Hi,

I'm using a set of (self-written) scripts to do some db-interactions (eg.
reading recordsets from a DB2 (AS/400), converting
them and inserting them in a MySQL-DB). These scripts where written on my
office-workstation where they run very well.
When i put them on our intranet-server the script performance dropped to a
point where it was not even possible to import
all of the data because of the massive runtime (around 6-7 records/sec
compared to 400-500 on my workstation).

I did some research and stumbled across some benchmark-scripts written by
Sebastian Bergmann ( http://www.sebastian-bergmann.de/PHP_Benchmark/ ) which
i used to benchmark the two systems.

My office-workstation: AthlonXP 2400+ / 512 / 5400rpm HDD IDE / Win2k Pro /
PHP 4.3.4
Intranet Server: 2x 2.8GHz Xeon / 2048 / 5x 15.000rpm Raid5 / WinNT4
Server-SP6a / PHP 4.3.4
Nearly the same php.ini is used on both systems (except for error-reporting
and stuff like that).

One should expect to see all of the above tests run faster on the server.
That is true for all but one test:
No5 (Array Index Access) shows terrible results (~740.000 on my workstation
/ ~10.000 on the server). Even if I don't use
this type of access in my scripts I think it is the explanation for the
terrible performance.

I tried older versions of PHP and found the "bug"(?) existing since 4.1.0

No Apache was involved while the scripts where run (cmd-line).


Does anyone have an idea what might cause this type of problem ?



Greetings and a happy new year

Jochen Boedeker

-- 
PHP General Mailing List (http://www.php.net/)
To unsubscribe, visit: http://www.php.net/unsub.php

Reply via email to