TDWI | Training & Research | Business Intelligence, Analytics, Big Data, Data Warehousing

IBM Touts All in One Information Integration Suite

IBM says its Information Server slices, dices, and delivers data of all kinds, from almost any source, and in almost any conceivable format.

• By Stephen Swoyer
• October 25, 2006

In a sense, IBM Corp.’s Information Server announcement event, held last week in Anaheim, Calif., had all the trappings of a late-night television infomercial—including the backing of a seven-piece band. Big Blue officially proclaimed that Information Server quite literally slices, dices, and delivers data of all kinds, from almost any source, and in almost any conceivable format.

It might sound like a pipedream—after all, IBM positions Information Server as a kind of Philosopher’s Stone for data integration, capable of transmuting just about any kind of information (structured, semi-structured, or unstructured) into something that can be exposed as part of an SOA and consumed by other service-enabled applications.

Pipedream or no, IBM officials point out that Information Server draws upon Big Blue’s own demonstrable expertise in data integration—which includes not just ETL, data quality, and data profiling (by virtue of its acquisition last year of the former Ascential Software Corp.), but also structured and unstructured content management (via its WebSphere Information Integrator tool, which incorporates assets from a passel of other acquisitions), document management (by way of its acquisition of the former FileNet Corp.), and, of course, service-enablement, thanks—once again—to its ubiquitous WebSphere-branded integration middleware stack.

That’s just the tip of the proverbial iceberg, of course. IBM officials say Information Server combines technology assets developed across 55 of its global software labs, along with software from its Venetica, Unicorn and CrossAccess acquisitions, in addition to Ascential.

So what is IBM Information Server? Not your grandfather’s—much less your enterprise data architect’s—trusty old ETL tool, for one thing. Officials bill it as the first all-in-one platform that’s able to facilitate heterogeneous information integration, enrichment, and delivery.

And they don’t shy away from saying as much. “Information Server is the industry’s first comprehensive unified foundation for enterprise information architectures, capable of scaling to meet any information volume requirement,” said Michael Curry, information integration program director with IBM, in a Webcast presentation.

At last week’s Information Server launch event, IBMers touted a bevy of potential use cases, such as compliance, enterprise-wide business intelligence (BI), and customer data integration (i.e., creating a single view of customer data) efforts.

The big bogeyman, IBM officials say, is information SILO-ing—the isolation of critical data in dozens (and sometimes hundreds or even thousands) of data sources, in a range of different formats, connected to an array of different platforms (of varying vintages), and often geographically distributed, at that. In spite of the promise of ETL, enterprise application integration (EAI), enterprise information integration (EII) and a host of other polysyllabic technology acronyms, SILO-ing remains a problem.

Enter IBM’s SOA-based Information Server, which more or less encompasses them all, providing all-in-one connectivity, cleansing, and delivery capabilities. Nor is Information Server an untested commodity, IBM officials insist: it’s already been through two beta programs (with more than 75 clients and partners).

Information Server has been a long time coming. In February, for example, IBM announced plans to ship it in the second half of this year. As it stands, Information Server still isn’t generally available; IBM plans to ship it next month. Much of the work stems from the fact that Big Blue’s inaugural all-in-one platform isn’t just an amalgam of separate IBM solutions: it’s a completely integrated offering, a bona-fide integration architecture, said IBM VP of information management strategy and marketing Tom Inman, at last week’s event: “This is a brand-new architecture, not a marketing bundle.”

Analysts seem to like what they see in Big Blue’s Information Server push. For one thing, notes James Kobielus, a principal analyst with consultancy Current Analysis, it’s based on established (and in many cases) best-of-breed technology assets—including Ascential’s ETL and data quality capabilities, as well as Information Integrator’s EII expertise.

In this respect, Kobielus notes, Information Server offers both new and incremental enticements, including feature and functionality updates for Big Blue’s re-branded WebSphere DataStage and Quality Stage tools, along with new modules that support collaborative data stewardship.

In addition, Information Server bundles a metadata repository and adapters for a number of data sources. “IBM’s launch of IBM Information Server was a necessity for the vendor to maintain its leadership in the enterprise [data integration] and [data quality] markets. With this upcoming release, IBM is first-to-market with a comprehensive, SOA-enabled DI/DQ product suite,” Kobielus points out. “IBM provides a strong value proposition for its existing DI customers to upgrade to IBM Information Server. The new suite provides many important feature enhancements over and above previously separate point products.”

There are some concerns, of course. For one thing, while IBM has been first to market with its all-in-one integration platform, other vendors—most notably Oracle Corp., Syabase Inc., and perhaps even Informatica Corp.—will no doubt follow suit. “IBM’s first-to-market advantage will be short-lived, as Sybase, Oracle, and other [data integration] rivals are expected to quickly follow with their own rollouts of increasingly feature-complete DI suites,” he points out.

Will the all-in-one Information Server push signal the end of standalone ETL? Kobielus says no, noting that purveyors of standalone ETL or EII tools can plausibly claim that the value of such all-in-one data integration suites—like the all-in-one BI platforms that rose to prominence last year—might be overblown, given that customers continue to favor point solutions for project-, system-, or application-specific purposes. “At the same time,” he concludes, standalone ETL or EII vendors “should explore mergers, acquisitions, cross-licensing, and other arrangements whereby they can individually or collectively address a broad range of point requirements.”