IBM Updates Information Integrator

Code-named Masala, the product is expected to ship in the fourth quarter, said Nelson Mattos, IBM Distinguished Engineer and director of Information Integration for IBM Software.

With this release, IBM is ratcheting up the volume around more "autonomic"--or self-monitoring, self-tuning and self-managing--functions. The product is designed to help DBAs figure out the most efficient way to cache data from disparate sources so it is more readily available to queries. Users also can expect big query speed improvements, with queries running up to 10 times faster than the current release, Mattos said.

One feature, Design Advisor, helps users figure out which caching scheme works best, while a new interface facilitates linkage to multiple legacy systems, where much data resides, Mattos said.

IBM, Armonk, N.Y., continues to pitch Information Integrator as a cornerstone of its federated data strategy, which is designed to enable customers to keep their data where it already resides yet tap into it more easily for data mining and analysis.

"Most integration projects require a lot of integration for deployment, and we're starting to see a large amount of our business driven by integrators who build solutions for retail, the financial sector, etc.," Mattos said. "They are definitely interested in new autonomic benefits that make their lives easier."

Masala's new interface takes some of the mystery out of what existing systems are in a given implementation, according to Mattos. "It lets you ask in English what is out there," he said. "The benefit is shorter deployment times. You get into SAP, PeopleSoft and Siebel without a lot of hand-coding."

Pat Hayes, data warehousing project lead at IndigeTech, a Richmond, Va.-based integrator, said Masala could help businesses extend and expand on the useful life span of their existing mainframe data. "[Many companies have] VSAM files or Adabas files that have been around for 15 years. The question is, how do we unlock them? Do we continue building that platform, or do we move everything off?" he said. "Now [with Masala] you don't have to move that data. You can use that as a repository and tap into it as needed using familiar Windows tools."

The autonomic stuff also is key, Hayes added. "As an integrator, I like to focus on a bigger solution than just making the connection. [The new] data-mapper tool makes it very easy to find and link to data sources, or you could use tools that are part of the databases to create the mappings."

Mattos declined to comment on pricing or packaging, but the current DB2 Information Integrator comes in several flavors. It starts at $5,000 per CPU for the replication edition, which is aimed at users who just want to replicate data across multiple databases. The standard edition adds federation capabilities and runs $15,000 per CPU. An advanced edition, which includes all of the above plus DB2, costs $40,000 per CPU. Connectors for tapping into non-IBM repositories are extra, but enterprises also can opt for an "unlimited" edition, including all necessary connectors, for $125,000 per CPU.

In the overall database market, Oracle remains in the top slot with 39 percent market share last year, up 8.6 percent year over year, according to research firm IDC. IBM was second with 31.3 percent share, up 5.1 percent, and Microsoft came in third at 12.1 percent, up 14.7 percent. The database market was up 7.6 percent overall, IDC said.

Oracle's big bet on Linux is paying off, according to company Vice President Robert Shimp. "We're very pleased with the progress we're seeing in Linux. Growth there is very strong," he said. "A lot of customers are moving from Unix to Linux, but we are also seeing migrations from Windows But Windows continues to be popular."

In late May, research firm Gartner posted its own take on the database market, showing Linux to be the high point in an otherwise flat field.