| View previous topic :: View next topic |
| Author |
Message |
mg66
-
Joined: 15 Aug 2004
Posts: 85
Location: USA, Illinois
|
|
| Back to top |
   |
 |
aprelium
-
Joined: 22 Mar 2002
Posts: 6800
|
|
| Back to top |
 |
|
mg66
-
Joined: 15 Aug 2004
Posts: 85
Location: USA, Illinois
|
|
| Back to top |
|
|
aprelium
-
Joined: 22 Mar 2002
Posts: 6800
|
 Posted: Mon Jun 05, 2006 10:37 am Post subject: |
 |
|
| mg66 wrote: | Thanx for the prompt reply.
Does Abyss fastcgi do the same sought of thing to a point by running processes for a certain time, like I have set my sites to 120 secs. |
FastCGI reduces startup time of the scripts since it will load the PHP processor once and reuse it for several requests. EAccelerator (and the likes) operate inside PHP and cache the compiled form of the scripts to reuse them later.
So to sum up. With CGI and no accelerator:
Script execution time = PHP interpreter loading time + PHP script reading + PHP script pseudo-compilation + PHP interpretation of the pseudo-compiled version
With FastCGI (starting from the second request):
Script execution time = PHP script reading + PHP script pseudo-compilation + PHP interpretation of the pseudo-compiled version
With FastCGI and EAccelerator (starting from the second request):
Script execution time = PHP interpretation of the pseudo-compiled version
| Quote: | | Since running fastcgi my load times seem faster but the stats dont seem to show any significant speed increase. |
Which stats are you referring to? Normal stats won't show you the difference. Only web server stress tools can.
_________________
Support Team
Aprelium - http://www.aprelium.com |
|
| Back to top |
|
|
|
