V G
Can you post your code?
Try putting `import guv; guv.monkey_patch()` at the very top, on the first line. It should be the first thing that runs so it can patch everything before anything else is...
I just tested: **Python 3.5 example:** ``` python import guv; guv.monkey_patch() import guv.wsgi import bottle @bottle.route('/') def index(): return 'Hello, world!' app = bottle.default_app() if __name__ == '__main__': server_sock =...
Something is _very_ wrong if you're only getting 300 req/sec. I'm easily able to get 5,000 req/sec. What are the specs on your computer? Also, make sure you're not installing...
Also note that in your eventlet example, 422 requests timed out. In any production situation, if you have some requests going through and others timing out, that's not an acceptable...
Yeah there's no way your little E-350 can handle 1,000 concurrent requests with CPython no matter what web framework/green thread library you use. Not even my big Core i7 5820k...
Ok I'll try putting the epoll backend back into guv (I removed it a while ago because I thought I'd never need it again). I remember it performed very well,...
Ok yeah something is definitely wrong with the latency, I'll take a look.
I didn't get a chance to look at it yet, but it doesn't require deep digging at all. The latency is primarily coming from the wsgi server, which is a...
This module is actually taken from gevent or eventlet (most likely an old eventlet version). I've never used it myself, so I'm not sure what the intended behaviour should be....