In-Depth

Informatica Overhauls PowerCenter

Version 7 features new Web services-oriented architecture and supports for LDAP and CWM

• By Stephen Swoyer
• 11/12/2003

When Informatica Corp. officially exited the analytic applications market in July, officials announced a return to the company's roots, so to speak, with a renewed emphasis on enhancing its bread-and-butter data integration products.

Since then, the company has made several significant data integration-related moves, first announcing its SuperGlue metadata management system in late August and then, in September, acquiring mainframe data specialist Striva.

But Redwood City, Calif.-based Informatica will make its most important data integration move to date this week, when it announces version 7 of its PowerCenter data integration platform. The new release represents a substantial overhaul of Informatica’s flagship extraction, transformation, and loading (ETL) suite, featuring a new Web services-oriented architecture (SOA) and support for standards such as the lightweight directory access protocol (LDAP) and the Common Warehouse Model (CWM).

“Customers will now be able to use standards-based APIs, like Web services, Common Warehouse Model, and LDAP, to adapt our product inside their environments, not just for data integration, but for everything from data integration to business intelligence,” says Sanjay Poonen, senior vice president of marketing with Informatica.

The rub, of course, is that an organization’s application infrastructure must already be Web services-enabled, and an organization itself must also have Web services development talent in-house. According to Wayne Eckerson, director of research at TDWI, that’s not such a significant problem, however, because most organizations will continue to exploit the low level PowerCenter APIs. “[M]ost folks want to use the lower-level APIs to exploit most functionality, [but] Web Services is becoming [the] lingua franca in many organizations that need to integrate apps from many vendors,” he writes.

What’s significant, he notes, is that Informatica is building a “developers network” around its Java and Web services interfaces that is similar to Microsoft’s own MSDN. “Web Services means that [Informatica] workflows can be deployed on any platform, not just the one it was developed on,” he observes.

Elsewhere in PowerCenter 7 are new data cleansing capabilities that, Poonen says, are enabled courtesy of technology Informatica licensed from data quality specialist FirstLogic. “It allows you to cleanse data inside an integration engine, through an embedded dictionary of data cleansing,” he comments, comparing PowerCenter’s new data cleansing capabilities with the spell check feature found in most word processors. “We don’t own the [name and address] dictionaries, we OEM them from [FirstLogic], but it gives us name and address cleansing that’s built into our word processor, so to speak—PowerCenter.”

But the most important component of the new PowerCenter is Informatica’s homegrown data profiling tool, which the company has embedded in PowerCenter’s integration engine. “There is some advantage to integrating profiling within an ETL tool—the profiling tool can use the ETL tool to connect to data sources, and integrate those data sources so they appear as one file to the profiling tool,” notes TDWI’s Eckerson. “Most profiling tools aren't directly linked to data cleansing or ETL tools, so you only do analysis in profiling and then you have to copy the rules somehow and insert them into the cleansing or ETL tool. This is how it should work!”

Poonen says that Informatica has overhauled PowerCenter’s development capabilities as well. “We were surprised at how little sophistication ETL tools have provided their developers for configuration management, version control, the kinds of things that developers have from Visual SourceSafe, Microsoft’s product,” he indicates.

As a result, PowerCenter supports team-based development capabilities, along with check-in/check-out and version control features. Eckerson says that this is a welcome feature, even though organizations don’t typically support a large number of ETL developers. “This is important to elevate ETL to a more rigorous development process. It's a foundational piece that customers are very excited about,” he says.

Among other features, PowerCenter 7 features improved security as well, with built-in LDAP authentication and a RSA support for encrypting traffic over WAN connections.