I am currently working on an app that will generate a resume
from a set of user defined input into several different formats
from an XML file (MS Word, PDF, TXT, HR-XML, and HTML). We are
thinking that we will write all the files once at publish time and
then store them (not sure where yet). We are doing this because we
will be hosting the online version of the resume as a CFM file with
access to all the other formats of the resume from their online
resume. We are assuming that there will be many more reads then
their will be writes over the life of the resume. So we don't want
to compile these each time a user requests one (that is a Word,
PDF, HTML, or HR-XML version).
The question I have now is should we store the files in the
database or the webserver.
I would think that it makes sense to store them on the
webserver. But as this will need to be in a clustered environment
then I will need to replicate these across the farm as each new
user creates a resume. So does anyone know if the penalty for
replicating these across the farm is higher then calling them from
database. Assuming that the average file size is 50K and on average
50 files will be called over the life of the resume. Thoughts?