You're way better off dropping the index(es), loading the table, then re-creating, clustered first.

P

-----Original Message-----
From: Ellingson Cory (RCIS)
[mailto:mssqldba-ezmlmshield-x5019876.[Email address protected]
Sent: Thursday, June 30, 2005 1:47 PM
To: LazyDBA Discussion
Subject: Index question...


Hello all!

I recently was informed of a database I now get to support. I have Lumigent
Log Explorer, and I have it set up to tell me when certain commands happed
(new tables, dropped tables, dbcc's, new indexes, etc) and I get an email
daily at the same time about an index that is dropped and recreated. After
some looking, I found a SP that issues this command, and it is part of a DTS
package. It appears the reason for this is there is a SP that appears to
look for duplicate records to remove before the final step to refresh a DB
user table. This index is a clustered index; the table is rebuilt each
night from numerous other tables...

Now, the question...

Is it necessary to drop and recreate this clustered index over a table that
is built fresh each night, or will the index continue to resort as it is
written?

I want to get rid of this statement...I was thinking my next step would be
to do a DBCC DBREINDEX (<table>, '<index name>'). I first wanted to see if
I needed it at all.

TIA

Cory Ellingson


---------------------------------------------------------------------
TO REPLY TO EVERBODY , PLEASE CLICK REPLY-ALL, NOT JUST REPLY
Website : http://www.LazyDBA.com
To unsubscribe: http://www.lazydba.com/unsubscribe.html

MS Sql Server LazyDBA home page