalived nginx	Centos8
RS01	192.168.92.132	nginx	Centos8
RS02	192.168.92.133	nginx	Centos8

需求：

LB01做主负载均衡器，LB02做备负载均衡器，VIP设为192.168.92.200。RS01与RS02做实际处理业务请求的服务器。

部署RS

RS01主机配置

#安装nginx
[root@RS01 ~]# yum -y install nginx

#先将原首页文件备份，再定义新的首页文件内容
[root@RS01 ~]# cd /usr/share/nginx/html/
[root@RS01 html]# ls
404.html  50x.html  index.html  nginx-logo.png  poweredby.png
[root@RS01 html]# mv index.html{,.bak}
[root@RS01 html]# echo 'This is RS01.' > index.html
[root@RS01 html]# ls
404.html  50x.html  index.html  index.html.bak  nginx-logo.png  poweredby.png

#启动nginx并设为开机自启
[root@RS01 html]# systemctl enable --now nginx.service

RS02主机配置

[root@RS02 ~]# dnf -y install nginx
[root@RS02 ~]# cd /usr/share/nginx/html/
[root@RS02 html]# mv index.html{,.bak}
[root@RS02 html]# echo "This is RS02." > index.html
[root@RS02 html]# ls
404.html  50x.html  index.html  index.html.bak  nginx-logo.png  poweredby.png
[root@RS02 html]# systemctl enable --now nginx.service

测试两台RS能否访问

[root@LB01 ~]# curl 192.168.92.132
This is RS01.
[root@LB01 ~]# curl 192.168.92.133
This is RS02.

部署LB

LB01主机做负载均衡

#安装nginx
[root@LB01 ~]# dnf -y install nginx

#修改配置文件前先对原文件做备份，养成身为运维的良好习惯
[root@LB01 ~]# cd /etc/nginx/
[root@LB01 nginx]# cp nginx.conf nginx.conf.bak
[root@LB01 nginx]# ls
conf.d                fastcgi_params          mime.types          nginx.conf.default   uwsgi_params.default
default.d             fastcgi_params.default  mime.types.default  scgi_params          win-utf
fastcgi.conf          koi-utf                 nginx.conf          scgi_params.default
fastcgi.conf.default  koi-win                 nginx.conf.bak      uwsgi_params

#配置负载均衡
[root@LB01 nginx]# vim nginx.conf
    upstream webserver {		#定义后端实际处理业务请求的服务器池
        server 192.168.92.132;	  #RS01的IP
        server 192.168.92.133;	  #RS02的IP
    }

    server {
        listen       80;
        server_name  _;
        root         /usr/share/nginx/html;

        include /etc/nginx/default.d/*.conf;

        location / {
            proxy_pass http://webserver;	
        }


[root@LB01 nginx]# systemctl enable --now nginx.service

测试负载均衡：

#因没有分配权重，默认是1:1轮询
[root@LB01 nginx]# curl 192.168.92.130
This is RS01.
[root@LB01 nginx]# curl 192.168.92.130
This is RS02.
[root@LB01 nginx]# curl 192.168.92.130
This is RS01.
[root@LB01 nginx]# curl 192.168.92.130
This is RS02.

LB02主机做负载均衡

#安装nginx
[root@LB02 ~]# dnf -y install nginx

[root@LB02 ~]# cd /etc/nginx/
[root@LB02 nginx]# cp nginx.conf nginx.conf.bak
[root@LB02 nginx]# vim nginx.conf

    upstream webserver {
        server 192.168.92.132;
        server 192.168.92.133;
    }

    server {
        listen       80;
        server_name  _;
        root         /usr/share/nginx/html;

        include /etc/nginx/default.d/*.conf;

        location / {
            proxy_pass http://webserver;
        }

[root@LB02 nginx]# systemctl start nginx.service

测试负载均衡：

[root@LB02 nginx]# curl 192.168.92.129
This is RS01.
[root@LB02 nginx]# curl 192.168.92.129
This is RS02.
[root@LB02 nginx]# curl 192.168.92.129
This is RS01.
[root@LB02 nginx]# curl 192.168.92.129
This is RS02.
#测试完停止nginx服务
[root@LB02 nginx]# systemctl stop nginx.service

部署HA

LB01做主LB

#下载做高可用的软件
[root@LB01 ~]# dnf -y install keepalived

#生成8位数的密码
[root@LB01 keepalived]# strings /dev/urandom |tr -dc A-Za-z0-9 | head -c8; echo
pP5ek1YA

#配置keepalived
[root@LB01 keepalived]# vim keepalived.conf
! Configuration File for keepalived

global_defs {
   router_id lb01
}

vrrp_instance VI_1 {
    state MASTER
    interface ens32
    virtual_router_id 81
    priority 100
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass pP5ek1YA
    }
    virtual_ipaddress {
        192.168.92.200
    }
}

virtual_server 192.168.92.200 80 {
    delay_loop 6
    lb_algo rr
    lb_kind DR
    persistence_timeout 50
    protocol TCP

    real_server 192.168.92.130 80 {
        weight 1
        TCP_CHECK {
            connect_port 80
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
        }
    }

    real_server 192.168.92.129 80 {
        weight 1
        TCP_CHECK {
            connect_port 80
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
        }
    }
}
#开启keepalived并设为开机自启
[root@LB01 ~]# systemctl enable --now keepalived.service

#可以看到VIP已经有了
[root@LB01 ~]# ip a s ens32
2: ens32: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
    link/ether 00:0c:29:9e:e3:c1 brd ff:ff:ff:ff:ff:ff
    inet 192.168.92.130/24 brd 192.168.92.255 scope global dynamic noprefixroute ens32
       valid_lft 1707sec preferred_lft 1707sec
    inet 192.168.92.200/32 scope global ens32
       valid_lft forever preferred_lft forever
    inet6 fe80::20c:29ff:fe9e:e3c1/64 scope link noprefixroute 
       valid_lft forever preferred_lft forever

#使用VIP进行访问。如果访问不了又确信配置无误，则极有可能是备负载均衡器服务没停止
[root@LB01 ~]# curl 192.168.92.200
This is RS01.
[root@LB01 ~]# curl 192.168.92.200
This is RS02.
[root@LB01 ~]# curl 192.168.92.200
This is RS01.
[root@LB01 ~]# curl 192.168.92.200
This is RS02.

验证究竟是否是LB01(主)主机在做反向代理

这里有必要简述一下nginx反向代理的工作流程：反向代理服务器接收访问用户的请求后，会代理用户重新发起请求代理下的节点服务器，最后把数据返回给客户端用。所以被代理的节点服务器并不知道客户端的存在，因为它所处理的全部请求都是由代理服务器请求的。

#在LB02主机上用VIP进行访问
[root@LB02 nginx]# curl 192.168.92.200
This is RS01.
[root@LB02 nginx]# curl 192.168.92.200
This is RS02.
[root@LB02 nginx]# curl 192.168.92.200
This is RS01.
[root@LB02 nginx]# curl 192.168.92.200
This is RS02.

#在RS01主机上查看日志
[root@RS01 html]# cd /var/log/nginx/
[root@RS01 nginx]# ls
access.log  error.log
#可以看到访问主机的IP确实是LB01这台
[root@RS01 nginx]# tail -f access.log
192.168.92.130 - - [17/Oct/2022:20:41:21 +0800] "GET / HTTP/1.0" 200 14 "-" "curl/7.61.1" "-"
192.168.92.130 - - [17/Oct/2022:20:41:23 +0800] "GET / HTTP/1.0" 200 14 "-" "curl/7.61.1" "-"

LB02做备LB

[root@LB02 ~]# dnf -y install keepalived

[root@LB02 ~]# cd /etc/keepalived/
[root@LB02 keepalived]# mv keepalived.conf{,.bak}
#将LB01主机的keepalived配置文件直接copy过来
[root@LB02 keepalived]# scp root@192.168.92.130:/etc/keepalived/keepalived.conf ./
[root@LB02 keepalived]# ls
keepalived.conf  keepalived.conf.bak

#修改配置文件。仅有两个地方需要注意，其一是state，设为backup。其二是priority，一定要比主低
[root@LB02 keepalived]# vim keepalived.conf
! Configuration File for keepalived

global_defs {
   router_id lb01
}

vrrp_instance VI_1 {
    state BACKUP
    interface ens32
    virtual_router_id 81
    priority 90
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass pP5ek1YA
    }
    virtual_ipaddress {
        192.168.92.200
    }
}

virtual_server 192.168.92.200 80 {
    delay_loop 6
    lb_algo rr
    lb_kind DR
    persistence_timeout 50
    protocol TCP

    real_server 192.168.92.130 80 {
        weight 1
        TCP_CHECK {
            connect_port 80
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
        }
    }

    real_server 192.168.92.129 80 {
        weight 1
        TCP_CHECK {
            connect_port 80
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
        }
    }
}
[root@LB02 keepalived]# systemctl enable --now keepalived.service

