i. Title: Infrastructure Inventory and Management System.
  
ii. Author: Tim Mabbott (tmabbott@post.harvard.edu)

iii. Abstract:
  This project will facilitate the management of a typical J2EE environment.
The main goal will be to provide administators with a way to track what 
components exist within the environment, and a way to represent the 
relationships between them with meaningful views.

v. Why:
  We have 100+ applications at HBS spread across 16 systems, some running 
on Weblogic, some on ATG Dynamo (Legacy), some clustered (most not).  Most 
apps have database dependancies, some have filesystem dependancies (including 
cross server NFS mounts), some have vendor dependancies.  This info is 
currently tracked in a couple of spreadsheets but representing the 
relationships (in a meaningful way) is problematic, as is keeping it current.

vi. How:
  The data about the environment will (obviously) be stored in XML.  Along 
with designing the Schema for the data I will also write a J2EE app so that 
data may be easily entered (as part of an applications deployment procedure 
for example) and also so that you can also get useful views (via XSLT) of the
environment ("what applications live on this host?" "what apps are using an 
old version of the JDBC drivers?")

vii. Questions
 I've already begun work on trying to represent the data and I'm not sure
when (if at all) I should be using attributes.  I know there is a lot of
discussion about this - but are there any good "rules of thumb" that you can
point me at?