LVS负载均衡验证轮询算法和加权轮询算法

实验介绍

本实验通过LVS负载均衡的方式去访问一个nginx集群，nginx1显示“web test page, my ip is 192.168.112.136 192.168.112.140 .”，nginx2显示“web test page, my ip is 192.168.112.137 192.168.112.140 .”

轮询算法： 轮询算法是将客户端的请求轮流分发给内部的服务器

实验环境

LVS主机	192.168.112.128
client	192.168.112.135
nginx1	192.168.112.136
nginx2	192.168.112.137

实验步骤

配置nginx1和nginx2：添加VIP地址

Nginx1：

[root@node3 ~]# nmcli connection add type dummy ifname dummy1 ipv4.method manual ipv4.addresses 192.168.112.140/24
[root@node3 ~]# 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: ens160: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000
    link/ether 00:0c:29:d8:7a:37 brd ff:ff:ff:ff:ff:ff
    inet 192.168.112.136/24 brd 192.168.112.255 scope global dynamic noprefixroute ens160
       valid_lft 1304sec preferred_lft 1304sec
    inet6 fe80::20c:29ff:fed8:7a37/64 scope link noprefixroute 
       valid_lft forever preferred_lft forever
3: dummy1: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 qdisc noqueue state UNKNOWN group default qlen 1000
    link/ether a6:59:c2:e9:00:7e brd ff:ff:ff:ff:ff:ff
    inet 192.168.112.140/24 brd 192.168.112.255 scope global noprefixroute dummy1
       valid_lft forever preferred_lft forever
    inet6 fe80::9f55:22b5:ca89:af68/64 scope link noprefixroute 
       valid_lft forever preferred_lft forever

Nginx2：

[root@node4 ~]# nmcli connection add type dummy ifname dummy1 ipv4.method manual ipv4.addresses 192.168.112.140/24

修改相应的arp内核配置

nginx1和nginx2两台机器共同执行命令

[root@node4 ~]# cat >> /etc/sysctl.conf << EOF
> net.ipv4.conf.all.arp_ignore = 1
> net.ipv4.conf.all.arp_announce = 2
> net.ipv4.conf.dummy1.arp_ignore = 1
> net.ipv4.conf.dummy1.arp_announce = 2
> EOF

[root@node4 ~]# sysctl -p | tail -4
net.ipv4.conf.all.arp_ignore = 1
net.ipv4.conf.all.arp_announce = 2
net.ipv4.conf.dummy1.arp_ignore = 1
net.ipv4.conf.dummy1.arp_announce = 2

安装Nginx服务，输入一条信息到/nginx/html/index.html中

nginx1和nginx2都需要执行

nginx2
[root@node4 ~]# dnf install -y nginx
[root@node4 ~]# echo "web test page, my ip is `hostname -I`. " > /usr/share//nginx/html/index.html 
[root@node4 ~]# systemctl restart nginx.service
[root@node4 ~]# curl 192.168.112.137
web test page, my ip is 192.168.112.137 192.168.112.140 .

nginx1
[root@node3 ~]#  echo "web test page, my ip is `hostname -I`. " > /usr/share//nginx/html/index.html
[root@node3 ~]# systemctl restart nginx.service
[root@node3 ~]# curl 192.168.112.140
web test page, my ip is 192.168.112.136 192.168.112.140 .

LVS配置

[root@node1 ~]# dnf install ipvsadm -y
[root@node1 ~]# ipvsadm -At 192.168.112.140:80
[root@node1 ~]# ipvsadm -at 192.168.112.140:80 -r 192.168.112.136:80
[root@node1 ~]# ipvsadm -at 192.168.112.140:80 -r 192.168.112.137:80
[root@node1 ~]# ipvsadm -Ln
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  192.168.112.140:80 wlc
  -> 192.168.112.136:80           Route   1      0          0         
  -> 192.168.112.137:80           Route   1      0          0   
      
-s 修改为轮询算法

[root@node1 ~]# ipvsadm -Et 192.168.112.140:80 -s rr
[root@node1 ~]# ipvsadm -Ln
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  192.168.112.140:80 rr
  -> 192.168.112.136:80           Route   1      0          0         
  -> 192.168.112.137:80           Route   1      0          0 

验证LVS的配置

在client客户端执行"curl 192.168.112.140",注意是vip的地址

[root@node2 ~]# curl 192.168.112.140
curl: (7) Failed to connect to 192.168.112.140 port 80 after 22 ms: No route to host

 出现报错“curl: (7) Failed to connect to 192.168.112.140 port 80 after 22 ms: No route to host”

检查防火墙是否关闭，没有关闭执行如下命令：

[root@node2 ~]# systemctl stop firewalld
[root@node2 ~]# setenforce 0

再次执行"curl 192.168.112.140"

[root@node2 ~]# curl 192.168.112.140
web test page, my ip is 192.168.112.136 192.168.112.140 . 
[root@node2 ~]# curl 192.168.112.140
web test page, my ip is 192.168.112.137 192.168.112.140 . 
[root@node2 ~]# for ((i=1;i<=6;i++)); do curl 192.168.112.140;done 
web test page, my ip is 192.168.112.136 192.168.112.140 . 
web test page, my ip is 192.168.112.137 192.168.112.140 . 
web test page, my ip is 192.168.112.136 192.168.112.140 . 
web test page, my ip is 192.168.112.137 192.168.112.140 . 
web test page, my ip is 192.168.112.136 192.168.112.140 . 
web test page, my ip is 192.168.112.137 192.168.112.140 . 

结果如上，轮询验证成功！

修改LVS配置验证加权轮询算法

在LVS主机执行

[root@node1 ~]#  ipvsadm -Et 192.168.112.140:80 -s wrr
[root@node1 ~]# ipvsadm -Ln
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  192.168.112.140:80 wrr
  -> 192.168.112.136:80           Route   1      0          0         
  -> 192.168.112.137:80           Route   1      0          0         
[root@node1 ~]# 
[root@node1 ~]# ipvsadm -et 192.168.112.140:80 -r 192.168.112.137:80 -w 2
[root@node1 ~]# ipvsadm -Ln
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  192.168.112.140:80 wrr
  -> 192.168.112.136:80           Route   1      0          0         
  -> 192.168.112.137:80           Route   2      0          0 

在client客户端查询

[root@node2 ~]# for ((i=1;i<=6;i++)); do curl 192.168.112.140;done 
web test page, my ip is 192.168.112.137 192.168.112.140 . 
web test page, my ip is 192.168.112.136 192.168.112.140 . 
web test page, my ip is 192.168.112.137 192.168.112.140 . 
web test page, my ip is 192.168.112.137 192.168.112.140 . 
web test page, my ip is 192.168.112.136 192.168.112.140 . 
web test page, my ip is 192.168.112.137 192.168.112.140 .