El 29/05/17 a les 13:58, Jan Grasnick | Grasbauer UG ha escrit:
Am 29.05.2017 um 11:58 schrieb Nicolas Évrard:
* Sergi Almacellas Abellana [2017-05-29 10:07 +0200]:
El 26/05/17 a les 17:51, Maximilian Holtzberg ha escrit:
You can try to run the client over mitmproxy like this:
mitmproxy -R http://localhost:8000 -p 8088
and connect the client to it. It gives you some timing informations
like:
First request byte 2017-05-26 12:07:21.961
Request complete 2017-05-26 12:07:21.985
First response byte 2017-05-26 12:07:22.027
Response complete 2017-05-26 12:07:22.043
You are still not proving any evidence that trytond is slower.
And also don't providing details about the setup-up.
I had the same issue with 4.2 and replaced werkzeug with cherrypi. Since
it happened in a very strange environment (windows, trytond in cygwin) I
didn't give it much thought. But the improvement was significant (from
mostly unusable to wow :) . I don't know your environment but you can
give it a try and report back your results . So we have an entry point
for further investigations.
trytond default comandline uses werkzeug run_simple [1] which as
explained on their docs [2] it performs poorly under high load. Indeed,
there is some thread on this list [3] explaining how to configure it as
WSGI server, which will provide better performance.
Hope it helps.
[1] http://hg.tryton.org/trytond/file/93b914901108/bin/trytond#l46
[2] http://werkzeug.pocoo.org/docs/0.12/serving/#werkzeug.serving.run_simple
[3] https://groups.google.com/d/msg/tryton-dev/GI2h2cSUG6A/e2F2H3nXBQAJ
--
Sergi Almacellas Abellana
www.koolpi.com
Twitter: @pokoli_srk
--
You received this message because you are subscribed to the Google Groups
"tryton-dev" group.
To view this discussion on the web visit
https://groups.google.com/d/msgid/tryton-dev/b0250ed5-87ac-b938-cbc5-10bcbb5933ce%40koolpi.com.
