I really like that Walmart/Sybase example, because Walmart is actually running the largest commercial data warehouse in the world including 2 years of detail data with tens of billions of detail rows. Of course, it's not using an OLTP system like Sybase/Oracle, it's a decision support database, Teradata.
-- Dieter Noeth, May 14, 2003
I would dispute that: http://www.wintercorp.com/vldb/2003_TopTen_Survey/TopTenWinners.asp Shows that France Telecom has the largest DSS system in Oracle. Walmart is not in the top-ten list, and surprise, surprise, the squashed competitor Kmart is.
-- what ever, January 7, 2004
Your comments about Sybase are naive and incorrect. Perhaps you had a bad run dealing with Sybase support, not sure if I can influence it otherwise. Sybase IQ for years now has been the bane some of the World's Largest data warehouses. Query performance and scalability are notably the highlights of all Sybase IQ implementations. Sybase customer, comScore Networks, received the Grand Prize in the 2003 Winter Corporation TopTen Program for Largest Database Size and Most Rows/Records for Microsoft Windows-based systems using Sybase IQ, the highly scalable analytics engine. Other Winter Corporation TopTen award-winning Sybase customers in UNIX categories include Nielsen Media Research and Korean-based customers Health Insurance Review Agency (HIRA), LG Card, Samsung Card and Chohung Bank. http://www.sybase.be/belgium/press/20031111-ipg-IQwintercorp.jsp I appreciate that you have given an opportunity to comment on this page.
-- Subraya Pai, May 9, 2004
Crystal Reports is not the reporting tool usually choosen for ad-hoc querying of datawarehouses, so maybe that's the reason your end customers weren't very happy about it. Tools better suited to this task are BusinessObjects (who also aquired CrystalReports a couple of month ago), Brio (aquired by Hyperion about 1 year ago), Microstrategy, Oracle Discoverer and Cognos. All of them allow you to build metadata about the datamodel of the datawarehouse and present the end user with the world in terms known to them (no criptic database table column names, predefined filter conditions and so on). For the end-user it's really only a matter of dragging and dropping the "objects" in their report and "pressing" a button. The tool will the generate the proper SQL, query the database (some of them even rewrite the query if you have aggregate tables, allow you to "join" in the report query results from different queries and databases or take a stored procedure as the data source) allow you to do simple computations (excel-like) on the result set etc.Regards, Georg
-- Georg Breazu, November 15, 2004
this article considered as a good article in data warehousing.I want to say that there is a query language called multidimensional Expression MDX .It is now the standard query language for OLAP.it is like Sql but with more capabilities in Grouping and more functions to facilitate OLAP operations.The most common function are ROLLUP and CUBE>>>>>>>>>>>>>>>>>>>>>>>>>>>etc. thanks to all.....
-- drtech dr, December 22, 2008
good n useful article..tnx for sharing..
-- mitesh trivedi, February 9, 2010