Jeff and Łukasz, Is there any possibility that nginx can not send requests to upstreams (uwsgi servers) so uwsgi servers are idle and have not so many requests to handle?
I checked uwsgi broken pipe errors, which are not huge, compared to the number of timed out requests in nginx error log. It looks like ddos attack, but I'm sure they are normal requests. I logged user sessions and usernames in access log. Our users often learning on my site around 9pm to 11pm in China time. I don't know why, but they just go to site at this time every day. The number of requests can be 20 times than the other time in a day. On Wed, Nov 14, 2012 at 12:18 AM, Jeff Van Voorst <[email protected]>wrote: > You may try using bees with machine guns (https://github.com/newsapps/** > beeswithmachineguns <https://github.com/newsapps/beeswithmachineguns>) to > perform a potentially large DDOS against your site to see how it holds up. > Unless apache benchmark has been extended it may not be able to fully load > your machine. > > I don't know if bees with machine guns requires the use of Amazon's EC2 > instances or if the code can be readily changed to use any distributed > system. > > --Jeff > > ______________________________**_________________ > uWSGI mailing list > [email protected] > http://lists.unbit.it/cgi-bin/**mailman/listinfo/uwsgi<http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi> > -- *吴焱红(Samuel)* 博客: blog.shanbay.com
_______________________________________________ uWSGI mailing list [email protected] http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi
