标签：stat 影响 rem 通过 comm document 缓存 服务器 图片
通常情况下缓存是用来减少后端压力，将压力尽可能的往前推，减少后端压力，提高网络并发延时！
服务端缓存
代理缓存
客户端缓存
Nginx代理缓存原理
proxy_cache配置语法：
Syntax: proxy_cache zone | off;
Default: proxy_cache off;
Context: http, server, location
//缓存路径
Syntax: proxy_cache_path path [levels=levels]
[use_temp_path=on|off] keys_zone=name:size [inactive=time]
[max_size=size] [manager_files=number] [manager_sleep=time][manager_threshold=time]
[loader_files=number] [loader_sleep=time] [loader_threshold=time] [purger=on|off]
[purger_files=number] [purger_sleep=time] [purger_threshold=time];
Default: —
Context: http
缓存过期周期：
Syntax: proxy_cache_valid [code ...] time;
Default: —
Context: http, server, location
//示例
proxy_cache_valid 200 302 10m;
proxy_cache_valid 404 1m;
缓存的维度：
Syntax: proxy_cache_key string;
Default: proxy_cache_key $scheme$proxy_host$request_uri;
Context: http, server, location
//示例
proxy_cache_key "$host$request_uri $cookie_user";
proxy_cache_key $scheme$proxy_host$uri$is_args$args;
| OS | service | IP |
|---|---|---|
| Centos 7.2 | Nginx Proxy | 192.168.1.1 |
| Centos 7.2 | Nginx web | 192.168.1.2 |
$ mkdir -p /soft/code{1..3}
$ for i in {1..3};do echo Code1-Url$i > /soft/code1/url$i.html;done
$ for i in {1..3};do echo Code2-Url$i > /soft/code2/url$i.html;done
$ for i in {1..3};do echo Code3-Url$i > /soft/code3/url$i.html;done
$ vim /etc/nginx/conf.d/web_node.conf
server {
listen 8081;
root /soft/code1;
index index.html;
}
server {
listen 8082;
root /soft/code2;
index index.html;
}
server {
listen 8083;
root /soft/code3;
index index.html;
}
$ nginx -t
$ systemctl start nginx
$ ss -lnt | grep 80
LISTEN 0 128 *:8081 *:*
LISTEN 0 128 *:8082 *:*
LISTEN 0 128 *:8083 *:*
$ mkdir /soft/cache -p # 创建缓存所需目录
$ vim /etc/nginx/conf.d/proxy_cache.conf
upstream cache {
server 192.168.1.2:8081;
server 192.168.1.2:8082;
server 192.168.1.2:8083;
}
proxy_cache_path /soft/cache levels=1:2 keys_zone=code_cache:10m max_size=10g inactive=60m use_temp_path=off;
# proxy_cache_path: 存放缓存临时文件
# levels=1:2：按照两层目录分级
# keys_zone=code_cache:10m：开辟空间名, 10m:开辟空间大小, 1m可存放8000key
# max_size=10g：控制最大大小，超过后Nginx会启动淘汰规则
# inactive=60m：60分钟没有被访问缓存就会被清理
# use_temp_path=off：临时文件，会影响性能，建议关闭
server {
listen 80;
server_name 192.168.1.1;
location / {
proxy_pass http://cache;
proxy_cache code_cache; # proxy_cache：开启缓存
proxy_cache_valid 200 304 12h;
proxy_cache_valid any 10m; # proxy_cache_valid：状态码200|304的过期时间为12h，其余状态码过期时间为10分钟
add_header Nginx-Cache "$upstream_cache_status"; # add_header：添加头部信息，观察客户端respoce是否命中
proxy_next_upstream error timeout invalid_header http_500 http_502 http_503 http_504; # proxy_next_upstream：出>现500、502、503、504等错误页面，会跳过此台服务器访问下台
include proxy_params;
}
}
$ nginx -t
$ systemctl start nginx
$ curl -s -I http://192.168.1.1/url1.html | grep "Nginx-Cache"
Nginx-Cache: MISS
# MISS表示没有命中缓存
$ curl -s -I http://192.168.1.1/url1.html | grep "Nginx-Cache"
Nginx-Cache: HIT
# HIT表示命中缓存
清理proxy_cache代理缓存！
$ rm -rf /soft/cache/*
