温馨提示×

温馨提示×

您好,登录后才能下订单哦!

密码登录×
登录注册×
其他方式登录
点击 登录注册 即表示同意《亿速云用户服务条款》

keepalived+nginx

发布时间:2020-06-28 20:53:15 来源:网络 阅读:635 作者:季凉末一 栏目:建站服务器


我们先来谈谈keepalived的工作原理:

keepalived的工作原理基于三个层面,分别时IP层,TCP层和应用层
IP 层:keepalived会定期向服务器集群中的服务器发送一个类似于ping程序的ICMP数据包。通俗理解就是我keepalived来ping你服 务器,我要是ping通了,那我知道你还活着,咱们可以继续合作,如果不幸发现你挂了,那就得把这台服务器踢了。该层就是确保服务器ip地址有效,一次来 判断这台服务器可否继续在岗谋职。
TCP层:这一层和IP层类似,只不过IP层检查的是IP,该层检查TCP端口。keepalived在IP层 且报你服务器的IP有效,然后继续看你的TCP端口,比如说你是web server,那你的80端口就应该开启,如果keepalived发现你的80端口没有启动。那么它就会认为你已经挂掉,然后剔除。
应用层:这一层就是具体的工作程序了,keepalived是个恪尽职守的检察官,应用层的程序与用户的设定息息相关,keepalived则通过检查服务器的运行是否和用户设定一致,如果不一致,keepalived就会宣判这台服务器牺牲。

keepalived + nginx

主机:172.25.44.3    server3
备机:172.25.44.4    server4

配置虚拟ip

[root@server3 ~]# vim /etc/sysconfig/network-scripts/ifcfg-eth0:0DEVICE=eth0:0TYPE=EthernetONBOOT=yesBOOTPROTO=staticDNS1=172.25.44.44IPADDR=172.25.44.100NETMASK=255.255.255.0GETWAY=172.25.44.44[root@server3 ~]# /etc/init.d/network restart[root@server3 ~]# ifconfigeth0      Link encap:Ethernet  HWaddr 52:54:00:48:60:3E          inet addr:172.25.44.4  Bcast:172.25.44.255  Mask:255.255.255.0          inet6 addr: fe80::5054:ff:fe48:603e/64 Scope:Link          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1          RX packets:11926 errors:0 dropped:0 overruns:0 frame:0          TX packets:6533 errors:0 dropped:0 overruns:0 carrier:0          collisions:0 txqueuelen:1000          RX bytes:25662550 (24.4 MiB)  TX bytes:876645 (856.0 KiB)
eth0:0    Link encap:Ethernet  HWaddr 52:54:00:48:60:3E          inet addr:172.25.44.100  Bcast:172.25.44.255  Mask:255.255.255.0          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
lo        Link encap:Local Loopback          inet addr:127.0.0.1  Mask:255.0.0.0          inet6 addr: ::1/128 Scope:Host          UP LOOPBACK RUNNING  MTU:16436  Metric:1          RX packets:36283 errors:0 dropped:0 overruns:0 frame:0          TX packets:36283 errors:0 dropped:0 overruns:0 carrier:0          collisions:0 txqueuelen:0          RX bytes:1982812 (1.8 MiB)  TX bytes:1982812 (1.8 MiB)
主备机均需配置

安装nginx

安装包:nginx-1.10.3.tar.gz
解压安装包:tar zxf nginx-1.10.3.tar.gz

[root@server3 software]# cd nginx-1.10.3 —prefix=/usr/local/lnmp/nignx
[root@server3 nginx-1.10.3]# ./configure
[root@server3 nginx-1.10.3]# make
[root@server3 nginx-1.10.3]# make install
[root@server3 nginx-1.10.3]# ln -s /usr/local/lnmp/nginx/sbin/nginx /usr/local/sbin/
#创建nginx的软链接


配置nginx
[root@server3 nginx-1.10.3]# cd /usr/local/lnmp/nginx/conf[root@server3 conf]# vim nginx.conf/#访问8088端口是的发布目录为/var/www/html,而nginx的默认发布目录还是/usr/local/lnmp/nginx/html
server {        listen       8088;#监听8088端口        server_name  localhost;
    location / {
        root   /var/www/html;
        index  index.html index.htm;
    }
    error_page   500 502 503 504  /50x.html;
    location = /50x.html {
        root   /usr/local/lnmp/nginx/html;
    }
}
[root@server3 conf]# cd /var/www/html[root@server3 html]# vim index.html
<html>
<head>
<title>Welcome to nginx!</title>
</head>
<body bgcolor="white" text="black">
<center><h2>Welcome to nginx! 172.25.44.3</h2></center>
</body>
</html>
[root@server3 html]# nginx

