Centos7 builds LVS cluster-NAT mode, DR mode

One, build LVS-NAT cluster

1. Preparation

Host	operating system	Virtual network card	IP address	Gateway
LVS load balancing scheduler	Centos7	VM1，2	192.168.1.10，192.168.100.10	Act as a gateway
Web-AAA	Centos7	VM1	192.168.1.3	192.168.1.10
Web-BBB	Centos7	VM1	192.168.1.4	192.168.1.10
NFS shared storage	Centos7	VM1	192.168.1.110	192.168.1.10
Extranet client	Win7	VM2	192.168.100.20	192.168.100.10

2. Configure LVS server

1) Load the module

[root@localhost ~]# modprobe ip_vs

2) Confirm the kernel's support for LVS

[root@localhost ~]# modprobe ip_vs

IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
-> RemoteAddress:Port Forward Weight ActiveConn InActConn
3) Enable routing and forwarding

[root@localhost ~]# vim /etc/sysctl.conf
net.ipv4.ip_forward = 1
[root@localhost ~]# sysctl -p

4) Use ipvsadm tool (LVS cluster management tool)
挂载光盘，并配置yum源

[root@localhost ~]# yum -y install ipvsadm

To create a virtual server, the VIP must be the real IP address of the machine

[root@localhost ~]# ipvsadm -C    （清除原有的策略）
[root@localhost ~]# ipvsadm -A -t 192.168.100.10:80 -s rr

1. -A add virtual server
2. -t VIP address and port number
3. -s load scheduling algorithm (polling rr, weighted polling wrr, least connection lc, weighted least connection wlc)
   add server node Web-AAA, Web-BBB

[root@localhost ~]# ipvsadm -a -t 192.168.100.10:80 -r 192.168.1.3:80 -m -w 1
[root@localhost ~]# ipvsadm -a -t 192.168.100.10:80 -r 192.168.1.4:80 -m -w 1

1. -a Add real server
2. -t vip address and tcp port number
3. -r rip address and port number
4. -m Use nat cluster default (-g DR mode, -i TUN mode)
5. -w Set the weight (0 is to suspend the node) to
   save the strategy, and set the boot to start

[root@localhost ~]# ipvsadm-save
[root@localhost ~]# systemctl enable ipvsadm

3. NFS shared storage server configuration

挂载光盘，并配置yum源
1) Install nfs-utils and rpcbind software packages

[root@localhost ~]# yum -y install nfs-utils rpcbind
[root@localhost ~]# systemctl enable nfs
[root@localhost ~]# systemctl enable rpcbind

2) Set up a shared directory

[root@localhost ~]# mkdir -p /opt/web1 /opt/web2

3) Set up nfs shared files
将文件夹web1,web2分别共享给web1，web2主机 使用，允许读写操作

[root@localhost ~]# vi /etc/exports
添加：
/opt/web1 192.168.1.3(rw,sync,no_root_squash)
/opt/web2 192.168.1.4(rw,sync,no_root_squash)

Note:
rw 读写
sync 同步写入
no_root_squash 客户机以root身份访问时赋予本地root权限
4) Start the NFS service program. Note: Start rpcbind first, then nfs

[root@localhost ~]# systemctl start rpcbind
[root@localhost ~]# systemctl start nfs
[root@localhost ~]# netstat -anpu | grep rpcbind

5) View the NFS shared directory published by this machine

[root@localhost ~]# showmount -e

4.Web-AAA host

挂载光盘，并配置yum源
1) Install httpd service

[root@Web-AAA ~]# yum -y install httpd

2) Configure nfs shared storage client

[root@Web-AAA ~]# yum -y install rpcbind nfs-utils
[root@Web-AAA ~]# systemctl enable rpcbind
[root@Web-AAA ~]# systemctl start rpcbind
[root@Web-AAA ~]# showmount -e 192.168.1.110
[root@Web-AAA ~]# mount 192.168.1.110:/opt/web1 /var/www/html/
[root@Web-AAA ~]# echo "<h1>this is web1</h1>" > /var/www/html/index.html
[root@Web-AAA ~]# echo "192.168.1.110:/opt/web1 /var/www/html nfs defaults,_netdev 0 0" > /etc/fstab
[root@Web-AAA ~]# tail -1 /etc/fstab
[root@Web-AAA ~]# systemctl enable httpd
[root@Web-AAA ~]# systemctl start httpd

