目录:
1.安装nginx
2.配置nginx
3.调优nginx
4.性能测试
ps:为了方便，文档使用docker容器来操作的。
1.安装nginx
1.1 启动容器、download nginx 以及编译前的修改
启动容器
liwangdeMacBook-Air:~ liwang$ docker run -i -t --name nginx_server_01 -v /Users/liwang/docker/nginx_data:/data -p : centos /bin/bash
将文件copy至容器中
liwangdeMacBook-Air:~ liwang$ docker ..tar.gz
解压
[root@e46ae471064e soft]# .tar.gz -C /soft/
安装之前修改源码屏蔽名称以及版本号
[root@e46ae471064e core]# ls -l "/soft/nginx-1.14.0/src/core/nginx.h" -rw-r--r-- Apr : /soft/nginx-/src/core/nginx.h [root@e46ae471064e core]# ------------------------------------------------------ #define NGINX_VERSION "0.0.1" #define NGINX_VER "HaiYan/" NGINX_VERSION #define NGINX_VAR "HaiYan" ------------------------------------------------------ #define NGINX_VERSION "1.14.0" #define NGINX_VER "nginx/" NGINX_VERSION #define NGINX_VAR "NGINX" ------------------------------------------------------
1.2 安装nginx
安装插件
[root@e46ae471064e ~]# yum install pcre pcre-devel gcc gcc-c++ make openssl openssl-devel -y
添加用户
[root@e46ae471064e nginx-]# useradd nginx -s /sbin/nologin -M
安装
[root@e46ae471064e nginx-]# ./configure --prefix=/usr/local/nginx --with-http_ssl_module --with-http_stub_status_module --user=nginx --group=nginx [root@e46ae471064e nginx-]# make [root@e46ae471064e nginx-]# make install
启动nginx,利用curl工具获取服务器信息
[root@e46ae471064e conf]# /usr/local/nginx/sbin/nginx [root@e46ae471064e nginx]# curl -I localhost HTTP/ OK Server: HaiYan/ Date: Tue, May :: GMT Content-Type: text/html Content-Length: Last-Modified: Tue, May :: GMT Connection: keep-alive ETag: "5ae87ab6-264" Accept-Ranges: bytes [root@e46ae471064e nginx]#
详细可安装信息可用./configure --help获取
2.配置nginx
1、nginx虚拟主机配置
基于域名的虚拟主机
[root@e46ae471064e conf]# cat nginx.conf | grep -v "#" | grep -v "^$" >> extra/www.conf
修改www.conf效果如下:
[root@e46ae471064e extra]# cat www.conf
server {
listen ;
server_name www.wang-li.top;
location / {
root html/www;
index index.html index.htm;
}
error_page /50x.html;
location = /50x.html {
root html;
}
}
[root@e46ae471064e extra]#
修改nginx.conf 在http区间增加,include extra/*; 这行
[root@e46ae471064e conf]# cat nginx.conf | grep -v "#" | grep -v "^$"
worker_processes ;
events {
worker_connections ;
}
http {
include mime.types;
default_type application/octet-stream;
sendfile on;
keepalive_timeout ;
server {
listen ;
server_name localhost;
location / {
root html;
index index.html index.htm;
}
error_page /50x.html;
location = /50x.html {
root html;
}
}
include extra/*;
}
[root@e46ae471064e conf]#
增加html/www/index.html文件
[root@e46ae471064e conf]# ls /usr/local/nginx/html/www/index.html -l -rw-r--r-- root root May : /usr/local/nginx/html/www/index.html [root@e46ae471064e conf]#
修改hosts文件，访问信息如下
