Informatica Pushes ILM -- With an Application-focused Twist

With ILM, Informatica says it has the answer for harried data management pros. Analysts caution that Informatica's strategy will take time to gestate.

Pity the modern data management professional: he or she must serve many demanding taskmasters -- such as line-of-business (LOB) customers, C-level executives, and compliance and security officers -- without completely pleasing any of them.

LOB customers want anytime access to all of their data -- along with the ability to run a dizzying variety of reports or queries against that data regardless of cost or complexity -- while they also clamor for new features or functionality, as well as access to an ever-larger variety of data sources.

C-level types want real-time results, but always within the bounds of regulatory propriety. It's the same for compliance officers, who often want to restrict access to data -- particularly to potentially sensitive data such as information that could run afoul of any regulations. Security officers, of course, would probably prefer that all data were siloed, encrypted, disconnected from a network, and housed (permanently offline) under lock and key, or in a Sensitive Compartmented Information Facility.

Data management pros can't meet all of these demands, of course, but according to officials from data integration and information lifecycle management (ILM) specialist Informatica Corp., they can perform in most respects. Officials point to Informatica's acquisition of Applimation, a provider of enterprise data management software for Oracle (and its ERP applications) and SAP platforms, as just what the doctor ordered.

Applimation markets tools for data archiving and data migration, as well as for the rapid generation of representative data sets for testing, development, or training. It has special expertise on metadata, with what officials claim are best-in-class metadata extraction capabilities. According to parent company Informatica, the combination of Applimation's data management know-how with its own data integration solutions results in a one-stop shop for application-focused ILM.

"They've been a leader in the application lifecycle market for a number of years. Applimation specializes in pulling [transaction] data out of structured applications, many of which are packaged applications -- stuff like Oracle e-Business, SAP, PeopleSoft, Siebel -- but it can also be custom applications or applications that that you build on any relational databases," says Adam Wilson, vice president of product management and marketing at Informatica.

"[Customers need] to pull that data out to do a few things with it: they need to archive their data so they can lower the TCO [of their application infrastructure] by moving dormant data to lower-cost storage and by limiting the growth in relational database licenses. "One byproduct of this is improved application performance. Many of these packaged applications, when we start getting up to 250 GB to 1 TB [of archive data], start to really choke. It's not that the underlying databases are choking, it's that the applications themselves can't handle these incredibly large data sets or data volumes. Companies need a way to resolve that. Traditionally, what they've been doing is sort of addressing the symptoms as opposed to the root cause of the problem. Applimation gets them to the root."

Analysts like what they see. Ted Friedman and Carolyn DiCenzo of Gartner Inc., who -- in addition to noting that the earlier acquisition of the former Princeton Softech by IBM Corp. made Applimation's acquisition all but inevitable -- argued that Informatica basically got itself a $40 million bargain in Applimation.

"Applimation's solutions will enable Informatica to address its customers' data life cycle management requirements, as part of the broader data management discipline," they write, observing that Applimation's "archiving technology appeals to information governance needs, as life cycle management of data is another part of information governance."

There's also the question of Applimation's subsetting and masking technologies, which, the Gartner duo say, are just the ticket for overworked enterprise development or QA teams in the age of Sarbanes-Oxley. "Applimation's subsetting and masking technology will be useful in facilitating management of test data in data integration and application development/deployment work," Friedman and DiCenzo say.

Gartner sees the acquisition as a particular win for Informatica customers, who -- in addition to PowerCenter's vaunted ETL, data replication and data migration, and data federation capabilities -- can optionally tap Applimation's masking and subsetting capabilities, along with its metadata management features.

On the other hand, Informatica's ILM coup isn't yet a fait accompli. The company must first integrate and reconcile Applimation's technology, which -- with connectivity to many of the same data sources that Informatica itself delivers -- must inevitably entail some degree of redundancy. For this reason, although Informatica customers should "look for opportunities to leverage the Applimation technology to facilitate management of test data for data integration development activities and beyond," they must also solicit "product road maps for Informatica's integration of Applimation technology" with its PowerCenter data integration platform.

Informatica's Wilson says the question of redundancy between Applimation and Informatica is largely a non-issue. "From a connectivity perspective, Informatica has a broader set of connectivity than Applimation had as a standalone company, so we can connect to more things," he asserts. "As they move over to PowerCenter, we'll be able to bolt on a much broader range of connectivity than what [Applimation has] been able to deliver."

Synergy's the thing, contends Wilson, who outlines several scenarios in which a combined Applimation and Informatica stack can help organizations size and scale their information infrastructures. "If I'm sourcing a specific transactional application and I'm building a data warehouse on top of that, if I decide [that] that's growing too much or I need to create lean copies of product that are obfuscated for test and [development] purposes, I can use Applimation for that. It can generate cleaned or masked [subsets] of my data -- with sensitive information such as social security numbers taken out -- just like that," he comments.

"What's even better is that I can reuse a lot of that logic and that metadata to do the archiving and create the lean copies. The more you use the platform across more project types, the more value it delivers and the more it accelerates [your operations]."

In many instances, Wilson says, Applimation's customers were first attracted by its data archiving capability; over time, he argues, they started tapping it for masking and prototyping, too. "We're increasingly running into cases where folks are looking to create copies of production [data] for development testing or training purposes. If you're a large retailer and … you have 12 TB of transactional data, and you're doing four releases per year while maintaining separate development, test, and training environments for each of those releases, all of a sudden you have … 144 TB of data that you're moving around. The costs can explode."

There are also the requisite data privacy concerns, Wilson allows. "[In this scenario] you have full production copies where none of the data is masked or secured in any way, you start to run into legal questions," he points out.

Wilson also praises Applimation's metadata management features, which he claims are superior to Informatica's own capabilities. "Because [Applimation does] custom applications, because they have a focus on packaged application… they've built accelerators -- we [i.e., Informatica] call them accelerators, anyway -- that are application-aware, so they understand the metadata or the content that lives in the underlying applications. So when I say I want to mask HR data coming out of PeopleSoft, I don't have to pick and choose which is the most appropriate mask. They've already done that hard work of saying, 'Here's all of the things that we think you're going to want to mask because we already know where all of the sensitive data is.'"