On сре, 15 јан 2014 05:08:49 CET, Roberto De Ioris wrote:
Hi there, we have an API that requires consistent high-performance, and as
part of profiling we've identified that large writes to memcache cause
problems. We would like to write to memcache after we have transmitted our
response to the client. My understanding is that (per the WSGI spec) we
ought to be able to yield our response and have that returned immediately
to Nginx, then do our memcache write after data has been transmitted to
the
client.
Something like:
def handle_request():
response = make_response()
yield response # Response data goes out over the network
memcache.write(response)
Hi, it is a perfectly valid usage
Another option would be to use the close() method supported by WSGI
»
If the iterable returned by the application has a close() method, the
server or gateway must call that method upon completion of the current
request, whether the request was completed normally, or terminated
early due to an error. (This is to support resource release by the
application. This protocol is intended to complement PEP 325's
generator support, and other common iterables with close() methods.
«
Here's a simple example. Unfortunately uWSGI doesn't return the HTTP
result (and close the http connection) until the sleep finishes, and my
reading of the PEP should allow that.
# app.wsgi
# uwsgi --plugins python2,gevent -M --http-socket :8080 --wsgi-file
app.wsgi --gevent 100
import gevent
class App(object):
def __call__(self, env, start_response):
start_response('200 OK', [('Content-Type','text/plain')])
return self
def __iter__(self):
yield "hello world"
def close(self):
gevent.sleep(4)
print "finished!"
application = App()
--
дамјан
_______________________________________________
uWSGI mailing list
[email protected]
http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi
