CentOS7.4+Keepalived+HAproxy主主和主从架构实验部署手册

请叫我水哥 2020-08-26

563

一、Haproxy+Keepalived主主模式的高可用环境

1.1 环境准备

1.2 NFS配置

1.3 web服务器配置

1. 4. haproxy安装配置

1.4.1 haproxy安装

1.4.2 haproxy配置内容

1.5 keepalived安装配置

1.6 配置结果确认

1.7 访问测试

二、Haproxy+Keepalived主从模式的高可用环境

一、Haproxy+Keepalived主主模式的高可用环境

1.1 环境准备

1.2 NFS配置

//以web01作为服务端，web02作为客户端，进行网站目录数据同步
1）NFS安装
[root@web01 /]# yum -y install nfs-utils rpcbind
[root@web01 /]# mkdir /data
[root@web01 /]# cat /etc/exports
/data  192.168.10.0/24(rw,sync,root_squash)  //可读写，同步模式，客户端以root访问时被限制为普通用户
[root@web01 /]# systemctl start rpcbind.service 
[root@web01 /]# systemctl enable rpcbind.service 
[root@web01 /]# systemctl start nfs.service
[root@web01 /]# systemctl enable nfs.service
2) 客户端挂载
[root@web02 ~]# showmount -e 192.168.10.182
Export list for 192.168.10.182:
/data 192.168.10.0/24
[root@web02 ~]# mkdir /data
[root@web02 ~]# mount -t nfs 192.168.10.182:/data /data
[root@web02 ~]# df -hT |grep /data
192.168.10.182:/data    nfs4      2.0T  1.5G  2.0T    1% /data

1.3 web服务器配置

1）修改后端web01 web02两个网站根目录
[root@web01 ~]# cat /etc/nginx/conf.d/default.conf |grep "/data" 
        root   /data;
[root@web01 ~]# nginx -s reload    
[root@web02 ~]# cat /etc/nginx/conf.d/default.conf |grep "/data" 
        root   /data;
[root@web02 ~]# nginx -s reload
2）添加测试页
[root@web01 ~]# cat /data/index.html 
this is test page

1. 4. haproxy安装配置

1.4.1 haproxy安装

//haproxy01在之前的文章里面已经配置，现在准备haproxy02
[root@localhost ~]# yum -y install gcc gcc-c++ pcre-devel openssl-devel systemd-devel
[root@haproxy02 opt]# tar zxvf haproxy-2.2.2.tar.gz
[root@haproxy02 opt]# cd haproxy-2.2.2/
[root@haproxy02 haproxy-2.2.2]# make TARGET=linux-glibc PREFIX=/usr/local/haproxy USE_OPENSSL=1 ADDLIB=-lz USE_PCRE=1 USE_ZLIB=1 USE_SYSTEMD=1
[root@haproxy02 haproxy-2.2.2]# make install PREFIX=/usr/local/haproxy
[root@haproxy02 haproxy-2.2.2]# cp contrib/systemd/haproxy.service.in /usr/lib/systemd/system/haproxy.service
[root@haproxy02 haproxy-2.2.2]# ln -s /usr/local/haproxy/sbin/haproxy /usr/sbin/haproxy
[root@haproxy02 haproxy-2.2.2]# cd /usr/local/haproxy/
[root@haproxy02 haproxy]# mkdir conf
[root@haproxy02 conf]# cat /usr/lib/systemd/system/haproxy.service
[Unit]
Description=HAProxy Load Balancer
After=network-online.target
Wants=network-online.target
[Service]
ExecStartPre=/usr/sbin/haproxy -f /usr/local/haproxy/conf/haproxy.cfg -c -q
ExecStart=/usr/sbin/haproxy -Ws -f /usr/local/haproxy/conf/haproxy.cfg -p /run/haproxy.pid
ExecReload=/bin/kill -USR2 $MAINPID
KillMode=mixed
Restart=always
SuccessExitStatus=143
Type=notify
[Install]
WantedBy=multi-user.target
[root@haproxy02 etc]# cat /etc/rsyslog.conf
$ModLoad imudp
$UDPServerRun 514 
local0.*            /var/log/haproxy.log
[root@haproxy02 etc]# touch /var/log/haproxy.log
[root@haproxy02 etc]# systemctl restart rsyslog.service
//如果没有配置文件，可以拷贝示例文件    
[root@haproxy02 haproxy]# cp /opt/haproxy-2.2.2/examples/option-http_proxy.cfg conf/haproxy.cfg
//这里使用haproxy01已经准备好的文件
[root@haproxy01 haproxy]# scp conf/haproxy.cfg root@192.168.48.181:/usr/local/haproxy/conf/
//启动haproxy服务
[root@haproxy02 conf]# systemctl status haproxy && systemctl enable haproxy

1.4.2 haproxy配置内容

两个haproxy配置需要一样

global
        maxconn         20000
        ulimit-n  65535
        log             127.0.0.1 local0 info
        uid             1
        gid             1
        chroot          /usr/local/haproxy
        nbproc    1
        daemon
       pidfile    /run/haproxy.pid
defaults
  log    global
  mode    http
  option    redispatch
  stats  uri  /haproxy
  stats  refresh  30s
  stats   realm  haproxy-status
  stats  auth  admin:123456
  stats    hide-version
  timeout  connect  5000
  timeout  client  50000
  timeout  server  50000
  option  httpchk  GET  /index.html


frontend test-http
  bind    :80
   mode            http
   log             global
   option          httplog
   option          dontlognull
  option    forwardfor
  option    httpclose
  default_backend  nginx


backend nginx
  mode            http
  balance    roundrobin
  option    redispatch
  option    httpclose
  option    forwardfor
  retries         2
  server  web01  192.168.10.182:80 weight 1 check inter 2000 rise 3 fall 5
  server  web02  192.168.10.183:80 weight 1 check inter 2000 rise 3 fall 5

