IBM Flexes XML Muscle

By Brian Fonseca  |  Posted 2005-01-03

IBM Flexes XML Muscle

In an effort to give customers more focused control over internal and non-IBM external enterprise data stores, IBM is planning to add native XML support to its relational DB2 Universal Database wrapped around new indexing capabilities.

An expanded beta test for the enhanced DB2 technology, which is currently available in alpha format for a few select IBM customers, is planned for the first half of this year. No official timetable has been set for general release of the new technology—which may be offered as part of an integrated DB2 upgrade or as a stand-alone product, according to Janet Perna, general manager of data management solutions at IBM Software Group, in Armonk, N.Y.

As the proliferation of SOAs (service-oriented architectures) continues, XML will play a prime role in business process operations due to its ties to content management, analytics, information integration and Web services. As more XML is created and its flexibility requirements grow, Perna said, IBMs new DB2 tools will let users mix and match patterns and data with native relational and stored native XML. As a result, queries will be able to encompass both of those data types across applications in the same database.

Click here to read about IBMs strategies for service-oriented architectures.

"We dont believe that all information will ever be in one place, and whats going to be important is providing companies with the ability to hide the complexity of information and where it is with integration middleware," said Perna. "We want IBM databases to be the natural choice for all these new XML applications that are going to be written."

To achieve maximum XML performance, bolstered indexing attributes in the technology will enable advanced search functions and a higher degree of filtering. IBM is also adding support for XPath and XQuery data models. This will allow users to create views that involve SQL and XQuery by sending the protocol through DB2s query optimizer for a unified query plan.

Dave Richards, chief technology officer for The Research Libraries Group Inc., of Mountain View, Calif., said his organization uses XML to store bibliographic data, which requires a great deal of auxiliary table construction to search and access records. Richards said the native XML support improvements for DB2 will allow digital records to be stored in a preparsed form, enabling streamlined and more efficient searches.

"We wanted to be able to support queries that just were based on information in the e-records that had not been indexed. The way we have to do that at the moment is not terribly efficient," said Richards. "[Native XML support] is going to enable us to store things more compactly and access them easier ... and make it easy for us to be able to ingest and then export data in XML when were able to migrate to that version of DB2."

Richards said RLG is implementing a 1.5TB database featuring 140 million records representing books, serials, maps, films and music scores.

Next Page: IBM differentiates itself on XML.

IBM differentiates itself on


PureEdge Solutions Inc., which adds audit trails, security and self-archive features to XML electronic forms, plans to integrate DB2s native XML support with its code base as soon as it becomes available, according to company officials.

"From our perspective, there is no choice. This is something we have to offer to customers and play with because it unlocks a lot of value from our own solutions," said Mike Rowling, senior solutions architect for PureEdge, of Victoria, British Columbia. "Organizations have lots of structured and semistructured data that is not always predefined for how youre going to use and access that information and how youre going to leverage that through the organization. This is going to be pretty monumental."

Analysts say IBMs heavily federated stance toward native XML support will be a major differentiation from approaches from other large DBMS vendors such as Oracle Corp. and, to a lesser degree, Microsoft Corp. Over the last few years, DBMS vendors have provided users with an XML support layer that maps or "shreds" an XML document onto an underlying relational storage model.

At its Oracle OpenWorld conference last month, Oracle officials said its Database 10g Release 2 will feature support for XQuery to access XML data.

Responding to criticism from customers that had difficulty using SQL Server 2000s XML mapping technology to get back stored nonrelational XML documents, Microsoft is planning to add native XML data type and corresponding data structures to query support to its SQL Server 2005 product, code-named Yukon, which will debut in the first half of this year.

Yukon allows users to register an assembly inside the database and store a CLR (Common Language Runtime) user-defined type or take an XML representation of an object and store that in a database, said Michael Rys, program manager for SQL Server Engine for XML Technologies at Microsoft, in Redmond, Wash.

In addition, SQL Server 2005 will feature a modifiable subset of the XQuery standard, which has yet to be officially signed off on by the World Wide Web Consortium, Rys said.

Check out eWEEK.coms for the latest database news, reviews and analysis.

Rocket Fuel