CGI vs twisted vs mod_python

I've run a few tests with a neat idea I found on Bill's blog (timeurl.py).

Here a few numbers comparing cgi, twisted and mod_python:
Between the 2 test runs apache and twisted were restarted
and all *.pyc file's were removed.


---------- With 5 blog entries ----------

1 request: -----------------
first run:
CGI         : 0.627060890198
WSGI-CGI    : 0.57884311676
twisted     : 0.722832918167
mod_python  : 0.524918079376

second run:
CGI         : 0.489893913269
WSGI-CGI    : 0.534939050674
twisted     : 0.77103304863
mod_python  : 0.0480790138245


100 requests: --------------
first run:
CGI         : 48.792345047
WSGI-CGI    : 48.2621901035
twisted     : 99.5655610561
mod_python  : 7.2256770134

second run:
CGI         : 49.7383179665
WSGI-CGI    : 50.3378739357
twisted     : 100.032269955
mod_python  : 4.65755105019



---------- With 2795 blog entries ----------

1 request: -----------------
first run:
CGI         : 2.38962101936
WSGI-CGI    : 2.3695230484
twisted     : 2.65319085121
mod_python  : 2.41627717018

second run:
CGI         : 2.24685406685
WSGI-CGI    : 2.31155514717
twisted     : 1.37736415863
mod_python  : 1.12132692337


100 requests: --------------
first run:
CGI         : 235.439440966
WSGI-CGI    : 236.187808037
twisted     : 199.473860025
mod_python  : 117.538961172

second run:
CGI         : 229.891646862
WSGI-CGI    : 227.116400003
twisted     : 199.379323006
mod_python  : 106.703619957


Interesting was that with CGI and mod_python the CPU was at 99% while
with twisted it was at +-50%.

cheers
Steven


here's the script:

#!/usr/bin/env python
import urllib
import timeit

url = urllib.URLopener()

test_cases = [
    {'label': 'CGI         :', 'page': "http://www.c-area.lan/blog/"},
    {'label': 'WSGI-CGI    :', 'page': "http://www.c-area.lan/weblog-cgi/"},
    {'label': 'twisted     :', 'page': "http://localhost:8080/"},
    {'label': 'mod_python  :', 'page': "http://www.c-area.lan/weblog/"}
]

def run_test_cases(cases, num_of_tests):
    for case in cases:
        t = timeit.Timer('url.open("%s").read()' % case['page'], \
            'from __main__ import url')
        print case['label'],
        print t.timeit(num_of_tests)

print "1 request: -----------------"
print "first run:"
run_test_cases(test_cases, 1)
print
print "second run:"
run_test_cases(test_cases, 1)
print "\n"

print "100 requests: --------------"
print "first run:"
run_test_cases(test_cases, 100)
print
print "second run:"
run_test_cases(test_cases, 100)



-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click