5.Web-BBB host

挂载光盘，并配置yum源
1) Install httpd service

[root@Web-BBB ~]# yum -y install httpd

2) Configure nfs shared storage client

[root@Web-BBB ~]# yum -y install rpcbind nfs-utils
[root@Web-BBB ~]# systemctl enable rpcbind
[root@Web-BBB ~]# systemctl start rpcbind
[root@Web-BBB ~]# showmount -e 192.168.1.110
[root@Web-BBB ~]# mount 192.168.1.110:/opt/web2 /var/www/html/
[root@Web-BBB ~]# echo "<h1>this is web2</h1>" > /var/www/html/index.html
[root@Web-BBB ~]# echo "192.168.1.110:/opt/web2 /var/www/html nfs defaults,_netdev 0 0" > /etc/fstab
[root@Web-BBB ~]# tail -1 /etc/fstab
[root@Web-BBB ~]# systemctl enable httpd
[root@Web-BBB ~]# systemctl start httpd

6. External network client test

访问：http：//192.168.100.10
多访问几次，不然看不到效果

2. Direct routing mode (LVS-DR)

1. Preparation

You can continue to use the above experimental environment, but do the following modifications
将调度器的外网网卡删掉，就是VM2那块网卡
将Win客户机的VM2改为VM1，IP改为：192.168.1.88

2. Configure the load scheduler

1) Configure Virtual IP Address (VIP)
采用虚接口的方式，为ens33绑定VIP地址，以便响应群集访问
[root@localhost ~]# cd /etc/sysconfig/network-scripts/
[root@localhost network-scripts]# rm -rf ifcfg-ens37
[root@localhost network-scripts]# cp ifcfg-ens33 ifcfg-ens33:0
[root@localhost network-scripts]# vi ifcfg-ens33:0
First delete the UUID configuration items, and then change the following configuration items to the following
NAME="ens33:0"
DEVICE="ens33:0"
IPADDR="192.168.1.188"
[root@localhost network-scripts]# systemctl restart network
[ root@localhost network-scripts]# ifconfig ens33:0

2) Adjust the /proc response parameters
对于DR群集模式来说，用于LVS负载调度器和各节点需要公用VIP地址，应该关闭linux内核重定向参数

[root@localhost ~]# vi /etc/sysctl.conf
删除原有的：net.ipv4.ip_forward = 1
添加：
net.ipv4.conf.all.send_redirects = 0
net.ipv4.conf.default.send_redirects = 0
net.ipv4.conf.ens33.send_redirects = 0
[root@localhost ~]# sysctl -p

3) Configure load distribution strategy

[root@localhost ~]# ipvsadm -C
[root@localhost ~]# ipvsadm -A -t 192.168.1.188:80 -s rr
[root@localhost ~]# ipvsadm -a -t 192.168.1.188:80 -r 192.168.1.3 -g -w 1
[root@localhost ~]# ipvsadm -a -t 192.168.1.188:80 -r 192.168.1.4 -g -w 1
[root@localhost ~]# ipvsadm-save

3. Configure the web node server

Web-AAA和WebBBB服务器操作都如下：
1) Configure the virtual ip address VIP
使用DR模式，节点服务器，也需要配置VIP地址，使用这个地址做web响应的源地址

[root@Web-AAA ~]# cd /etc/sysconfig/network-scripts/
[root@Web-AAA network-scripts]# cp ifcfg-lo ifcfg-lo:0
[root@Web-AAA network-scripts]# vi ifcfg-lo:0
修改：
修改DEVICE=lo:0
NETMASK=255.255.255.255           \\子网掩码必须为255.255.255.255
IPADDR=192.168.1.188               \\改为VIP地址
[root@Web-AAA network-scripts]# systemctl restart network
[root@Web-AAA network-scripts]# ip a

Add a routing record for this machine to limit the VIP data locally to avoid communication chaos

[root@Web-AAA ~]# yum -y install net-tools
[root@Web-AAA ~]# vi /etc/rc.local
添加：
/sbin/route add -host 192.168.1.188 dev lo:0
[root@Web-AAA ~]# route add -host 192.168.1.188 dev lo:0

2) Adjust the kernel parameters

[root@Web-AAA ~]# vi /etc/sysctl.conf
添加：
net.ipv4.conf.all.arp_ignore = 1
net.ipv4.conf.all.arp_announce = 2
net.ipv4.conf.default.arp_ignore = 1
net.ipv4.conf.default.arp_announce = 2
net.ipv4.conf.lo.arp_ignore = 1
net.ipv4.conf.lo.arp_announce = 2
[root@Web-AAA ~]# sysctl -p

3) Install httpd to create a test web page, and then mount the NFS share
4) Enable httpd service.
Since the experiment in Figure 1 is used, it has already been done, so we will not do steps 3 or 4

Test the LVS cluster

在win7客户机上将VM2改为VM1，并设置IP为192.168.1.88

测试时频繁刷新页面


You can also use the script to test:

5) View the LVS strategy on the LVS scheduler

1. Masq-address masquerading NAT mode
2. Route-DR mode
3. Weight—weight