安装keepalived

安装包:keepalived-1.2.2.tar.gz

[root@server3 software]# wget 
http://www.keepalived.org/software/keepalived-1.2.2.tar.gz
[root@server3 software]# tar zxf keepalived-1.2.2.tar.gz
[root@server3 software]# cd keepalived-1.2.2
[root@server3 keepalived-1.2.2]# ./configure
[root@server3 keepalived-1.2.2]# make
[root@server3 keepalived-1.2.2]# make install
[root@server3 keepalived-1.2.2]# cp /usr/local/sbin/keepalived /usr/sbin/
[root@server3 keepalived-1.2.2]# cp /usr/local/etc/rc.d/init.d/keepalived /etc/init.d/
[root@server3 keepalived-1.2.2]# cp /usr/local/etc/sysconfig/keepalived /etc/sysconfig/
[root@server3 keepalived-1.2.2]# cp -r /usr/local/etc/keepalived/ /etc/
配置leepalived
[root@server3 keepalived-1.2.2]# vim /var/www/html/index.html
global_defs {
    router_id LVS_DEVEL
}
vrrp_script chk_http_port {
    script “</dev/tcp/127.0.0.1/8088”
    interval 1
    weight -2
}
vrrp_instance VI_1 {
    state MASTER
    interface eth0
    virtual_router_id 51
    priority 100
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {
        172.25.44.100
    }
    track_script {
        chk_http_port
    }
#state MASTER:MASTER主机,BACKUP备机#priority 100:优先级指定,数字越大优先级越高#virtual_ipaddress:指定vip#track_scrip:检查端口
[root@server3 keepalived-1.2.2]# /etc/init.d/keepalived start

备机次同步主机的操作步骤:keepalived的配置文件中,主机的state为MASTER,备机的state为BACKUP,备机的priority要小于主机

测试

验证nginx启动正常

访问主机nginx的默认页

keepalived+nginx

访问主机nginx的指定页

keepalived+nginx

访问备机nginx的默认页

keepalived+nginx

访问备机nginx的指定页

keepalived+nginx

访问vip

主机server3的优先级更高,在主机完好的情况下,有server3来负载nginx运行

keepalived+nginx

[root@server3 software]# /etc/init.d/keepalived stop
Stopping keepalived:                                       [  OK  ]
当主机的keeplived服务挂掉的时候,在则访问vip,则由备机自动负载nginx运行
[root@server4 conf]# ip a
1: lo:  mtu 16436 qdisc noqueue state UNKNOWN
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever
2: eth0:  mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 52:54:00:48:60:3e brd ff:ff:ff:ff:ff:ff
    inet 172.25.44.4/24 brd 172.25.44.255 scope global eth0
    inet 172.25.254.104/24 brd 172.25.254.255 scope global eth0
    inet 172.25.44.100/32 scope global eth0
    inet 172.25.44.100/24 brd 172.25.44.255 scope global secondary eth0:0
    inet6 fe80::5054:ff:fe48:603e/64 scope link
       valid_lft forever preferred_lft forever
同时可以看到vip浮动在备机server4上

keepalived+nginx

[root@server3 software]# /etc/init.d/keepalived start
Starting keepalived:                                       [  OK  ]
启动主机server3上的keepalived服务,再次访问vip,则nginx的负载运行又回到主机server3上
[root@server3 software]# ip a
1: lo:  mtu 16436 qdisc noqueue state UNKNOWN
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever
2: eth0:  mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 52:54:00:b4:14:0d brd ff:ff:ff:ff:ff:ff
    inet 172.25.44.3/24 brd 172.25.44.255 scope global eth0
    inet 172.25.254.103/24 brd 172.25.254.255 scope global eth0
    inet 172.25.44.100/32 scope global eth0
    inet6 fe80::5054:ff:feb4:140d/64 scope link
       valid_lft forever preferred_lft forever
该情况下查看ip,发现vip浮动在主机server3上

keepalived+nginx

在应用层出问题的情况下,如停掉主机server3的nginx服务,则备机server4也会自动接管nginx支持前端的访问
[root@server3 software]# killall nginx

keepalived+nginx

问题补充:在备机keepalived的配合配置文件中,如果补修改state的情况下实验也是可以成功的,我觉得应该时两个机子通过priority优先级竞争。


向AI问一下细节

免责声明:本站发布的内容(图片、视频和文字)以原创、转载和分享为主,文章观点不代表本网站立场,如果涉及侵权请联系站长邮箱:is@yisu.com进行举报,并提供相关证据,一经查实,将立刻删除涉嫌侵权内容。

AI