Thanks for the replies.  I have a couple of additional questions.  My original concern came about when I compared the memory usage of the above SQL 2000 sever to a...

Tom, good suggestion on the one proc calling 25 - I agree.  Also, thanks for the info on the optimizer hint.  I hadn't heard of it before. RH

Yes, I have considered it.  The proc is a vendor supplied proc that sorts through about 25 versions of a query and executes one using an IF condition against the...

Nevermind.  I found that answer too.

http://www.sql-server-performance.com/rd_data_cache.asp 

RH

I found what I need in BOL under SQL Server Architecture: Execution Plan Caching and Reuse.

Now, does anyone know how to expose the aging metrics?  I would like to interogate...

sp_recompile has not been specified within the procedure, nor is anyone specifying a recompile manually.

The proc seems to get recompiled several times per day...not every time.

RH

Sold. RH

OK, so I found information on the net that says SQL Maint uses the standard DBCC REINDEX statement.  So, I lookup DBCC REINDEX in BOL (which it DOES list) but it...

Mark, yes, I know this can be solved using a set.  Although, based on the characterstics of the forms and the speed of the initial load, etc., the only way...

As a note...if I had any indexes whatsoever on the table I would assume the issue was with forward-pointers.  Without indexes, I assumed that forward pointers wouldn't be neccessary and...

Eric, I'm not sue I agree.  Dynamic SQL 100 times faster...really?  Are you simply trying to say T-SQL won't loop to 10 million quite as fast as truly compliled code. ...

I found some information on this that primarily cleared up my question.  It seems that because of the efficiency of plan reuse as of SQL2000, the PREPARE/EXECUTE model serves only...

It's very hard for me to say what the traffic volume will be.  This will be a very small ecommerce site with a very specific product.  Based on your statement,...

You're right, TLOG writes are sequential so I should alter my statement.  90% of my activity will be very small singleton reads/writes.  There will be very little activity with large...

I couldn't find a way to do this via an input mask so I just wrote a string maniplulation function to deal with these fields plus a couple of others.

Thanks...