Information Lifecycle Management Lives, Finally

Talking Points

  • Information lifecycle management seems to be catching on finally, because enterprises must adhere to compliance regulations.
  • The big idea is to take data that must be readily available and store costly systems, and to move data that must be stored to meet regulatory requirements on slow but inexpensive systems.
  • Enterprise ILM initiatives today are in varying states of disarray, now doing tiered storage, and that's where SRM comes in.

We have about 100 terabytes on our NAS, about 500 terabytes on our SAN and 1.5 petabytes of tape backup,” says Chris Wintheiser, a systems engineer at TLRTS Enterprise Storage. TLRTS is the IT unit of Thomson’s Legal & Regulatory business unit.

Information lifecycle management seems to be catching on finally, mainly because enterprises must adhere to compliance regulations. “ILM has been mostly vendor hype” until recently, says Brian Garrett, an analyst at Enterprise Strategy Group.

“Whenever we do anything, we do it in terabyte sizes,” Wintheiser says. “And more and more, we need to be working within a compliance framework.” Perhaps most impressive of all is the size of the team handling this massive storage load: eight full-time employees.

As you might imagine from an enterprise specifically focused on legal and regulatory issues, compliance is the biggest factor in TLRTS’ storage picture. “The biggest issue is backups, restoring and archiving solutions,” Wintheiser says. To meet regulatory dictums, the company’s users “want to keep seven years’ data and archive [it] offsite, and there are problems with that.”

He points out that because of the pace of change in today’s technology world, “even if you could archive it, seven years later, when you want to recover [the data], you’re not even using that technology anymore,” Wintheiser says. “And most database solutions don’t backlog seven years back, so how reliable is that data anyway?”

ILM’s big idea is to take data that must be readily available and store it on fast but costly systems, and to move data that must be stored to meet regulatory requirements on slow but inexpensive systems. ILM provides management of information, starting with the data’s creation and extending through securing the information—updating, maintaining, archiving and, ultimately, deleting it.

To deploy ILM, enterprises must first understand and classify their data, then store it according to factors such as value, risk and the likelihood the information will be needed in a hurry.

The job of deciding what to move when and where is invariably automated, using storage resource management and similar tools. SRM manages the storage infrastructure—hardware, switches, files and file systems—and optimizes the efficiency with which available storage space is used. The goal is to automate tasks—data collection, backup and recovery, SAN performance analysis, virtualization, provisioning, forecasting and so on—which storage administrators had to be perform manually until recently.

SRM is essentially a prerequisite to ILM. That’s how enterprises are using SRM: to get a full picture of their storage infrastructure. And it’s usually a chaotic picture, cluttered by myriad undocumented device types, data types and practices.

“Enterprise ILM initiatives today are in varying states of disarray,” Garrett says, “but people are now doing tiered storage. That’s where SRM comes in.”

“Any enterprise that’s experienced growth will have data splattered across different storage devices and platforms,” says Kris Domich, principal consultant at Dimension Data, an IT consultancy. “SRM by itself is not a magic pill. Companies are using it to take in a holistic view of all data across the enterprise—how it needs to be classified, its retention period, its termination point.” (See related story, “Information has a shelf life.”)

Keeping the business running
Although compliance is a major force behind the growth of ILM and SRM, it’s hardly the only factor. “We also see a lot of [storage] demand driven by business-continuity models,” says Pam Taylor. As VP of SHARE, the venerable (and largest, with more than 20,000 members) IBM user group, Taylor gets an earful on the topics most affecting member companies.

Post-9/11, Taylor says, enterprises determined that simple disaster recovery isn’t enough. “We see an emphasis on geographically dispersed mirror sites,” she says, “and naturally that has a big impact on storage.” Additionally, analysts point to increasing reliance on imaging as a driver.

Whatever the drivers, many ITorganizations are taking their first hard look at SRM software. They have a multitude of options: many large enterprise application vendors, such as Computer Associates International, offer SRM tools, as do EMC, IBM and Hewlett-Packard. In addition, there are SRM applications from specialists including Storability, AppIQ, FalconStor and CreekPath Systems.

Benefits aplenty
Businesses that use SRM software to get an ILM initiative started often do so because of compliance issues, but they tend to find plenty of other benefits. “With SRM, you…ID the data you need to manage in a compliant form, and then you need to manage it,” Domich says. “By IDing, reclassifying and moving data to the appropriate device, you tend to make space where you thought you were running out of space.”

For example, Domich says that when Dimension Data clients closely examine their storage picture as part of the baseline-creation process, they’re often stunned at how much non-essential data resides on their high-speed SAN—the most expensive flavor of storage. Shifting such data to a more appropriate medium, be it a mid-level SAN, a NAS, disk or tape, not only creates a more logical structure—it saves money. “We find our clients can often delay storage purchases 1 to 3 years,” Domich says.

Star had its own motive for implementing SRM. The $80-million managed services company, which is based in the U.K. and has about 4,200 customers, “saw an opportunity to package IBM’s on-demand capability,” including storage, and resell it, says Richard Ellis, Star’s VP of IT. The company opted for IBMTivoli SRM tools because they scaled effectively and allowed Star’s IT group to effectively manage such small parcels of storage, according to Ellis.

Last year, TLRTS Enterprise Storage opted for Storability’s Global Storage Manager, switching from IBM SRM software—which in TLRTS’ environment could hang during server reboots, Wintheiser says, due to the IBM host/agent model. So far, Wintheiser says, scalability and customizability are the big benefits of the Storability tools.

Because Storability’s data repository is a simple Microsoft SQL database, “You can go in and generate your own report,” Wintheiser says. That was TLRTS’ most important requirement, he adds, because “custom reporting is a huge need for us. When you’re generating reports, you can’t anticipate every [type of] report or field you would need. Every [business] unit has unique needs and forms. They also have application data that’s unique. We need to be able to correlate trends and peaks, put it all on a timeline and see what’s happening on a CPU at that time. So if the repository is closed to any other tool, it’s not useable.”

At TLRTS, a new Storability configuration-management tool has quickly become a favorite of Wintheiser’s group. “It lets you correlate the DB, file system and server metrics on a single page,” he says. “Because we’re a small group, we can’t be creating reports for all development groups, so we need the ability to pull metrics off everything. Then, [development groups] can look at any metrics anytime they want.”

As ILM becomes a standard part of the IT landscape, more businesses will turn to SRM software. “The big value of SRM right now is to help you understand your storage needs and do an assessment,” says Enterprise Strategy Group’s Garrett. “The majority of both the market and the software’s potential is still to be tapped, but we’re starting to see awareness reach down.”

Sidebar: Information has a shelf life

Tools to Master the Sarbanes-Oxley Challenge
By Alan R. Earls

The Search for Extra-Relation BI
By Stephen Swoyer

Managing Unstructured Information
By John K. Waters