Microsoft to build common data store

Microsoft Corp. is working to use the data store and engine currently in its SQL Server database across several of its server products.

The idea is to make the products work together more effectively and to eliminate overlap wherein several of the products have different data stores and engines, according to Barry Goffe, Microsoft group product manager in the .Net enterprise server group, in Redmond, Wash.

“We’re always looking for ways to make (the products) work better together,” Goffe said.

Microsoft is starting this project with its Exchange e-mail server, SharePoint Portal Server, and a forthcoming operating system.

“We’re rebuilding Exchange on top of the SQL Server data store,” said Bill Gates, Microsoft’s chairman and chief software architect, at an event last week to honor people who don’t work for Microsoft but volunteer their time on the Internet helping others work through problems with Microsoft products. “One of the great Holy Grails inside Microsoft is the vision of a unified file system.”

Gates continued that the strategy is to move from five or six data stores in separate products to one single data store. “This is probably one of the toughest things we’re doing,” Gates said.

Gates added that Microsoft is working with the next generation of Windows, code-named Longhorn, to make the SQL Server data store part of the file system.

Stan Sorenson, Microsoft’s director of server marketing, said that the company is also working to unify its SharePoint Portal Server with the search engine in Yukon.

Yukon is the code name Microsoft uses for the next version of SQL Server. Sorenson said that Microsoft will issue the first beta version of Yukon in the second quarter of next year, and ship the final version in the first half of calendar 2003.

“We look at our (SQL Server) storage engine becoming a storage engine for all of our products,” Sorenson said.

Peter Urban, an analyst at Boston consultancy AMR Research Inc., likened the idea of a common data store to Oracle’s iFS (Internet File System), a centralized data repository.

“If you have a centralized repository, you have one place to search against. That makes searching easier and more efficient,” Urban said.

Microsoft’s Goffe said that the common data store strategy is part of a broader effort to combine the realms of structured and unstructured data.

Indeed, Microsoft’s database rivals IBM Corp., Oracle Corp., NCR Corp., and Sybase Inc. also are working to merge structured and unstructured data sets with the ultimate goal of tapping into a more comprehensive knowledge base including information residing in Word, Excel and e-mail documents.

AMR’s Urban added that by making a common data store, Microsoft will be effectively enabling its other applications to harness the power of a relational database, thereby making the applications themselves more scalable. Also, it will lead to better interoperability with other applications because of the XML technologies in SQL Server.

“This would address some of the security issues Microsoft faces. Storing information in a database is more secure,” Urban said.