[root@e46ae471064e conf]# cat /etc/hosts | grep www.wang-li.top 127.0.0.1 www.wang-li.top [root@e46ae471064e conf]# curl -I www.wang-li.top HTTP/ OK Server: HaiYan/ Date: Tue, May :: GMT Content-Type: text/html Content-Length: Last-Modified: Tue, May :: GMT Connection: keep-alive ETag: "5ae87ead-265" Accept-Ranges: bytes [root@e46ae471064e conf]#
基于端口的虚拟主机
修改www.conf基于端口的访问，如下:
[root@e46ae471064e extra]# cat www.conf
server {
listen ;
server_name www.wang-li.top;
location / {
root html/www;
index index.html index.htm;
}
error_page /50x.html;
location = /50x.html {
root html;
}
}
[root@e46ae471064e extra]#
reload nginx 访问如下
[root@e46ae471064e extra]# /usr/local/nginx/sbin/nginx -t nginx: the configuration file /usr/local/nginx/conf/nginx.conf syntax is ok nginx: configuration file /usr/local/nginx/conf/nginx.conf test is successful [root@e46ae471064e extra]# /usr/local/nginx/sbin/nginx -s reload [root@e46ae471064e extra]# curl -I www.wang-li.top: HTTP/ OK Server: HaiYan/ Date: Tue, May :: GMT Content-Type: text/html Content-Length: Last-Modified: Tue, May :: GMT Connection: keep-alive ETag: "5ae87ead-265" Accept-Ranges: bytes [root@e46ae471064e extra]#
基于IP的虚拟主机
获取容器ip
liwangdeMacBook-Air:~ liwang$ docker inspect -f {{.NetworkSettings.IPAddress}} nginx_server_01
172.17.0.2
liwangdeMacBook-Air:~ liwang$
修改www.conf如下
[root@e46ae471064e extra]# cat www.conf
server {
listen ;
server_name www.wang-li.top;
location / {
root html/www;
index index.html index.htm;
}
error_page /50x.html;
location = /50x.html {
root html;
}
}
[root@e46ae471064e extra]#
reload后访问如下:
[root@e46ae471064e extra]# curl -I HTTP/ OK Server: HaiYan/ Date: Tue, May :: GMT Content-Type: text/html Content-Length: Last-Modified: Tue, May :: GMT Connection: keep-alive ETag: "5ae87ead-265" Accept-Ranges: bytes [root@e46ae471064e extra]#
2、nginx stub_status配置
查看系统是否支持此模块
[root@e46ae471064e extra]# /usr/local/nginx/sbin/nginx -V nginx version: HaiYan/ built by (Red Hat -) (GCC) built with OpenSSL Jan TLS SNI support enabled configure arguments: --prefix=/usr/local/nginx --with-http_ssl_module --with-http_stub_status_module --user=nginx --group=nginx [root@e46ae471064e extra]#
发现有--with-http_stub_status_module就是支持的
配置www.conf如下:
[root@e46ae471064e extra]# cat www.conf
server {
listen ;
server_name www.wang-li.top;
location / {
# root html/www;
# index index.html index.htm;
stub_status on;
access_log off;
}
}
[root@e46ae471064e extra]#
reload后访问如下
root@e46ae471064e extra]# curl www.wang-li.top Active connections: server accepts handled requests Reading: Writing: Waiting: [root@e46ae471064e extra]#
三个值分别代表：
active connections:正在处理的连接活动数
21 21 20
第一个表示从启动到现在一共处理了21次请求
第二个表示建立了21次握手
第三个表示总共处理了20次请求
丢包数 = 握手数-连接数 可见，并未丢包
reading: nginx 读取到客户端的header信息数
writing: Nginx 返回给客户端的header信息数
waiting: Nginx 处理完等待下一次请求指令的驻留连接，在开启keep-alive的情况下，这个值等于active - (reading + writing)
3、nginx_location:
作用：根据不同的URI来执行不同的应用
语法:
location [ = | ~ | ~* | ^- ] url {
...
