> Hi,
>
> Below is how I am launching uWSGI via supervisor for the gevent loop
>
> /usr/local/bin/uwsgi --loop gevent --socket 127.0.0.1:8070 --processes
> 3 --pp /home/ubuntu/workspace/rtbopsConfig/rtbServers/rtbUwsgiPixelServer/
> --wsgi-file
> /home/ubuntu/workspace/rtbopsConfig/rtbServers/rtbUwsgiPixelServer/uwsgiPixelServer.py
> -b 32768 --master --async 20 --enable-threads --listen 2048
>
> I get about 1-2K QPS. I note that my ram and swap tend to grow to
> being full so I restart uWSGI. I have 4 servers, with 3 cores
> dedicated for each server.
>
>
> Here are the relevant ngix confing for nginx
> user www-data;
> worker_processes 4; # 2*number of cpus
> pid /var/run/nginx.pid;
> events {
> worker_connections 30000;
> multi_accept on;
> use epoll;
> }
>
>
> upstream uwsgi_pixel {
> server 127.0.0.1:8070;
> }
> location /pixel/ {
> include uwsgi_params;
> uwsgi_buffering off;
> #uwsgi_param X-Real-IP $remote_addr;
> #uwsgi_param Host $http_host;
> #uwsgi_pass uwsgi_bidder;
> uwsgi_pass 127.0.0.1:8070;
> }
>
> How can I modify how I launch uWSGI so its aware and restart? Or any
> other options I can use?
> _______________________________________________
> uWSGI mailing list
> [email protected]
> http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi
>
Try --reload-on-rss <X> where X is the maximum amount of per-process memory.
If you do not want your app to swap, add --never-swap (this will trigger a
brutal kill in case of 'out of memory')
--
Roberto De Ioris
http://unbit.it
_______________________________________________
uWSGI mailing list
[email protected]
http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi
