-1

I have nodejs app that sending > 800 mongodb documents on client startup (execute only when a client access my app for the first time).

Nginx as reverse proxy in front of node server.

App server spec

  • Digital Ocean
  • CentOS 7.2
  • 2GB Ram
  • 2CPU

MongoDB server spec

  • Digital Ocean
  • Ubuntu 14.04
  • 512 RAM
  • 1 CPU

nginx -v // nginx version: nginx/1.8.1

nginx config

user  nginx;
worker_processes  2;
worker_rlimit_nofile 100480;

error_log  /var/log/nginx/error.log;

pid        /run/nginx.pid;

events {
  worker_connections  1024;
}

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;

  index   index.html index.htm;

  server {
    server_name 128.199.139.xxx;

    root /var/www/myapp/bundle/public;

    module_app_type node;

    module_startup_file main.js;

    module_env_var MONGO_URL mongodb://{username}:{password}@128.199.139.xxx:27017/;

    module_env_var ROOT_URL http://128.199.139.xxx;

    location / {
      proxy_pass http://128.199.139.xxx;
      proxy_http_version 1.1;

      #proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
      #proxy_set_header X-Real-IP $remote_addr;

      # pass the host header - http://wiki.nginx.org/HttpProxyModule#proxy_pass
      #proxy_set_header Host $host;

      # WebSocket proxying - from http://nginx.org/en/docs/http/websocket.html
      proxy_set_header Upgrade "upgrade";
      proxy_set_header Connection $http_upgrade;

      #add_header 'Access-Control-Allow-Origin' '*';
    }
  }

  server {
    listen       80 default_server;
    server_name  localhost;
    root         /usr/share/nginx/html;

    #charset koi8-r;

    #access_log  /var/log/nginx/host.access.log  main;

    # Load configuration files for the default server block.
    include /etc/nginx/default.d/*.conf;

    location / {
    }

    # redirect server error pages to the static page /40x.html
    #
    error_page  404              /404.html;
    location = /40x.html {
    }

    # redirect server error pages to the static page /50x.html
    #
    error_page   500 502 503 504  /50x.html;
    location = /50x.html {
    }
  }
}

Error Log

Below error log:

2016/03/17 09:46:00 [crit] 10295#0: accept4() failed (24: Too many open files)
2016/03/17 09:46:01 [crit] 10295#0: accept4() failed (24: Too many open files)
2016/03/17 09:46:01 [crit] 10295#0: accept4() failed (24: Too many open files)
.....many duplicate error as above
2016/03/17 09:47:35 [error] 10295#0: *4064 recv() failed (104: Connection reset by peer) while reading response header from upstream, client: 128.199.139.160, server: 128.199.139.1$
2016/03/17 09:47:35 [alert] 10295#0: accept4() failed (9: Bad file descriptor)
[ 2016-03-17 09:53:47.8144 10403/7f2833c9a700 age/Ust/UstRouterMain.cpp:422 ]: Signal received. Gracefully shutting down... (send signal 2 more time(s) to force shutdown)
[ 2016-03-17 09:53:47.8145 10403/7f2839500880 age/Ust/UstRouterMain.cpp:492 ]: Received command to shutdown gracefully. Waiting until all clients have disconnected...
[ 2016-03-17 09:53:47.8146 10403/7f2833c9a700 Ser/Server.h:464 ]: [UstRouter] Shutdown finished
2016/03/17 09:54:11 [alert] 10549#0: 1024 worker_connections are not enough
2016/03/17 09:54:11 [error] 10549#0: *1021 recv() failed (104: Connection reset by peer) while reading response header from upstream, client: 128.199.139.160, server: 128.199.139.1$
2016/03/17 09:54:12 [alert] 10549#0: 1024 worker_connections are not enough
2016/03/17 09:54:12 [error] 10549#0: *2043 recv() failed (104: Connection reset by peer) while reading response header from upstream, client: 128.199.139.160, server: 128.199.139.1$
2016/03/17 11:43:20 [alert] 10549#0: 1024 worker_connections are not enough
2016/03/17 11:43:20 [error] 10549#0: *3069 recv() failed (104: Connection reset by peer) while reading response header from upstream, client: 128.199.139.160, server: 128.199.139.1$
2016/03/17 13:49:54 [error] 10549#0: *3071 open() "/usr/share/nginx/html/robots.txt" failed (2: No such file or directory), client: 180.97.106.xx, server: localhost, request: "GET

Appreciated for any help or clue

PROBLEM

The problem is I'm getting following error according to above setup:

  • Too many open files
  • worker_connections are not enough
  • (104: Connection reset by peer) while reading response header from upstream

How to solve this issue? Thanks

prime
  • 1
  • 1

1 Answers1

0

Well it seems like googling a little bit, like "worker connections are not enough" or "nginx too many open files" already gives some clues, like maybe proxying to port 80 (on the public IP):

  proxy_pass http://128.199.139.xxx;

leads to a loop (hence "too many open files" and "1024 workers are not enough").

Maybe you should try these hints and tell, if they didn't solve your problem.

https://stackoverflow.com/questions/28265717/worker-connections-are-not-enough

Too many open files with nginx, can't seem to raise limit

http://www.cyberciti.biz/faq/linux-unix-nginx-too-many-open-files/

https://anothersysadmin.wordpress.com/2013/08/09/nginx-and-the-too-many-open-files-limit/

Community
  • 1
zezollo
  • 390
  • 1
  • 3
  • 10
  • Well you don't know anything yet you answering and downvoting my question. Look at my nginx config. I have applied all google tell me! – prime Mar 20 '16 at 00:32
  • No I didn't downvote (I even can't). You should state clearly what you have tried (and did fail) in your question. Whatever, seing this: `http { ... server { server_name 128.199.139.xxx; ... location / { proxy_pass http://128.199.139.xxx; ...` let me think you're actually proxying 128.199.139.xxx:80 to itself (loop) and that it should be fixed. – zezollo Mar 20 '16 at 05:45