I am going to answer a completely different question than you asked, but it may help.

I have often been able to break these types of queries into 2 groups of users: simple 
and complex.

Many (upwards of 70%) of the users choose one of a couple "default" configurations. I 
handle these by querying once, and writing the emails in batch mode.

Once these are removed and shunted into a single query, I deal with the complex users 
as a separate list.

So I only need to do the complex queries on a third of my users, which saves a lot of 
time.

Just a thought,
Jerry Johnson

>>> [EMAIL PROTECTED] 11/05/03 12:18PM >>>
This is not a question for the HoF system but instead one that I'm passing
along.

Here's the problem. We have 4000 users getting email during the day and each
email is generated from a complex search that checks on about 300 events in
a day. Each email is unique and is based on one or more searches that are
specific to each user. No user has the same search criteria. Additionally,
this data is real time with a single digest a day for users.
This is rather stressful on the server and a solution is needed.
Additionally, we're dealing with a legacy DB system with 10 tables of 200
total columns joined together (don't ask). I've suggested having a second
machine with its own SQL server. This server will receive a replicated feed
from the main DB and be used only to generate these emails.
Is there a better way to do this? A better solution?
Thanks

p.s. SQL 2000 pro


~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|



Message: http://www.houseoffusion.com/lists.cfm?link=i:6:1711
Archives: http://www.houseoffusion.com/lists.cfm?link=t:6
Subscription: http://www.houseoffusion.com/lists.cfm?link=s:6
Unsubscribe: 
http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=<:emailid:>.<:userid:>.<:listid:>

                        

Reply via email to