Double checked Python benchmark script
epo001 at hotmail.com
Wed Jun 25 11:30:52 CEST 2003
I used to be a C wizard, will check it out later on in my test environment.
This is much simpler but in mine I wanted to test the effect of requesting
all all versus single attributes and to do random lookups. This is good
however for initial high level investigations.
>I wrote my own test script (attached). It's less complex to read but more
>flexible (through LDAP URL as parameter). Please review.
>Usage: pref_test.py <LDAP URL> <number of tests>
>Grab whole sub-tree 100 times (no server-side limit or logging, ~500
>$ python perf_test.py "ldap://localhost:1390/dc=stroeder,dc=de?*?sub" 100
>Opening connection each time: 20.4206629992
>Reusing connection: 16.9708769321
>Using ldapsearch: 6.0001270771
>I'd appreciate if anyone with C knowledge could review what is done in
>ldap_result() (LDAPObject.c) and LDAPmessage_to_python() (message.c).
More information about the python-ldap