实验要求,至少三台虚拟机,在同一网段172.18/16,172.18.10.10/11,其中172.18.200.100做为DR
并且配置两块网卡,另一块网卡的地址不能再171.18/16网段内
[[email protected] ~]# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 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: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 00:0c:29:23:f3:8d brd ff:ff:ff:ff:ff:ff
inet 172.18.200.100/16 brd 172.18.255.255 scope global eth1
inet6 fe80::20c:29ff:fe23:f38d/64 scope link
valid_lft forever preferred_lft forever
3: eth2: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 00:0c:29:23:f3:97 brd ff:ff:ff:ff:ff:ff
inet 192.168.10.100/24 brd 192.168.10.255 scope global eth2
inet6 fe80::20c:29ff:fe23:f397/64 scope link
valid_lft forever preferred_lft forever
1、在10/11上使用yum安装php和httpd
[[email protected] ~]# yum install php httpd
2、配置时间同步,先安装chrony安装包,再进行配置
[[email protected] ~]# yum install -y chrony
[[email protected] ~]# vim /etc/chrony.conf
# Allow NTP client access from local network.
#allow 192.168/16
allow 172.18/16
# Serve time even if not synchronized to any NTP server.
local stratum 10
3、启动chrony服务
[[email protected] ~]# service chronyd start
Starting chronyd: [ OK ]
4、安装同步时间的守护进程,为的是每过一段时间同步时间
分别在两台Vs上安装chrony
[[email protected] ~]# yum install -y chrony
[[email protected] ~]# yum install -y chrony
编辑配置文件,两台VS机子一样的配置
[[email protected] ~]# vim /etc/chrony.conf
#server 0.rhel.pool.ntp.org iburst
#server 1.rhel.pool.ntp.org iburst
#server 2.rhel.pool.ntp.org iburst
#server 3.rhel.pool.ntp.org iburst
server 172.18.200.100 iburst
启动chrony服务,两台VS机子一样的配置
[[email protected] ~]# service chronyd start
Starting chronyd: [ OK ]
5、与服务器同步时间
[[email protected] ~]# ntpdate 172.18.200.100
10 May 02:13:57 ntpdate[3324]: step time server 172.18.200.100 offset -2587.052960 sec
[[email protected] ~]# ntpdate 172.18.200.100
10 May 02:13:57 ntpdate[3324]: step time server 172.18.200.100 offset -2587.052960 sec
[[email protected] ~]# date
Wed May 10 02:31:05 CST 2017
[[email protected] ~]# date
Wed May 10 02:31:08 CST 2017
6、在两台VS主机上提供相关主页
这里用简单的for循环快速生成10个页面测试文件
[[email protected] ~]# for i in {1..20}; do echo "Test Page $i on UpStream Server 1 (172.18.10.10)" > /var/www/html/test$i.html;done
[[email protected] html]# ls /var/www/html/
test10.html test12.html test14.html test16.html test18.html test1.html test2.html test4.html test6.html test8.html
test11.html test13.html test15.html test17.html test19.html test20.html test3.html test5.html test7.html test9.html
另一台机子做同样的操作
7、启动httpd服务(10/11)
[email protected] html]# service httpd start
8、在DR服务端测试,使用curl测试页面是否能访问
[[email protected] ~]# curl http://172.18.10.10/test1.html
Test Page 1 on UpStream Server 1 (172.18.10.10)
[[email protected] ~]# curl http://172.18.10.11/test1.html
Test Page 1 on UpStream Server 1 (172.18.10.11)
9、在DR端下载并安装nginx,由于使用centos6.8,无法安装nginx1.10以上的版本
lftp 172.18.0.1:/pub/Sources/6.x86_64/nginx> ls
-rw-r--r-- 1 500 500 714233 Jul 25 2013 nginx-1.0.15-5.el6.src.rpm
-rwxr--r-- 1 500 500 319456 Apr 24 2014 nginx-1.4.7-1.el6.ngx.x86_64.rpm
-rw-r--r-- 1 0 0 344416 Sep 16 2014 nginx-1.6.2-1.el6.ngx.x86_64.rpm
lftp 172.18.0.1:/pub/Sources/6.x86_64/nginx> mget nginx-1.6.2-1.el6.ngx.x86_64.rpm
344416 bytes transferred
lftp 172.18.0.1:/pub/Sources/6.x86_64/nginx> bye
安装nginx
[[email protected] ~]# yum install nginx-1.6.2-1.el6.ngx.x86_64.rpm
10、配置nginx,仅仅作为反向代理服务器
[[email protected] ~]# cd /etc/nginx/conf.d/
[[email protected] conf.d]# ls
default.conf example_ssl.conf
[[email protected] conf.d]# cp default.conf default.conf.bak
[[email protected] conf.d]# ls
default.conf default.conf.bak example_ssl.conf
因为要先定义组,所以首先编辑nginx.conf文件
[[email protected] nginx]# ls
conf.d fastcgi_params koi-utf koi-win mime.types nginx.conf scgi_params uwsgi_params win-utf
[[email protected] nginx]# vim nginx.conf
在http上下文中定义upstream 模块,操作如下
http {
include /etc/nginx/mime.types;
default_type application/octet-stream;
log_format main ‘$remote_addr - $remote_user [$time_local] "$request" ‘
‘$status $body_bytes_sent "$http_referer" ‘
‘"$http_user_agent" "$http_x_forwarded_for"‘;
access_log /var/log/nginx/access.log main;
sendfile on;
#tcp_nopush on;
keepalive_timeout 65;
#gzip on;
upstream websrvs {
server 172.18.10.10:80;
server 172.18.10.11:80;
}
include /etc/nginx/conf.d/*.conf;
}编辑conf.d文件中的default.conf文件
[[email protected] conf.d]# vim default.conf
在location上下文中编辑proxy_pass反向代理文件
server {
listen 80;
server_name localhost;
#charset koi8-r;
#access_log /var/log/nginx/log/host.access.log main;
location / {
proxy_pass http://websrvs;
root /usr/share/nginx/html;
index index.html index.htm;
nginx -t 检查配置文件语法
10、启动nginx服务
[[email protected] conf.d]# nginx
查看监听端口80是否开启
[[email protected] conf.d]# ss -tnl
State Recv-Q Send-Q Local Address:Port Peer Address:Port
LISTEN 0 128 *:80 *:*
LISTEN 0 128 :::22 :::*
LISTEN 0 128 *:22 *:*
LISTEN 0 100 ::1:25 :::*
LISTEN 0 100 127.0.0.1:25
11、在客户端上使用curl命令一次性访问10次页面,看看效果
[[email protected] ~]# for ((i=1;i<=10;i++));do curl http://172.18.200.100:80/test1.html; done
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 2 (172.18.10.11)
发现实现负载均衡
12、若期望负载权限有所不同,则进行如下配置
[[email protected] conf.d]# cd ..
