> We are not in production yet and we have just made some load testing
for
> the moment.
> But, I don't see why we would have scalability problems directly
linked to
> the choice of CFCs-as-models (except that it might consume more RAM
than
> before).
>
We all know what the problems with assumptions are. Why don't you load
test your application and report on what percentage of total page
serving time is consumed by CFCs on average for a page request.

> It is true that I was not sure about this ... until the new
macromedia.com
> was launched. It is the best proof of scalability of CFMX (and CFCs).
> Figures show that CFCs don't add so much overhead compared to
cfinclude or
> cfmodule.
> 
Which figures do you refer to? The published information from Macromedia
only compares the time is takes to call a CFC method vs. a UDF and
custom tag. Unfortunately, that data doesn't appear to take into account
complex inheritance hierarchies or any overhead associated with
persisted CFCs.

More than one person however has posted data to public mailing lists
showing the overhead associated with the CFC use cases I mentioned. It
is not that hard to see; try it for yourself.

-Matt

----------------------------------------------------------
You are subscribed to cfcdev. To unsubscribe, send an email
to [EMAIL PROTECTED] with the word 'unsubscribe cfcdev' 
in the message of the email.

CFCDev is run by CFCZone (www.cfczone.org) and supported
by Mindtool, Corporation (www.mindtool.com).

Reply via email to