In order to benchmark the web server performance, I wrote a simple PHP script to poll a website with CURL. It might not be the most accurate, but all I need is a baseline to measure the average load time.
After some testing, here's result:
no acceleration: ~0.550 sec
eaccelerator 32Mb cache: ~0.475 sec
eaccelerator 64Mb cache: ~0.455 sec
APC 32Mb cache: ~0.417 sec
APC 64Mb cache: ~0.413 sec
APC is also come with a control panel:
From the performance point of view, any opcode cache helps server performance as expected. APC, from the result, has a edge over eaccelerator. In addition, APC has a more informative control panel. APC also a part of PECL, which is semi-standard to PHP.
Therefore, I switched to APC.
BTW, I also use my benchmark script on Google.... its load time is a wicked 0.112 sec. I guess I still have a long way to go.
My passion has always been on software development, and I know it since I
wrote my first program on an Apple II. I worked on cellular wireless system
in the past (C/C++) and now focus on web application
(LAMP, PHP, MySQL, CakePHP, Symfony, jQuery, Google AppEngine/python).
![[Valid XHTML 1.0 Strict]](https://edmondhui.ddns.net/blog/skins/admun3/valid-xhtml.png)
Comments
Leave comment
Trackback