Finding the right data warehouse appliance is largely a matter of due diligence and proper testing, according to two instructors from business intelligence (BI) consultancy, and Kim Stanick, vice president of marketing with ParAccel Inc., a San Diego, Calif.-based analytic database vendor. Schiff and Stanick taught a course on demystifying data warehouse appliances at a recent TDWI conference in Orlando.
Stanick said it's important to set up data warehouse appliance testing environments ahead of time, and those environments should take into account how things change over time because of such factors as additional users or BI subject areas.
"Doing your due diligence means scoping your benchmarking and proof-of-concept process very well and then setting up your standards for those evaluation scenarios so that you don't basically get manipulated by vendors who would prefer not to show you their warts," Stanick said.
There are many vendor and analyst opinions about what a data warehouse appliance actually is, but the technology can generally be defined as a combination of integrated hardware and software designed specifically for analytical processing, according to Schiff.
The idea behind data warehouse appliances is that an architecturally integrated package of hardware, database technology and storage that makes use of massively parallel processing is uniquely optimized for query processing. And data warehouse appliances have evolved significantly from humble beginnings.
"Initially, some of the very early data warehouse appliance vendors came in there and said: 'Let's solve a specific purpose with maybe a departmental or functional solution.' They got in there and they had some real quick wins and for the most part they were successful," Schiff said. "Now they're moving to where they're powerful enough in many cases to take on [enterprise-wide] data warehouse characteristics, where they can do multiple subjects areas."
Data warehouse appliances have taken heat from technology purists who like a great deal of freedom to customize. But generally speaking, Stanick said, the idea of customizing data warehouse appliances kind of misses the point.
"Data warehouse appliances are an attempt to standardize more of the stack so you have to customize less," she said. "If you start to customize these things heavily, you're not benefiting from what standardization can do for you."
Data warehouse appliance providers include newer, pure-play vendors like Calpont Corp. and Greenplum, as well as well-known database management system vendors like IBM and Oracle.
Data warehouse appliance shopping: Things to remember
When getting ready to shop for a data warehouse appliance, the first question to ask yourself is whether or not the application you plan to use it for is mission critical, according to Stanick. If it is, she added, you then need to decide whether your organization can stomach the idea of putting a relatively new technology in a mission-critical environment.
When evaluating vendors, all the standard criteria apply, she continued. Check to see whether the vendor has positive references and, more important, whether there are any reference customers that are trying to do the same sorts of things as your organization. Other standard questions to answer include whether or not the technology is in the vendor's "sweet spot" and whether or not the vendor is eager to win your business.
"Some people can take advantage of the fact that if they're willing to take on a little bit of risk or innovation, they can actually help it to grow attention for them in their industry," Stanick said. "[And] any emerging technology vendor is often very willing to build out their product to support the handful of customers that are so dear to them."
Next, she said, check to see whether the data warehouse appliance in question conforms to your standard operating environment -- and make sure the vendor is willing to do a free proof-of-concept demonstration.
"Even the really established vendors that are now trying to offer an appliance product really should be willing to participate in a proof-of-concept to show you exactly what they can do," she said.
Finally, check into the viability of the vendor, because newer vendors sometimes have a tendency to go away.
"[You don't want to end up] with dinosaurs that are really hard to get maintenance and stuff for," she said.