测试主备切换

#模拟主负载均衡器出现故障
[root@LB01 ~]# systemctl stop nginx keepalived.service

#去到备负载均衡器上查看VIP
[root@LB02 ~]# ip a s ens32
2: ens32: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
    link/ether 00:0c:29:e2:b1:9f brd ff:ff:ff:ff:ff:ff
    inet 192.168.92.129/24 brd 192.168.92.255 scope global dynamic noprefixroute ens32
       valid_lft 1317sec preferred_lft 1317sec
    inet 192.168.92.200/32 scope global ens32
       valid_lft forever preferred_lft forever
    inet6 fe80::20c:29ff:fee2:b19f/64 scope link noprefixroute 
       valid_lft forever preferred_lft forever

#启动nginx进行负载均衡
[root@LB02 ~]# systemctl start nginx.service
[root@LB02 ~]# curl 192.168.92.200
This is RS01.
[root@LB02 ~]# curl 192.168.92.200
This is RS02.
[root@LB02 ~]# curl 192.168.92.200
This is RS01.
[root@LB02 ~]# curl 192.168.92.200
This is RS02.

#来到RS01主机上查看访问日志，可以看到此时显示源IP是LB02
[root@RS01 nginx]# tail -f access.log
192.168.92.129 - - [17/Oct/2022:21:10:31 +0800] "GET / HTTP/1.0" 200 14 "-" "curl/7.61.1" "-"
192.168.92.129 - - [17/Oct/2022:21:10:33 +0800] "GET / HTTP/1.0" 200 14 "-" "curl/7.61.1" "-"

#如果你想继续做监控脚本实现半自动主备切换，那么请恢复到LB01为主负载均衡器
[root@LB02 ~]# systemctl stop nginx.service
[root@LB01 ~]# systemctl start nginx.service keepalived.service
[root@LB01 ~]# ip a s ens32
2: ens32: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
    link/ether 00:0c:29:9e:e3:c1 brd ff:ff:ff:ff:ff:ff
    inet 192.168.92.130/24 brd 192.168.92.255 scope global dynamic noprefixroute ens32
       valid_lft 1205sec preferred_lft 1205sec
    inet 192.168.92.200/32 scope global ens32
       valid_lft forever preferred_lft forever
    inet6 fe80::20c:29ff:fe9e:e3c1/64 scope link noprefixroute 
       valid_lft forever preferred_lft forever

配置监控脚本实现半自动主备切换

所谓半自动主备切换意思是，当主ka(keepalived)挂掉了，监控脚本检测到后，备ka会自动成为新的主ka。当旧主ka恢复后想要重新成为主卡时需要系统管理员手动切换。

LB01主机配置

[root@LB01 ~]# mkdir /scripts && cd /scripts
[root@LB01 scripts]# vim check_nginx.sh
#!/bin/bash
nginx_status=$(ps -ef|grep -Ev "grep|$0"|grep '\bnginx\b'|wc -l)
if [ $nginx_status -lt 1 ];then
        systemctl stop keepalived
fi

[root@LB01 scripts]# vim notify.sh
#!/bin/bash
case "$1" in
    master)
        nginx_status=$(ps -ef|grep -Ev "grep|$0"|grep '\bnginx\b'|wc -l)
        if [ $nginx_status -lt 1 ];then
            systemctl start nginx
        fi
    ;;
    backup)
        nginx_status=$(ps -ef|grep -Ev "grep|$0"|grep '\bnginx\b'|wc -l)
        if [ $nginx_status -gt 0 ];then
            systemctl stop nginx
        fi
    ;;
    *)
        echo "Usage:$0 master|backup VIP"
    ;;
esac

[root@LB01 scripts]# chmod +x check_nginx.sh notify.sh 
[root@LB01 scripts]# ll
total 8
-rwxr-xr-x 1 root root 139 Oct 17 23:09 check_nginx.sh
-rwxr-xr-x 1 root root 392 Oct 17 23:20 notify.sh

#将监控脚本配置到keepalived
[root@LB01 ~]# vim /etc/keepalived/keepalived.conf
! Configuration File for keepalived

global_defs {
   router_id lb01
}
#填空以下这五行
vrrp_script nginx_check {
    script "/scripts/check_nginx.sh"
    interval 1
    weight -20
}

vrrp_instance VI_1 {
    state MASTER
    interface ens32
    virtual_router_id 81
    priority 100
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass pP5ek1YA
    }
    virtual_ipaddress {
        192.168.92.200
    }
    track_ipaddress{	#添加以下四行
        nginx_check
    }
    notify_master "/scripts/notify.sh master"
}

virtual_server 192.168.92.200 80 {
    delay_loop 6
    lb_algo rr
    lb_kind DR
    persistence_timeout 50
    protocol TCP

    real_server 192.168.92.130 80 {
        weight 1
        TCP_CHECK {
            connect_port 80
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
        }
    }

    real_server 192.168.92.129 80 {
        weight 1
        TCP_CHECK {
            connect_port 80
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
        }
    }
}
[root@LB01 ~]# systemctl restart keepalived.service

