SAP customers have been told that with HANA, SAP’s in-memory database platform, terabytes of their data can be processed without breaking a sweat. At TechEd we learned that even a petabyte of “raw, uncompressed data” isn’t a problem.

In fact, SAP execs say HANA has created a magical big data processing and analysis environment in which customers are limited only by their imaginations.

If you can dream it, we can do it.

But one thing that all customers approaching HANA must remember is that those use-case dreams will turn out to be an expense of nightmarish proportions if they don’t find a way to rationalize their data volumes before they move to HANA. That is because SAP charges for HANA based on volume—the more data you’ve got to process, the more expensive it’s going to be.

“Memory is just much more expensive than disk,” says Rob Jackson, a solutions architect at Dolphin, an SAP partner that specializes in information lifecycle management and BPM software and services. (Jackson gave a presentation at the ASUG Arizona chapter meeting in October, which I attended.)

Even without considering any HANA scenarios, data archiving should make a lot of sense to SAP customers who are seeing ballooning amounts of structured and unstructured data flows. But as SAP expands HANA’s footprint and capabilities—from standalone apps and BW on HANA to the core Business Suite on HANA in 2013—the need to reduce data requirements is a critical yet not-often-discussed best practice.

The economics of SAP HANA dictate that customers should shrink the amount of data they put into their HANA system, according to Jackson. This starts with “putting a value on their data”—creating a plan to archive older, seldom-used data that’s not critical to ongoing, real-time operations.

Resistance to data archiving appears to be widespread, however. During Jackson’s presentation, ASUG members in attendance bemoaned the excuses that they hear from department heads and users unwilling to modify their data hoarding ways. Jackson listed many reasons for resistance, including: a worry that archiving means deleting data (it doesn’t); that it’ll create “holes” in business document flow or violate corporate retention policies (no and no); that it’ll send them to “tax jail” (not unless they’re fudging the numbers to begin with); or that retrieving archived data takes too long (nope).

There are lots more, but they’re more crutches than reality, Jackson says. “There are technical solutions that overcome every one of these [concerns].”

“Lean Systems” Cut HANA Costs

Again, HANA conversations aside, the “economics of archiving” can result in substantial savings. “When you archive data, you’re not deleting it. You’re moving it to low cost storage,” Jackson says. “You can ensure that the business can access it if need be—and in many cases, the business won’t even know it’s archived.”

Archiving can compress data to 90 percent, according to Jackson, and cloud storage costs “pennies a gigabyte.” The savings, then, come from reductions in storage and hardware costs. And that helps when customers are eyeing HANA. “A lean system minimizes the cost impact when migrating to SAP HANA,” he adds. (For its part, SAP has stated“the more of HANA you buy, the cheaper it gets.”)

The requirements for HANA migration (i.e., “large, high volume systems”) include secure, long-term retention of data, transparent access to archived data, high-retrieval performance, scalability for high data volume and concurrent access, and low administration overhead, Jackson says.

While those are critical, he contends that the technical side of an archiving plan is less onerous than the administrative and political side of the change—such as sourcing storage products, conducting workshops and getting everyone to sign off on the plan. The goal is one uniform and repeatable policy, though turf battles are inevitable. The challenge, then, as stated by an ASUG member who’d been through it, was: “Harmonizing business-retention policies.”

It’s never too late to commence an archiving project, because if you’re like most companies today, you’re collecting more and more data every day—and you’ve probably got a ton of older data that can be archived. “It’s like blowing up a balloon,” Jackson says, “that data has got to go somewhere.”