1. 首页
  2. 负载均衡
  3. nginx+keepalived实现nginx双主高可用的负载均衡

原  nginx+keepalived实现nginx双主高可用的负载均衡

 王同学  负载均衡  2016-03-16  7059  发表评论

一、前言：

在互联网上面，网站为用户提供原始的内容访问，同时为用户提供交互操作。提供稳定可靠的服务，可以给用户带来良好的体验，保证用户的正常访问使用，在网站的可靠性方面，有很多的技术可以来提供，这类技术可以分为：

高可用：保证服务的可靠，稳定，实现故障的屏蔽，避免了单点故障。

高性能：多台服务器连接起来，处理一个复杂的计算问题。

负载均衡：将用户请求引导到后端多台服务器，实现服务器请求的负载。

我们将这类技术称之为集群负载均衡，可以提供负载均衡和高可用的有硬件和软件，软件方面有haproxy，lvs，keepalived，nginx，heartbeat，corosync等等，而这里我们采用的是nginx-keepalived来构建。

Nginx有很强的代理功能，但是一台nginx 就形成了单点，现在使用keepalived来解决这个问题，keepalived可以实现故障转移切换，实现后端的健康检查，前端的高可用，使网站故障记录大大降低，避免了单点故障造成网站无法访问的问题，确保了网站业务的正常运行。

二、Nginx+keepalived有两种配置方案：

2.1、Nginx+keepalived 主从配置

这种方案，使用一个vip地址，前端使用2台机器，一台做主，一台做备，但同时只有一台机器工作，另一台备份机器在主机器不出现故障的时候，永远处于浪费状态，对于服务器不多的网站，该方案不经济实惠，所以本次不予采用。

2.2、Nginx+keepalived 双主配置

这种方案，使用两个vip地址，前端使用2台机器，互为主备，同时有两台机器工作，当其中一台机器出现故障，两台机器的请求转移到一台机器负担，非常适合于当前架构环境，故本次采用此方案对网站进行高可用架构。

三、Nginx+keepalived 主从配置

3.1、Nginx+keepalived 主从配置详情请见http://kling.blog.51cto.com/3320545/1240359

这里不做重点介绍。

四、Ningx+Keepalived 双主配置

4.1、拓扑结构

4.2、测试环境如下：

系统：Ceentos 6.4 64位

前端node1服务器：

DIP: 192.168.122.2

VIP: 192.168.122.22

前端node2服务器：

DIP: 192.168.122.3

VIP：192.168.122.23

后端服务器：

web server01：192.168.122.4

web server02：192.168.122.5

web server03：192.168.122.6

4.3、软件安装

分别在两台前端服务器上安装nginx+keepalived，使用脚本如下：

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
#!/bin/bash
# author: kuangl
# mail: kuangl@orient-media.com
# description: The installation of Nginx files.
# -------------------------------------------------------- #
         ## Nginx_install
# -------------------------------------------------------- #
# Nginx installation
#CURRENT_PATH=$(pwd)
for i in $(rpm -q gcc gcc-c++ kernel-devel openssl-devel zlib-devel popt-devel popt-static libnl-devel wget make |grep 'not installed' | awk '{print $2}')
do
    yum -y install $i
done
[ -d /root/software ]
[ "$?" != 0 ] && mkdir /root/software
cd /root/software
[ !  -e pcre-8.33.tar.gz ] && wget ftp://ftp.csx.cam.ac.uk/pub/software/programming/pcre/pcre-8.33.tar.gz
tar -zxvf pcre-8.33.tar.gz
cd pcre-8.33
./configure
make && make install
echo $? || [ $? != 0  ] || echo  " installation pcre  failed" || exit 1
cd /root/software
[ ! -e nginx-1.2.9.tar.gz ] && wget http://nginx.org/download/nginx-1.2.9.tar.gz
tar -zxvf nginx-1.2.9.tar.gz
cd nginx-1.2.9
./configure  --prefix=/usr/local/nginx --with-http_ssl_module --with-http_sub_module --with-http_stub_status_module  --with-http_gzip_static_module
make && make install
echo $? || [ $? != 0  ] || echo  " installation  nginx  failed" || exit 1
# -------------------------------------------------------- #
            ## Keepalived_intsall
# -------------------------------------------------------- #
# Keepalived installation
cd /root/softwarae
[ ! -e keepalived-1.2.4.tar.gz ] &&  wget http://www.keepalived.org/software/keepalived-1.2.4.tar.gz
tar -zxvf keepalived-1.2.4.tar.gz
cd keepalived-1.2.4
ln -s /usr/src/kernels/$(uname -r) /usr/src/kernels/linux
./configure --prefix=/usr  --bindir=/usr/bin  --sbindir=/usr/bin  --libexecdir=/usr/libexec --localstatedir=/var --libdir=/lib64  --infodir=/usr/share/info  --sysconfdir=/etc --mandir=/usr/local/share/man   --with-kernel-dir=/usr/src/kernels/linux
make && make install
echo $? || [ $? != 0  ] || print " installation keepalived  failed" || exit 1
chkconfig --add keepalived
chkconfig --level 345 keepalived on