}
语法规则： location [=|~|~*|^~] /uri/ { … }
= 开头表示精确匹配
^~ 开头表示uri以某个常规字符串开头，理解为匹配 url路径即可。nginx不对url做编码，因此请求为/static/20%/aa，可以被规则^~ /static/ /aa匹配到（注意是空格）。
~ 开头表示区分大小写的正则匹配
~* 开头表示不区分大小写的正则匹配
!~和!~*分别为区分大小写不匹配及不区分大小写不匹配 的正则
/ 通用匹配，任何请求都会匹配到。
~和~*的区别: ~区分大小写， ~* 不区分大小写
= :为精确配置
/ :为默认配置
案例:
利用location来实现根据URI来访问不同的内容，实现功能:访问www.wang-li.top/bbs/XXX时，能够抓取和www.wang-li.top不同的位置
配置如下
[root@e46ae471064e conf]# sed -n "47,59p" nginx.conf
location ^~ /bbs/ {
root html/bbs;
index index.html;
#
}
location / {
root html;
index index.html index.htm;
#
}
[root@e46ae471064e conf]# tree /usr/local/nginx/html/
/usr/local/nginx/html/
|-- 50x.html
|-- bbs
| `-- bbs
| `-- index.html
|-- index.html
`-- www
`-- index.html
directories, files
[root@e46ae471064e conf]#
访问如下:
[root@e46ae471064e conf]# curl localhost
<!DOCTYPE html>
<html>
<head>
<title>Welcome to nginx!</title>
<style>
body {
width: 35em;
margin: auto;
font-family: Tahoma, Verdana, Arial, sans-serif;
}
</style>
</head>
<body>
<h1>Welcome to nginx!</h1>
<p>If you see this page, the nginx web server is successfully installed and
working. Further configuration is required.</p>
<p>For online documentation and support please refer to
<a href="http://nginx.org/">nginx.org</a>.<br/>
Commercial support is available at
<a href="http://nginx.com/">nginx.com</a>.</p>
<p><em>Thank you for using nginx.</em></p>
</body>
</html>
[root@e46ae471064e conf]# curl localhost/bbs/
Welcome www.wang-li.top PPS
[root@e46ae471064e conf]#
4、nginx alias 和 root的区别
基于上诉的nginx location,location 和 alias的差别就在于
root会带入上诉的url路径去判断，而alias则不会，举例:
location ^~ /bbs/ {
root html/bbs;
index index.html;
#
}
这段说明:当我的url中包含bbs时，会去执行location的内容，而内容包含root html/bbs,现在如果输入localhost\bbs\时，系统会去检查本地文件html/bbs/bbs/...这下面的内容
而alias则不会
修改如下:
[root@e46ae471064e bbs]# sed -n "48,58p" /usr/local/nginx/conf/nginx.conf
location ^~ /bbs/ {
alias html/bbs/;
index index.html;
#
}
location / {
root html;
index index.html index.htm;
#
}
[root@e46ae471064e bbs]# cat /usr/local/nginx/html/bbs/index.html
alias liwag bbs site
[root@e46ae471064e bbs]#
[root@e46ae471064e bbs]# curl localhost/bbs/
alias liwag bbs site
[root@e46ae471064e bbs]#
5、nginx rewrite:
语法:
rewrite ^/(.*) url/$1 permanent;
意思是只要匹配如下^/(.*)，就跳转至url上，且$1是取前面regex部分括号的内容，结尾的primanent是永久301重定向标记。
regex语法:
\：去掉特殊字符
^：起始位置
$：结束位置
*：匹配0次或多次
+：匹配前面的字符一次或多次
?:匹配前面的字符0次或1次
.:匹配\n之外的任何字符
(pattern):匹配括号内字符，并可以在后面匹配，常用$0-$9属性获取值
rewrite flag标记
last:本条规则匹配完成后，继续向下匹配
break:本条规则匹配完成后，不再匹配后面的规则
redirect:返回302临时重定向，
permanent:返回301永久重定向
例子,为实现需求，访问 网址/blog 时跳转至www.cnblogs.com:
[root@e46ae471064e conf]# sed -n "33,39p" nginx.conf
#gzip on;
server {
listen ;
server_name localhost;
rewrite ^/(.*)/blog http://www.cnblogs.com/$1 permanent;
[root@e46ae471064e conf]# curl -I localhost/wang-li/blog
HTTP/ Moved Permanently
Server: HaiYan/
Date: Fri, May :: GMT
Content-Type: text/html
Content-Length:
Connection: keep-alive
Location: http://www.cnblogs.com/wang-li
[root@e46ae471064e conf]#
3.nginx调优
1.修改默认用户，有两种方式其一是修改nginx.conf中的内容，但是事先用户必须在系统中存在，其二为编译时就指定用户和用户组
其一：
[root@e46ae471064e conf]# grep "nginx" /etc/passwd nginx:x::::/home/nginx:/sbin/nologin [root@e46ae471064e conf]# sed -n "3,6p" nginx.conf user nginx nginx; worker_processes ; #error_log logs/error.log; [root@e46ae471064e conf]#
其二:
