|
courtney ludwin
courtneyludwin at yahoo.com
Thu Aug 12 15:06:38 EDT 2004
I still think something is not right,
I am able to fill a cheetah template well over 1000
times a second (just wrote a script to test this out),
and I know mod_python is fast so I think that it is
more of a configuration issue...
Cheetah is fast when filling templates, but slow to
load templates therefore it may be a importing issue,
but I don't see how in my code, I am only importing
the template once:
example handler:
#---------------------------------------------------------
from ads.tmpl.index import index
from common.ui import navigation as nav
from mod_python import apache
def handler(req, sesh):
tmpl = index(searchList=[{
"sesh": sesh,
"site_nav":
nav.get_site_nav("adhome"),
"ad_nav": nav.get_ad_nav("adhome")}])
return apache.OK, tmpl
#---------------------------------------------------------
--- Nick <nick at dd.revealed.net> wrote:
> courtney ludwin wrote:
> > The reason I ask is that I have not been able to
> get
> > mp+cheetah to perform well together even after
> > compiling templates. On my P3-933 w/ 1Gig of ram
> I
> > can serve ~10 pages/sec using Cheetah and ~200
> > pages/sec NOT using Cheetah.
>
> That doesn't sound completely unreasonable for a
> benchmark figure (i.e.
> constantly hitting the server with simultaneous
> requests and driving up the
> CPU). My benchmarks typically don't even do that
> well, with say 10 clients
> and 0 delay between each request.
>
> Nick
> _______________________________________________
> Mod_python mailing list
> Mod_python at modpython.org
>
http://mailman.modpython.org/mailman/listinfo/mod_python
>
__________________________________
Do you Yahoo!?
New and Improved Yahoo! Mail - Send 10MB messages!
http://promotions.yahoo.com/new_mail
|