One of a Kind: Teradata Touts Active Data Warehousing

Only Teradata does ADW, officials say, and ADW isn't just marketing-speak.

Legendary rock and roll promoter Bill Graham used to say that the Grateful Dead aren't (just) the best at what they do, they're the only ones who do what they do.

Another Graham -- namely Dan Graham, senior marketing manager with Teradata Corp. -- likes to make a similar claim about his company.

Graham heads up Teradata's active data warehousing (ADW) practice, which he says is about as different from vanilla data warehousing -- or data warehousing as it's practiced by all of Teradata's would-be competitors -- as most bands are different from the Grateful Dead.

Similarly, Graham says, only Teradata does ADW, and ADW, he insists, isn't just marketing-speak. "Active data warehousing is … where you take the operational data store [ODS] and put it inside the warehouse. When you do this, you're able to load data in near-real time or real time and you're also able to [run queries against this data] and maintain a sub-second response time, so it looks a little like online transaction processing, but in fact it's read mostly," he explains.

Although this might not sound particularly complicated -- would-be Teradata competitors Dataupia Inc., Greenplum Software Inc., Netezza Inc., ParAccel Inc., and Vertica Inc. all claim to achieve similar sub-second performance -- there's a lot happening underneath the covers, Graham contends.

"This is not simply plopping an ODS workload inside the data warehouse. It's rather getting to the point where we're mirroring operational data as it's being used in the operational systems with the historical data," he points out.

The real difference between Teradata and its would-be competitors is one of scale, Graham insists, reprising a familiar Teradata line: Netezza got its foot in the door by pushing supercharged data mart systems designed to address tactical problems. Its competitors followed suit, targeting tactical Big Data issues in specific vertical markets. Teradata, for its part, pushes an ADW scheme that emphasizes the benefits of a unified enterprise data warehouse (EDW).

"What we're saying with Teradata is that you can actually have this one machine do all of these things in a single box, therefore you don't have to manage all of these ETL streams [feeding into separate data marts]," he claims.

The salient point, Graham argues, is that none of Teradata's would-be competitors -- not Netezza, not Hewlett-Packard Co. (HP), not Oracle Corp., and not IBM Corp. -- can counter its ADW pitch. He cites Teradata's query-level support for mixed workloads, a differentiator he claims is matched by only two other vendors -- in very different market segments.

"There's only a handful of machines in the world that can do this. The IBM mainframe has done mixed workload management for 30 years. The Tandem machine can do it, but it's still mostly stuck in the investment [banking] space. I can assure you that what … [HP is marketing] with Neoview is not a mixed workload [solution]," Graham maintains. "Teradata [supports] mixed workloads at the query level. We're the only ones at the decision-support level doing this. We're on our fourth major generation. We've been doing this since the early 90's."

That's why Graham and other Teradata officials like to describe the other players in the DW segment as asymmetric competitors: they don't use the same weapons, bring the same forces to bear, or attack (issues) in the same ways, he points out. It's also why Graham -- and to a degree, other officials -- concede that there just might be a place for Teradata's asymmetric rivals. It was largely thanks to pressure from Netezza (and the recent proliferation of Netezza-esque analytic database technologies) that Teradata last year introduced its first appliance systems. It's likewise thanks to pressure from Netezza and others that Teradata -- long an EDW-or-bust proponent -- is increasingly willing to countenance the existence (if not the desirability) of data marts, too.

"I think in most cases the competition has made us stronger. I believe a couple of our competitors have made us change direction and made us come up with better offers for our customers," Graham acknowledges.

"We count more than 150 database vendors now. Five years ago, if you went to something like the Gartner Magic Quadrant, you'd see eight," he continues. "They're not all going to survive. We know that. … We know that this is innovation in progress. Their solutions often have a trick -- a capability -- that we respect: they're good at what they've done. When you get right down to it, however, they're a one trick pony. They're a technology solution and nothing more."

Of course, Teradata is a technology solution, too, Graham concedes, but he distinguishes between empirical (i.e., field-tested) and speculative (i.e., cooked up in a lab) technologies.

"This [mixed workload] stuff really is rocket science. This is the brainy stuff that you really have to be good to implement this inside the machine. Then, after you do that, you have to suffer through real customer implementations for a while, because the customers throw things at it you never imagined," he argues. "So [a mixed workload capability is] something that has to mature over time. You just don't cook this up in the lab … and release it as a finished product."