Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > comp.lang.python > #44667 > unrolled thread

Highest performance Python framework for API exposure?

Started byAlec Taylor <alec.taylor6@gmail.com>
First post2013-05-04 00:22 +1000
Last post2013-05-03 20:18 -0400
Articles 2 — 2 participants

Back to article view | Back to comp.lang.python


Contents

  Highest performance Python framework for API exposure? Alec Taylor <alec.taylor6@gmail.com> - 2013-05-04 00:22 +1000
    Re: Highest performance Python framework for API exposure? Roy Smith <roy@panix.com> - 2013-05-03 20:18 -0400

#44667 — Highest performance Python framework for API exposure?

FromAlec Taylor <alec.taylor6@gmail.com>
Date2013-05-04 00:22 +1000
SubjectHighest performance Python framework for API exposure?
Message-ID<mailman.1269.1367590981.3114.python-list@python.org>
Which Python framework has benchmarked the highest?

I am looking for a solution to initially run on Heroku or OpenShift;
but eventually run on a pure IaaS.

Will expose the interfaces RESTfully as JSON, purposed to act as the
"glue" between the endpoints, databases (incl. caches) and queues.

An additional requirement is support for serving static, downloadable
content efficiently.

Currently looking at Falcon, Bottle and Werkzeug.

Suggestions are greatly appreciated.

Thanks,

Alec Taylor

[toc] | [next] | [standalone]


#44696

FromRoy Smith <roy@panix.com>
Date2013-05-03 20:18 -0400
Message-ID<roy-4B8DD5.20184903052013@news.panix.com>
In reply to#44667
In article <mailman.1269.1367590981.3114.python-list@python.org>,
 Alec Taylor <alec.taylor6@gmail.com> wrote:

> Which Python framework has benchmarked the highest?
> 
> I am looking for a solution to initially run on Heroku or OpenShift;
> but eventually run on a pure IaaS.
> 
> Will expose the interfaces RESTfully as JSON, purposed to act as the
> "glue" between the endpoints, databases (incl. caches) and queues.

Based on my experience, it doesn't matter what framework you use.  All 
that's going to matter is how fast you can pump data in and out of your 
database.  The overhead added by the Python front-end probably won't be 
significant.

Pick a framework that you find easy to use and don't sweat it.  Invest 
your time worrying about your database.

> An additional requirement is support for serving static, downloadable
> content efficiently.

You don't want to do that with Python.  You want to do that with 
something that's optimized for serving static content (such as nginx).

[toc] | [prev] | [standalone]


Back to top | Article view | comp.lang.python


csiph-web