Indexing option on numeric fact referred to in DW Toolkit 2nd edition
2 posters
Page 1 of 1
Indexing option on numeric fact referred to in DW Toolkit 2nd edition
In the 2nd edition of THE DATA WAREHOUSE TOOLKIT, Chapter 9, page 208 at the end of the first paragraph, there is a reference to an indexing option on a numeric fact pioneered by SYBASE IQ product which has become a standard indexing option in competing databases. I would like to know what is the equivalent indexing option in Oracle?
dennisf- Posts : 9
Join date : 2009-04-14
Re: Indexing option on numeric fact referred to in DW Toolkit 2nd edition
I believe it is referring to bitmap indexes.
Sybase IQ is a 'column oriented' data base. It implements tables as a series of bitmap vectors for each unique value in each column in the table. The vectors are compressed, so overall, the database is smaller than the raw data it holds (if stored in a traditional row).
In Oracle, implementing bitmap indexes allow you to take advantage of the star query optimiser. Fact tables should be defined with bitmap indexes on each foreign key column (local indexes if the table is partitioned), one index per column. If all goes well, Oracle should resolve predicates against the dimensions, then using the list of subject dimension foreign keys, apply them en-masse against the bit vectors to isolate the rows in the fact table. It provides fast, consistant query times regardless of the query.
Sybase IQ is a 'column oriented' data base. It implements tables as a series of bitmap vectors for each unique value in each column in the table. The vectors are compressed, so overall, the database is smaller than the raw data it holds (if stored in a traditional row).
In Oracle, implementing bitmap indexes allow you to take advantage of the star query optimiser. Fact tables should be defined with bitmap indexes on each foreign key column (local indexes if the table is partitioned), one index per column. If all goes well, Oracle should resolve predicates against the dimensions, then using the list of subject dimension foreign keys, apply them en-masse against the bit vectors to isolate the rows in the fact table. It provides fast, consistant query times regardless of the query.
Similar topics
» crm issue in The DataWareHouse Toolkit 2nd edition
» Fact Indexing -SQL Server 2008
» The great SQL query "template" from edition 1 of Data Warehouse Toolkit
» Ebook The Data Warehouse Lifecycle Toolkit, 2nd Edition: Practical Techniques for Building Data Warehouse and Business Intelligence Systems
» Fact table with non-numeric measure
» Fact Indexing -SQL Server 2008
» The great SQL query "template" from edition 1 of Data Warehouse Toolkit
» Ebook The Data Warehouse Lifecycle Toolkit, 2nd Edition: Practical Techniques for Building Data Warehouse and Business Intelligence Systems
» Fact table with non-numeric measure
Page 1 of 1
Permissions in this forum:
You cannot reply to topics in this forum
|
|