4.4、在后端服务器上安装apached

    后端node4

1
2
3
4
5
[root@node4 ~]# yum  -y install httpd
[root@node4 html]# echo "this is 192.168.122.4" > /var/www/htmlindex.html
[root@node4 ~]# service httpd start
[root@node4 html]# curl  192.168.122.4
this is 192.168.122.4

  后端node5

1
2
3
4
5
[root@node5 ~]# yum  -y install httpd
[root@node5 html]# echo "this is 192.168.122.5" > /var/www/htmlindex.html
[root@node5 ~]# service httpd start
[root@node5 html]# curl  192.168.122.5
this is 192.168.122.5

  后端node6

1
2
3
4
5
[root@node6 ~]# yum  -y install httpd
[root@node6 html]# echo "this is 192.168.122.6" > /var/www/htmlindex.html
[root@node6 ~]# service httpd start
[root@node6 html]# curl  192.168.122.6
this is 192.168.122.6

4.5、node2、node3上配置nginx

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
[root@node2 ~]# vim /usr/local/nginx/conf/nginx.conf
upstream web1       ##定义负载均衡组为web1
    {
        ip_hash;
        server 192.168.122.4:80;
        server 192.168.122.5:80;
        server 192.168.122.6:80;
    }
 server {
        listen       80;
        server_name  dev.test01.com;
        location /
        {
        root /home/kuangl/;
        index index.html index.htm;
        proxy_set_header Host $host;
        proxy_set_header X-Forwarded-For $remote_addr;
        proxy_pass http://web1;
        }
      }

4.6、在node2上配置keepalived

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
[root@node2 conf]# vim /etc/keepalived/keepalived.conf
! Configuration File for keepalived
global_defs {
   notification_email {
     404060945@qq.com
   }
   notification_email_from root@localhost
   smtp_server 127.0.0.1
   smtp_connect_timeout 30
   router_id LVS_DEVEL
}
vrrp_script chk_haproxy {
    script "/etc/keepalived/chk_nginx.sh"
    interval 2
    weight 2
}
vrrp_instance VI_1 {
    state MASTER
    interface eth0
    virtual_router_id 200
    priority 250
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass kuanglnginx
    }
   track_script {
        chk_nginx
    }
    virtual_ipaddress {
        192.168.122.22
    }
}
vrrp_instance VI_2 {
    state BACKUP
    interface eth0
    virtual_router_id 251
    priority 100
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass kuangl
    }
    track_script {
        chk_nginx
    }
    virtual_ipaddress {
        192.168.122.23
    }
}

4.7、在node3上配置keepalived

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
! Configuration File for keepalived
global_defs {
   notification_email {
     404060945@qq.com
   }
   notification_email_from root@localhost
   smtp_server 127.0.0.1
   smtp_connect_timeout 30
   router_id LVS_DEVEL
}
vrrp_script chk_haproxy {
    script "/etc/keepalived/chk_nginx.sh"
    interval 2
    weight 2
}
vrrp_instance VI_1 {
    state BACKUP
    interface eth0
    virtual_router_id 200
    priority 100
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass kuanglnginx
    }
    track_script {
        chk_nginx
    }
    virtual_ipaddress {
        192.168.122.22
    }
}
vrrp_instance VI_2 {
    state MASTER
    interface eth0
    virtual_router_id 251
    priority 250
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass kuangl
    }
    track_script {
        chk_nginx
    }
    virtual_ipaddress {
        192.168.122.23
    }
}

