Nginx TCP (WebSockets) 超时/Keepalive 配置 [英] Nginx TCP (WebSockets) Timeout / Keepalive Config

问题描述

我使用的是 nginx 版本:nginx/1.0.12

我的 nginx.conf 看起来像这样:

My nginx.conf looks like this:

#user  nobody;
worker_processes  1;  

error_log  logs/error.log;
#error_log  logs/error.log  notice;
#error_log  logs/error.log  info;

#pid        logs/nginx.pid;

events {
    worker_connections  1024;
}

tcp {

     upstream websockets {
      ## Play! WS location
       server 127.0.0.1:9000;
       check interval=3000 rise=2 fall=5 timeout=1000;
     }    

    server {
        listen 80; 
        listen 8000;
        server_name socket.domain.com;

        tcp_nodelay on; 
        proxy_pass websockets;
        proxy_send_timeout 300;

    }   

     # virtual hosting
     #include /usr/local/nginx/vhosts/*;
}

我的应用程序似乎每 75 秒(左右)丢弃一次 websocket 连接，我认为这是因为 Nginx 的默认 keepalive 配置.如何增加超时时间?

My application seems to be dropping websocket connnections every 75 sec (or so) which I think is because of Nginx's default keepalive config. How do increase the timeout?

推荐答案

我尝试了 nginx 1.7.1 不支持的 websocket_*_timeout(它给出:未知指令em>).

I tried the websocket_*_timeout which are not supported on nginx 1.7.1 (it gives: unknown directive).

但是设置高 proxy_*_timeout 是有效的:

However setting a high proxy_*_timeout works:

proxy_connect_timeout 7d;
proxy_send_timeout 7d;
proxy_read_timeout 7d;

7d 表示7天，见官方nginx配置参考

此外，您可能只需要设置 proxy_read_timeout 7d; 因为这通常很重要，除非代理背后的服务器非常慢.

Additionally you probably only have to set the proxy_read_timeout 7d; as that's the one that usually matter unless the server behind the proxy is very slow.

这篇关于Nginx TCP (WebSockets) 超时/Keepalive 配置的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！