LB02主机配置

backup无需检测nginx是否正常，当升级为MASTER时启动nginx，当降级为BACKUP时关闭

[root@LB02 ~]# mkdir /scripts && cd /scripts
[root@LB02 scripts]# scp root@192.168.92.130:/scripts/notify.sh ./
[root@LB02 scripts]# cat notify.sh 
#!/bin/bash
case "$1" in
    master)
        nginx_status=$(ps -ef|grep -Ev "grep|$0"|grep '\bnginx\b'|wc -l)
        if [ $nginx_status -lt 1 ];then
            systemctl start nginx
        fi
    ;;
    backup)
        nginx_status=$(ps -ef|grep -Ev "grep|$0"|grep '\bnginx\b'|wc -l)
        if [ $nginx_status -gt 0 ];then
            systemctl stop nginx
        fi
    ;;
    *)
        echo "Usage:$0 master|backup VIP"
    ;;
esac
[root@LB02 scripts]# ll
total 4
-rwxr-xr-x 1 root root 376 Oct 17 23:34 notify.sh

[root@LB02 scripts]# vim /etc/keepalived/keepalived.conf
! Configuration File for keepalived

global_defs {
   router_id lb01
}

vrrp_instance VI_1 {
    state BACKUP
    interface ens32
    virtual_router_id 81
    priority 90
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass pP5ek1YA
    }
    virtual_ipaddress {
        192.168.92.200
    }
    notify_master "/scripts/notify.sh master"	#添加这两行
    notify_backup "/scripts/notify.sh backup"
}

virtual_server 192.168.92.200 80 {
    delay_loop 6
    lb_algo rr
    lb_kind DR
    persistence_timeout 50
    protocol TCP

    real_server 192.168.92.130 80 {
        weight 1
        TCP_CHECK {
            connect_port 80
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
        }
    }

    real_server 192.168.92.129 80 {
        weight 1
        TCP_CHECK {
            connect_port 80
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
        }
    }
}
[root@LB02 scripts]# systemctl restart keepalived.service

测试配置监控脚本是否能自动进行主备切换

#目前VIP在LB01主机上，说明此时还是主
[root@LB01 ~]# ip a s ens32
2: ens32: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
    link/ether 00:0c:29:9e:e3:c1 brd ff:ff:ff:ff:ff:ff
    inet 192.168.92.130/24 brd 192.168.92.255 scope global dynamic noprefixroute ens32
       valid_lft 1534sec preferred_lft 1534sec
    inet 192.168.92.200/32 scope global ens32
       valid_lft forever preferred_lft forever
    inet6 fe80::20c:29ff:fe9e:e3c1/64 scope link noprefixroute 
       valid_lft forever preferred_lft forever

#手动停止nginx均衡负载器，模拟故障
[root@LB01 ~]# systemctl stop nginx.service
#可以看到由于负载均衡器挂掉了，运行脚本停掉了keepalived。VIP也不在了
[root@LB01 scripts]# systemctl status keepalived.service 
● keepalived.service - LVS and VRRP High Availability Monitor
   Loaded: loaded (/usr/lib/systemd/system/keepalived.service; enabled; vendor preset: disabled)
   Active: inactive (dead) since Mon 2022-10-17 23:42:38 CST; 10s ago
[root@LB01 scripts]# ip a s ens32
2: ens32: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
    link/ether 00:0c:29:9e:e3:c1 brd ff:ff:ff:ff:ff:ff
    inet 192.168.92.130/24 brd 192.168.92.255 scope global dynamic noprefixroute ens32
       valid_lft 1326sec preferred_lft 1326sec
    inet6 fe80::20c:29ff:fe9e:e3c1/64 scope link noprefixroute 
       valid_lft forever preferred_lft forever

#此时去到LB02查看VIP，可以看到VIP在这台负载均衡器上了
[root@LB02 ~]# ip a s ens32
2: ens32: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
    link/ether 00:0c:29:e2:b1:9f brd ff:ff:ff:ff:ff:ff
    inet 192.168.92.129/24 brd 192.168.92.255 scope global dynamic noprefixroute ens32
       valid_lft 1230sec preferred_lft 1230sec
    inet 192.168.92.200/32 scope global ens32
       valid_lft forever preferred_lft forever
    inet6 fe80::20c:29ff:fee2:b19f/64 scope link noprefixroute 
       valid_lft forever preferred_lft forever
#可以看到nginx的默认80也随之启用
[root@LB02 ~]# ss -anlt | grep 80
LISTEN 0      128          0.0.0.0:80        0.0.0.0:*    

#要想再次启用LB01为主，则需自行手动启动nginx与keepalived服务