John W. Colby
jwcolby at colbyconsulting.com
Thu Sep 2 00:01:55 CDT 2004
And in fact I won't even need to recalc nightly since the data simply doesn't change unless I make updates or merges. John W. Colby www.ColbyConsulting.com -----Original Message----- From: dba-sqlserver-bounces at databaseadvisors.com [mailto:dba-sqlserver-bounces at databaseadvisors.com] On Behalf Of Arthur Fuller Sent: Thursday, September 02, 2004 12:45 AM To: dba-sqlserver at databaseadvisors.com Subject: RE: [dba-SQLServer] VLDBs, the saga - Ramblings only Almost by definition a data warehouse is about as denormalized as you can get. Typically it is way bigger than the OLTP database from which it stems. It creates an un-normalized database from the central fact table and calculates all the totals for every dimension. The more dimensions, the larger the table, but the advantage is that everything is precalculated so nothing has to be derived ad-hoc. In the ideal world, you refresh the OLAP db every night when the managers are asleep and next morning the data is fresh again. Theoretically :) A.