Nginx反向代理websocket配置实例
投稿：junjie
最近有一个需求，就是需要使用 nginx 反向代理 websocket，经过查找一番资料，目前已经测试通过，本文只做一个记录
注： 看官方文档说 Nginx 在 1.3 以后的版本才支持 websocket 反向代理，所以要想使用支持 websocket 的功能，必须升级到 1.3 以后的版本，因此我这边是下载的 Tengine 的最新版本测试的
1.下载 tengine 最近的源码
wget http://tengine.taobao.org/download/tengine-2.0.3.tar.gz
2.安装基础的依赖包
yum -y install pcre*
yum -y install zlib*
yum -y install openssl*
3.解压编译安装
tar -zxvf tengine-2.0.3.tar.gz cd tengine-2.0.3 ./configure --prefix=安装目录 make sudo make install
nginx.conf 的配置如下：
user apps apps;
worker_processes 4; # 这个由于我是用的虚拟机，所以配置的 4 ，另外 tengine 可以自动根据CPU数目设置进程个数和绑定CPU亲缘性
# worker_processes auto
# worker_cpu_affinity auto
error_log logs/error.log;
pid logs/nginx.pid;
#Specifies the value for maximum file descriptors that can be opened by this process.
worker_rlimit_nofile 65535;
events {
use epoll;
worker_connections 65535;
}
# load modules compiled as Dynamic Shared Object (DSO)
#
#dso {
# load ngx_http_fastcgi_module.so;
# load ngx_http_rewrite_module.so;
#}
http {
include mime.types;
default_type application/octet-stream;
server_names_hash_bucket_size 128;
client_header_buffer_size 4k;
large_client_header_buffers 4 32k;
client_max_body_size 80m;
sendfile on;
tcp_nopush on;
client_body_timeout 5;
client_header_timeout 5;
keepalive_timeout 5;
send_timeout 5;
open_file_cache max=65535 inactive=20s;
open_file_cache_valid 30s;
open_file_cache_min_uses 1;
tcp_nodelay on;
fastcgi_connect_timeout 300;
fastcgi_send_timeout 300;
fastcgi_read_timeout 300;
fastcgi_buffer_size 64k;
fastcgi_buffers 4 64k;
fastcgi_busy_buffers_size 128k;
fastcgi_temp_file_write_size 128k;
client_body_buffer_size 512k;
proxy_connect_timeout 5;
proxy_read_timeout 60;
proxy_send_timeout 5;
proxy_buffer_size 16k;
proxy_buffers 4 64k;
proxy_busy_buffers_size 128k;
proxy_temp_file_write_size 128k;
gzip on;
gzip_min_length 1k;
gzip_buffers 4 16k;
gzip_http_version 1.0;
gzip_comp_level 2;
gzip_types text/plain application/x-javascript text/css application/xml;
gzip_vary on;
proxy_temp_path /dev/shm/temp;
proxy_cache_path /dev/shm/cache levels=2:2:2 keys_zone=cache_go:200m inactive=5d max_size=7g;
log_format log_access '$remote_addr - $remote_user [$time_local] "$request" "$request_time" "$upstream_response_time"'
'$status $body_bytes_sent "$http_referer" '
'"$http_user_agent" $http_x_forwarded_for $host $hostname' ;
#websocket 需要加下这个
map $http_upgrade $connection_upgrade {
default upgrade;
'' close;
}
include /home/apps/tengine/conf/test.com;
}
test.com 的配置文件内容：
upstream test.com {
server 192.168.1.5:9000;
}
server {
listen 80;
server_name test.com;
#charset koi8-r;
#access_log logs/host.access.log main;
location ^~ /websocket {
proxy_pass http://test.com;
proxy_redirect off;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header Host $host;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_http_version 1.1;
proxy_set_header Upgrade $http_upgrade;
proxy_set_header Connection "upgrade";
}
}
解析 map 指令
上面 nginx.conf 配置中的 map $http_upgrade $connection_upgrade 的作用，参考 http://www.ttlsa.com/nginx/using-nginx-map-method/
该作用主要是根据客户端请求中 $http_upgrade 的值，来构造改变 $connection_upgrade 的值，即根据变量 $http_upgrade 的值创建新的变量 $connection_upgrade，创建的规则就是 {} 里面的东西，请见配置：
map $http_upgrade $connection_upgrade {
default upgrade;
'' close;
}
其中的规则没有做匹配，因此使用默认的，即 $connection_upgrade 的值会一直是 upgrade。然后如果 $http_upgrade为空字符串的话，那值会是 close。个人的理解！