[root@e46ae471064e conf]# /usr/local/nginx/sbin/nginx -V nginx version: HaiYan/ built by (Red Hat -) (GCC) built with OpenSSL Jan TLS SNI support enabled configure arguments: --prefix=/usr/local/nginx --with-http_ssl_module --with-http_stub_status_module --user=nginx --group=nginx [root@e46ae471064e conf]#
在编译时指定user和group
2.设置nginx worker cpu核数和cpu affinity
设置nginx worker cpu核数一般设置为系统CPU核数或者是其倍数
查看系统cpu核数
[root@e46ae471064e conf]# cat /proc/cpuinfo | grep processor | wc -l [root@e46ae471064e conf]#
修改后nginx worker为:
[root@e46ae471064e conf]# sed -n "3,6p" nginx.conf worker_processes ; #error_log logs/error.log; #error_log logs/error.log notice; [root@e46ae471064e conf]#
cpu affinity([əˈfɪnɪti])的目的是为了设置cpu分布,让其均匀分布在各个cpu上
配置如下:
[root@e46ae471064e conf]# sed -n "3,5p" nginx.conf worker_processes ; worker_cpu_affinity ; [root@e46ae471064e conf]#
配置之后top命令如下:
[root@e46ae471064e conf]# top -n
top - :: up :, users, load average: 0.00, 0.00, 0.00
Tasks: total, running, sleeping, stopped, zombie
%Cpu(s): 3.2 us, 0.0 sy, 0.0 ni, 96.8 id, 0.0 wa, 0.0 hi, 0.0 si, 0.0 st
KiB Mem : total, used, buff/cache
KiB Swap: total, used. avail Mem
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
root S :00.56 bash
root S :00.01 nginx
nginx S :00.00 nginx
nginx S :00.00 nginx
nginx S :00.00 nginx
nginx S :00.00 nginx
root R :00.00 top
3.调整nginx 工作模式为epoll，来处理高并发事件
[root@e46ae471064e conf]# sed -n "12,17p" nginx.conf
events {
use epoll;
worker_connections ;
}
[root@e46ae471064e conf]#
4.设置nginx worker进程最大打开文件描述符数目,最好还是与系统ulimit保持一致
[root@e46ae471064e conf]# ulimit -n [root@e46ae471064e conf]#
配置conf如下:
[root@e46ae471064e conf]# sed -n "3,6p" nginx.conf worker_processes ; worker_cpu_affinity ; worker_rlimit_nofile ; [root@e46ae471064e conf]#
5.配置nginx最多的连接数
配置如下:
[root@e46ae471064e conf]# sed -n "10,20p" nginx.conf
#pid logs/nginx.pid;
events {
use epoll;
worker_connections ;
}
http {
[root@e46ae471064e conf]#
6.优化超时连接时间
当超过设定时间还未连接成功的话，则会断开
[root@e46ae471064e conf]# sed -n "30,35p" nginx.conf
sendfile on;
#tcp_nopush on;
#keepalive_timeout ;
keepalive_timeout ;
[root@e46ae471064e conf]#
7.优化客户端请求头超时时间以及请求body超时时间
此设置主要是为了防止客户端利用http进行攻击,如果在设置时间内没有发送完整的header或则body,则会返回request time out错误
[root@e46ae471064e conf]# sed -n "34,39p" nginx.conf
keepalive_timeout ;
client_header_timeout ;
client_body_timeout ;
#gzip on;
[root@e46ae471064e conf]#
8.优化客户端超时时间
如果超过设置时间，客户端还未有任何动作，nginx则会断掉此连接
[root@e46ae471064e conf]# sed -n "34,39p" nginx.conf
keepalive_timeout ;
client_header_timeout ;
client_body_timeout ;
send_timeout ;
[root@e46ae471064e conf]#
附上nginx.conf如下(仅供参考)
[root@e46ae471064e conf]# cat nginx.conf | grep -v "^$" | grep -v "#"
worker_processes ;
worker_cpu_affinity ;
worker_rlimit_nofile ;
events {
use epoll;
worker_connections ;
}
http {
include mime.types;
default_type application/octet-stream;
sendfile on;
keepalive_timeout ;
client_header_timeout ;
client_body_timeout ;
send_timeout ;
server {
listen ;
server_name localhost;
rewrite ^/(.*)/blog http://www.cnblogs.com/$1 permanent;
location ^~ /bbs/ {
alias html/bbs/;
index index.html;
}
location / {
root html;
index index.html index.htm;
}
}}
[root@e46ae471064e conf]#
4.性能测试
利用ab压力测试工具进行测试结果如下:
[root@e46ae471064e conf]# ab -c -n localhost/
参数说明，-c 为并发数 -n 为总请求数 总请求数需要大于并发数
This is ApacheBench, Version $>
Copyright Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Licensed to The Apache Software Foundation, http://www.apache.org/
上述介绍ab工具版本等等信息