4.8、在两台双主服务器上添加自动检测脚本

1
2
3
4
5
6
7
8
9
10
11
12
#!/bin/bash
# description:
# 定时查看nginx是否存在，如果不存在则启动nginx
# 如果启动失败，则停止keepalived
status=$(ps -C nginx --no-heading|wc -l)
if [ "${status}" = "0" ]; then
        /usr/local/nginx/sbin/nginx
        status2=$(ps -C nginx --no-heading|wc -l)
        if [ "${status2}" = "0"  ]; then
                /etc/init.d/keepalived stop
        fi
fi

4.9、开启nginx、keepalived服务

1
2
3
4
[root@node2 ~]# service keepalived start
[root@node2 ~]# /usr/local/nginx/sbin/nginx
[root@node3 ~]# service keepalived start
[root@node3 ~]# /usr/local/nginx/sbin/nginx

4.10、用 ip a 查看VIP

4.11、测试访问

1
2
3
4
5
6
7
8
9
10
11
12
[kuangl@node01 ~]$ curl http://192.168.122.22
this is 192.168.122.6
[kuangl@node01 ~]$ curl http://192.168.122.22
this is 192.168.122.4
[kuangl@node01 ~]$ curl http://192.168.122.22
this is 192.168.122.5
[kuangl@node01 ~]$ curl http://192.168.122.23
this is 192.168.122.6
[kuangl@node01 ~]$ curl http://192.168.122.23
this is 192.168.122.4
[kuangl@node01 ~]$ curl http://192.168.122.23
this is 192.168.122.5

五、后端用rsync做数据同步

  node5-node6上配置进程模式，以node5为例

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
[root@node5 ~]# yum -y install rsync
[root@node5 ~]# vim /etc/rsynsd.conf
uid = root
gid = root
use chroot = no
max connections = 5
pid file = /var/run/rsyncd.pid
lock file = /var/run/rsync.lock
log file = /var/log/rsyncd.log
[web01]                      
path=/home/kuangl/          
comment = update          
ignore errors              
read only = no             
list = no                 
hosts allow = 192.168.122.0/24
auth users = root        
uid = root
gid = root
secrets file = /etc/rsyncd.secrets
[root@node5 ~]# vim /etc/rsyncd.secrets
root:123456
[root@node5 ~]# chmod 0600 /etc/rsyncd.secrets
[root@node5 ~]# ll /etc/rsyncd.secrets
-rw-------. 1 root root 12 Jul 20 19:41 /etc/rsyncd.secrets
[root@node5 ~]# rsync --daemon
[root@node5 ~]# echo "rsync --daemon" >> /etc/rc.local

 node4上配置命令模式：

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
[root@node4 ~]# yum -y install rsync
[root@node4 ~]# vim /etc/rsyncd.secrets
123456
[root@node4 ~]# chmod 0600 /etc/rsyncd.secrets
root@node4 kuangl]# rsync -vzrtopg --delete --progress --password-file=/etc/rsyncd.secrets  rsync+inotify root@192.168.122.5::web01
sending incremental file list
rsync+inotify/
rsync+inotify/inotify-tools-3.14.tar.gz
      358772 100%    1.85MB/s    0:00:00 (xfer#1, to-check=2/4)
rsync+inotify/rsync+inotify_client.sh
         617 100%    3.11kB/s    0:00:00 (xfer#2, to-check=1/4)
rsync+inotify/rsync+inotify_server.sh
         900 100%    4.03kB/s    0:00:00 (xfer#3, to-check=0/4)
sent 360679 bytes  received 69 bytes  240498.67 bytes/sec
total size is 360289  speedup is 1.00

 查看结果

1
2
3
4
5
[root@node5 ~]# cd /home/kuangl/
[root@node5 kuangl]# ll
total 8
-rw-r--r--. 1 root root   22 Jul 20 15:16 index.html
drwxr-xr-x. 2 root root 4096 Nov 11  2012 rsync+inotify

本文出自 “&思远晨曦” 博客，请务必保留此出处http://kling.blog.51cto.com/3320545/1253474
所有评论
加载评论 ...
发表评论
匠心蓝  爱情红  清新绿  难看蓝 

© 2026  LINK ONE BLOG - 已稳定运行11年  闽ICP备15004025号

用户登录

用户注册