Hello Denny,

Thank you for your reply and for the references.

I agree with you about the lack or redundancy and the need for
RAID configurations and redundancy in a production environment. We are
using this configuration in a test scenario
to validate whether we can fulfill one specific deployment with
reduced resources (we don't need redundancy for this
test).

The database has 1 million records on one specific table and up to 100
clients. Each client performs 1-2 transactions per second.
Therefore, the estimated transaction rate is 200 transactions/sec.
This database is heavilly used for searches
and we have search tables that are fully indexed (100+ columns and
corresponding indices). Also, full-text search
is enabled and heavilly used.

Our server has 8 GB RAM and is a dual Quad core Xeon. we are running
Windows Server 2003 and SQL Server 2005 Standard.
At this point, I do not have numbers on the cache hit ratio but it is
one of the indicators that we will measure. Is there any target
value that we should aim to?

Also, we will be measuring the average disk queue, since according to
the BOL, we should keep it below 2. Do you agree with this?

I like the configuration that you suggested but have three questions:
- Do you think we would we have any gains by creating our indices with
the SORT_IN_TEMPDB ON and installing tempdb in its oun drive?
- How would you configure the two disks that you suggested for the
application db? Two filegroups?
- Although we are not using at the moment, there might be a need to
use xml in this database for processing via XQuery.
This will require the addition of yet more indices (e.g. Primary xml,
secondary for path, property, and value)  to the database.
Would you change the suggested configuration under this new scenario?

thank you and kind regards
CD