Benchmarking localhost (be patient)
Completed requests
Completed requests
Completed requests
Completed requests
Completed requests
Completed requests
Completed requests
Completed requests
Completed requests
Completed requests
Finished requests
Server Software: HaiYan/
Server Hostname: localhost
Server Port:
show 出服务器信息以及访问信息端口等
Document Path: /
Document Length: bytes
请求路劲以及文件大小
Concurrency Level:
Time taken for tests: 1.024 seconds
整个测试持续的时间
Complete requests:
完成的请求数量
Failed requests:
失败数
Write errors:
Total transferred: bytes
HTML transferred: bytes
Requests per second: 9769.13 [#/sec] (mean)
平均返回数据时间
Time per request: 1.024 [ms] (mean)
平均响应时间
Time per request: 0.102 [ms] (mean, across all concurrent requests)
平均并发响应时间
Transfer rate: 500944.38 [Kbytes/sec] received
Connection Times (ms)
min mean[+/-sd] median max
Connect:
Processing:
Waiting:
Total:
响应时间的值
Percentage of the requests served within a certain time (ms)
%
%
%
%
%
%
%
%
% (longest request)
请求的平均速度
[root@e46ae471064e conf]#
2.测试未调优之前的nginx
[root@e46ae471064e conf]# ab -c -n localhost/
This is ApacheBench, Version $>
Copyright Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Licensed to The Apache Software Foundation, http://www.apache.org/
Benchmarking localhost (be patient)
Completed requests
Completed requests
Completed requests
Completed requests
Completed requests
Completed requests
Completed requests
Completed requests
Completed requests
Completed requests
Finished requests
Server Software: HaiYan/
Server Hostname: localhost
Server Port:
Document Path: /
Document Length: bytes
Concurrency Level:
Time taken for tests: 3.910 seconds
Complete requests:
Failed requests:
(Connect: , Receive: , Length: , Exceptions: )
Write errors:
Total transferred: bytes
HTML transferred: bytes
Requests per second: 7672.50 [#/sec] (mean)
Time per request: 260.671 [ms] (mean)
Time per request: 0.130 [ms] (mean, across all concurrent requests)
Transfer rate: 379256.45 [Kbytes/sec] received
Connection Times (ms)
min mean[+/-sd] median max
Connect:
Processing:
Waiting:
Total:
Percentage of the requests served within a certain time (ms)
%
%
%
%
%
%
%
%
% (longest request)
[root@e46ae471064e conf]#
3.测试调优之后的nginx
[root@e46ae471064e conf]# ab -c -n localhost/
This is ApacheBench, Version $>
Copyright Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Licensed to The Apache Software Foundation, http://www.apache.org/
Benchmarking localhost (be patient)
Completed requests
Completed requests
Completed requests
Completed requests
Completed requests
Completed requests
Completed requests
Completed requests
Completed requests
Completed requests
Finished requests
Server Software: HaiYan/
Server Hostname: localhost
Server Port:
Document Path: /
Document Length: bytes
Concurrency Level:
Time taken for tests: 5.894 seconds
Complete requests:
Failed requests:
Write errors:
Total transferred: bytes
HTML transferred: bytes
Requests per second: 5090.28 [#/sec] (mean)
Time per request: 392.906 [ms] (mean)
Time per request: 0.196 [ms] (mean, across all concurrent requests)
Transfer rate: 261020.94 [Kbytes/sec] received
Connection Times (ms)
min mean[+/-sd] median max
Connect:
Processing:
Waiting:
Total:
Percentage of the requests served within a certain time (ms)
%
%
%
%
%
%
%
%
% (longest request)
[root@e46ae471064e conf]#
通过上述比对，即可发现调优之前和调优之后的差距，可以清晰的看到未调优前测试失败的请求数量29911，而调优后失败的请求数量为0。