[[email protected] nginx]# vim nginx.conf
#gzip on;
upstream websrvs {
server 172.18.10.10:80 weigth=2;
server 172.18.10.11:80 weight=3; #### 这里权重:综合2+3,意为第一台服务器虚拟为2台,第二台服务器虚拟为3台
在upstream 模块里增加权重设置
保存退出,重启服务
[[email protected] nginx]# nginx -t
nginx: the configuration file /etc/nginx/nginx.conf syntax is ok
nginx: configuration file /etc/nginx/nginx.conf test is successful
[[email protected] nginx]# nginx -s reload
13、客户端再次访问10次效果如下
[[email protected] ~]# for ((i=1;i<=10;i++));do curl http://172.18.200.100:80/test1.html; done
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 2 (172.18.10.11)
14、关闭其中一台VS,再次测试
[[email protected] ~]# service httpd stop
Stopping httpd: [ OK ]
[[email protected] ~]# for ((i=1;i<=10;i++));do curl http://172.18.200.100:80/test1.html; done
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
再次启动VS,在测试
[[email protected] ~]# service httpd start
[[email protected] ~]# for ((i=1;i<=10;i++));do curl http://172.18.200.100:80/test1.html; done
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 1 (172.18.10.10)
15、定义最大失败次数max_fails ,和超时时间间隔 fail_timeout
[[email protected] nginx]# vim nginx.conf
upstream websrvs {
server 172.18.10.10:80 weight=2 max_fails=2 fail_timeout=2;
server 172.18.10.11:80 weight=3 backup;
保存退出并重启服务
在客户端测试
[[email protected] ~]# for ((i=1;i<=10;i++));do curl http://172.18.200.100:80/test1.html; done
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
再次修改配置
[[email protected] ~]# for ((i=1;i<=10;i++));do curl http://172.18.200.100:80/test1.html; done
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 2 (172.18.10.11)
因此发现nginx默认调度规则为加权轮询wrr
16、实现ip_hash(ip哈希绑定)绑定,第一次访问哪个主机,随后都会访问该主机
vim nginx.conf
upstream websrvs {
ip_hash;
server 172.18.10.10:80 weight=2 max_fails=2 fail_timeout=2;
server 172.18.10.11:80 weight=3;
[[email protected] ~]# for ((i=1;i<=10;i++));do curl http://172.18.200.100:80/test1.html; done
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 1 (172.18.10.10)
17、实现least_conn(加权最少连接),相当于wlc,最少连接,只要weight不同就会考虑权重
upstream websrvs {
least_conn;
server 172.18.10.10:80 weight=2 max_fails=2 fail_timeout=2;
server 172.18.10.11:80 weight=3;
}
[[email protected] ~]# for ((i=1;i<=10;i++));do curl http://172.18.200.100:80/test1.html; done
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 2 (172.18.10.11)
Test Page 1 on UpStream Server 1 (172.18.10.10)
Test Page 1 on UpStream Server 2 (172.18.10.11)
18.哈希 hash key ,key可以是任何表示,url,地址等等,hash后跟什么就绑定什么
upstream websrvs {
hash $request;
server 172.18.10.10:80 weight=2 max_fails=2 fail_timeout=2;
server 172.18.10.11:80 weight=3;
}
第一次访问那个资源在哪台服务器,之后就绑定再该服务器上
upstream websrvs {
hash $request consistent; #####一致性哈希算法
server 172.18.10.10:80 weight=2 max_fails=2 fail_timeout=2;
server 172.18.10.11:80 weight=3;
}
19、keepalive connections
并发太高,给端口带来压力,为此避免这种情况
可以在后面这一侧,使用长连接
nginx的特点,不是每一个请求靠每一个进程来响应,
而是一个进程响应n个请求
因此大量请求都是由worker响应的
worker 4个
现在用一个来表示
保持一定数量的长连接
一个长连接只能响应一个请求
与第一个server保留32个,意味着一次性可以发出32个长连接,也不用建立新连接
与第二个server保留32个长连接,也是一样,不用建立新连接,一共就是64个
所以连接一直在,也不用三次握手和四次断开
这时候,worker在nginx端占用的端口数量也就不变了
因此,32个长连接连接着,大量的请求就可以通过这些长连接发送
不用每个连接都建立一个端口
可以极大地节约端口
keepalive connections 表示保留的空闲长连接有多少个
假如没人用这些长连接了,一直建立这些长连接也是对后端服务器的压力
解决办法就是超时关闭,最少保持一个关闭
后面访问就直接可以用了,不用再去建立长连接
keepalive connections ,因此将该设置数字调高一些,可以一定意义上提升服务器性能