1.5 keepalived安装配置

1)安装配置
[root@haproxy01 haproxy]# yum -y install keepalived
[root@haproxy01 keepalived]# cat keepalived.conf
! Configuration File for keepalived


global_defs {
   notification_email {
     root@localhost
   }
   notification_email_from Alexandre.Cassen@firewall.loc
   smtp_server 127.0.0.1
   smtp_connect_timeout 30
   router_id LVS_DEVEL
}
vrrp_script check_haproxy {
  script "/root/check_haproxy.sh"
  interval 2
  weight  2
}


vrrp_instance VI_1 {
    state MASTER
    interface ens33
    virtual_router_id 51
    priority 100
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    track_script {
  check_haproxy
}
    virtual_ipaddress {
        192.168.48.200
    }
notify_master "/root/clean_arp.sh 192.168.48.200"
}


vrrp_instance VI_2 {
    state BACKUP
    interface ens33
    virtual_router_id 52
    priority 99
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {
        192.168.48.100
    }
notify_master "/root/clean_arp.sh 192.168.48.100"
}
2)拷贝配置到备节点
[root@haproxy01 keepalived]# scp keepalived.conf root@192.168.48.181:/etc/keepalived/
//备节点配置文件主要调整
vrrp_instance VI_1{
state BACKUP
virtual_router_id 51
priority 99
}
vrrp_instance VI_2{
state MASTER
virtual_router_id 52
priority 100
}
3)脚本配置
[root@haproxy01 ~]# cat check_haproxy.sh //keepalived会定期检查脚本(判断haproxy是否正常运行)
#!/bin/bash
status=`systemctl is-active haproxy`
if [ ${status} != "active" ];then
systemctl start haproxy
sleep 3
if [ `systemctl is-active haproxy` != "active" ];then
systemctl stop haproxy
fi
fi
[root@haproxy01 ~]# cat clean_arp.sh //主要用作keepalived启动的时候判断ip是否存在，如果存在则不启动keepalived
#/bin/sh
VIP=$1
gateway=192.168.48.2
arping -I ens33 -c 5 -s $VIP $gateway &> /dev/null
//arping用来向局域网内的其它主机发送ARP请求的指令,它可以用来测试局域网内的某个IP是否已被使用
4)启动keepalived，haproxy
[root@haproxy01 ~]# systemctl start keepalived.service && systemctl enable keepalived.service
[root@haproxy01 ~]# systemctl start haproxy && systemctl enable haproxy

1.6 配置结果确认

1)主主架构模式下的vip是否生效
[root@haproxy01 ~]# ip addr |grep ens33
2: ens33: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    inet 192.168.48.180/24 brd 192.168.48.255 scope global ens33
    inet 192.168.48.200/32 scope global ens33
[root@haproxy02 ~]# ip addr |grep ens33
2: ens33: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    inet 192.168.48.181/24 brd 192.168.48.255 scope global ens33
    inet 192.168.48.100/32 scope global ens33 
2）keepalived主备切换测试，查看vip是否漂移
[root@haproxy01 ~]# systemctl stop keepalived  
[root@haproxy02 ~]# ip addr |grep ens33
2: ens33: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    inet 192.168.48.181/24 brd 192.168.48.255 scope global ens33
    inet 192.168.48.100/32 scope global ens33
    inet 192.168.48.200/32 scope global ens33  
//再次切换
[root@haproxy02 ~]# systemctl stop keepalived.service 
[root@haproxy01 ~]# ip addr |grep ens33
2: ens33: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    inet 192.168.48.180/24 brd 192.168.48.255 scope global ens33
    inet 192.168.48.200/32 scope global ens33
    inet 192.168.48.100/32 scope global ens33 
3）haproxy停止，测试keepalive的监测服务脚本是否生效
[root@haproxy01 ~]# systemctl stop haproxy  
[root@haproxy01 ~]# while true;do systemctl is-active haproxy && sleep 1;done //循环下，为了看效果
active
inactive //已经停止
activating //然后检测到，被keepalived启动
activating
active
active 
4)haproxy测似，关闭后端nginx
[root@web01 ~]# systemctl stop nginx
C:\Users\cs>curl 192.168.48.200
this is test page
C:\Users\cs>curl 192.168.48.100
this is test page

1.7 访问测试

//客户端访问
C:\Users\cs>curl 192.168.48.100
this is test page
C:\Users\cs>curl 192.168.48.200
this is test page

haproxy监控页面访问

vip192.168.48.200访问

vip192.168.48.100访问

二、Haproxy+Keepalived主从模式的高可用环境

主从模式相对于主主模式主要区别只在于keepalived的配置不同，其他的配置都和主主模式下的一样，并且主从环境下只有一个VIP（默认在Master端配置）

1）在上面的配置中主备删除keepalived.conf的vrrp_instace VI_2

2) 重新启动keepalived会发现在主节点上面有一个vip，备节点没有

3) 当主节点故障，vip会漂移到备用节点

和主主模式大同小异这里就不做演示了。

数据库

文章转载自请叫我水哥，如果涉嫌侵权，请发送邮件至：contact@modb.pro进行举报，并提供相关证据，一经查实，墨天轮将立刻删除相关内容。

CentOS7.4+Keepalived+HAproxy主主和主从架构实验部署手册

评论