$ curl -s -I http://192.168.1.1/url1.html | grep "Nginx-Cache"
Nginx-Cache: MISS
ngx_cache_pruge扩展模块，需要编译安装Nginx$ mkdir /soft/src && cd /soft/src # 创建响应目录
$ wget http://nginx.org/download/nginx-1.12.2.tar.gz # 下载Nginx包
$ tar xf nginx-1.12.2.tar.gz
$ wget http://labs.frickle.com/files/ngx_cache_purge-2.3.tar.gz # 下载ngx_cache_purge
$ tar xf ngx_cache_purge-2.3.tar.gz # 下载ngx_cache_purge
$ cd nginx-1.12.2/ && ./configure --prefix=/server/nginx --add-module=../ngx_cache_purge-2.3 --with-http_stub_status_module --with-http_ssl_module
$ make && make install
# 需要将上面的缓存proxy_cache.conf文件拷贝源码包中,并增加如下内容
location ~ /purge(/.*) {
allow 127.0.0.1;
allow 192.168.1.0/24;
deny all;
proxy_cache_purge code_cache $host$1$is_args$args;
}
$ /server/nginx/sbin/nginx -t
$ /server/nginx/sbin/nginx -s reload
# 检测配置重新加载
举例说明：浏览器访问http://192.168.1.1/url1.html ，访问http://192.168.1.1/purge/url1.html 清除url1.html页面缓存的内容！
指定部分页面不进行proxy-cache缓存！
$ vim /etc/nginx/conf.d/proxy_cache.conf
upstream cache {
server 192.168.1.2:8081;
server 192.168.1.2:8082;
server 192.168.1.3:8083;
}
proxy_cache_path /soft/cache levels=1:2 keys_zone=code_cache:10m max_size=10g inactive=60m use_temp_path=off;
server {
listen 80;
server_name 192.168.1.1;
if ($request_uri ~ ^/(url3|login|register|password)) {
set $cookie_nocache 1; # 访问的url包含url3|login|register|password这些关键字的不进行缓存
}
location / {
proxy_pass http://cache;
proxy_cache code_cache;
proxy_cache_valid 200 304 12h;
proxy_cache_valid any 10m;
proxy_cache_key $host$uri$is_args$args;
proxy_no_cache $cookie_nocache $arg_nocache $arg_comment; # 注意此行
proxy_no_cache $http_pargma $http_authorization; # 注意此行
add_header Nginx-Cache "$upstream_cache_status";
proxy_next_upstream error timeout invalid_header http_500 http_502 http_503 http_504;
include proxy_params;
}
}
$ nginx -t
$ systemtl reload nginx
$ curl -s -I http://192.168.1.1/url3.html | grep "Nginx-Cache"
Nginx-Cache: MISS
$ curl -s -I http://192.168.1.1/url3.html | grep "Nginx-Cache"
Nginx-Cache: MISS
# 测试访问定义不缓存的内容，从结果看出确实不缓存了
$ curl -s -I http://192.168.1.1/url1.html | grep "Nginx-Cache"
Nginx-Cache: MISS
$ curl -s -I http://192.168.1.1/url1.html | grep "Nginx-Cache"
Nginx-Cache: HIT
# 原本的内容依然会被缓存
‘
通过日志记录proxy_cache命中情况与对应url
# 修改/etc/nginx/nginx.conf中log_format格式
log_format main ‘$remote_addr - $remote_user [$time_local] "$request" ‘
‘$status $body_bytes_sent "$http_referer" ‘
‘"$http_user_agent" "$http_x_forwarded_for"‘ ‘"$upstream_cache_status"‘;
# 就是在原本的日志格式上增加$upstream_cache_status变量
# 修改proxy_cache.conf, 在server标签新增access日志
access_log /var/log/nginx/proxy_cache.log main;
# 使用curl访问，最后显示日志命令情况！
192.168.1.2 - - [27/Jun/2020:16:51:30 +0800] "HEAD /url1.html HTTP/1.1" 200 0 "-" "curl/7.29.0" "-""MISS"
192.168.1.2 - - [27/Jun/2020:16:51:31 +0800] "HEAD /url1.html HTTP/1.1" 200 0 "-" "curl/7.29.0" "-""HIT"
如果需要查看nginx cache的命中率可参考Nginx cache查看命中率
标签：stat 影响 rem 通过 comm document 缓存 服务器 图片
原文地址：https://www.cnblogs.com/lvzhenjiang/p/14022161.html
