Data Warehouse Appliances? Me too!

Just as every presidential candidate this cycle is the candidate of Change, it seems that all the DBMS vendors offer the preferred data-warehouse appliance solution. That's the message I heard from appliance panelists at today's TDWI Washington DC chapter meeting. For a couple of them it was a real stretch, which in one case wasn't a bad thing. The net take-away is that we are seeing Change in the DBMS world, even if for the politicians that word is still only a promise.

TDWI-DC's panel consisted of Doug Cardin from IBM, Victoria Eastwood from Infobright, Phil Francisco of Netezza, Foster Hinshaw of Dataupia, and Rita Sallam of Oracle. Now my definition of DW appliance is a packaging of processor, storage, operating system, and DBMS that is optimized for data warehousing. A scalability model is essential. And only one of the represented companies hits the mark: Netezza, with an asterisk for IBM.I don't have much to say about Netezza other than that they have convincing customer stories and their financials demonstrate market success. So I'll move on to the others. First the asterisk vendor, IBM:

Doug Cardin started by asking if attendees were aware that IBM sells a DW appliance. He wouldn't have asked if he hadn't known that very few people would raise their hands. We didn't disappoint. IBM is a hardware-software-services firm that is so big they couldn't fail to have an appliance. (I'd bet HP has one too.) Big Blue's is IBM Balanced Warehouse, "the complete data warehousing solution comprised of pre-tested, scalable and fully-integrated system components of DB2 Warehouse, Server, and Storage." DB2, like Netezza, sports a scalable MPP (parallel) architecture. Sounds good, but if a tree falls in the forest and noone's around, does it make a sound? IBM's caught in that koan.

Dataupia calls their Satori Server a DW appliance, and company founder Foster Hinshaw should know: he defined the field as co-founder of Netezza. Yet Satori isn't self-contained. In a December meeting I had with him and Marketing VP Samantha Stone, Hinshaw said their goal is to "enable MPP for any database, in an appliance form factor, at [an appropriate] price." The on-line product overview refers to "existing database infrastructure," and that's the rub: do we have a DW appliance here or a query accelerator? But whichever Satori is, it's far, far more appliance like than anything I see that Oracle has to offer.

Oracle's panel rep was in solid me-too territory. Oracle couldn't possibly admit that they don't have a solution for any problem or category remotely connected to databases. And Rita Sallam didn't have anything close to an answer to Foster Hinshaw's contention that the performance of shared-memory, SMP architectures like Oracle's doesn't scale for large-volume data warehousing.

Infobright's Victoria Eastwood was the other odd-person-out on the panel: Her company sells neither hardware nor a free-standing DBMS. Instead, they sell a column-store engine for MySQL, which (otherwise) has been a decided innovation laggard, especially where data warehousing has been concerned, for a couple of years now.

I asked TDWI-DC board member Cheryl Hannan why non-appliance-vendor Infobright was on the panel, and her answer made sense: "Infobright is positioned to compete against the others. They're trying to solve a similar business problem." I'll cite one distinction, relayed to me by company CEO Miriam Tuerk in a late-January briefing and reiterated by Victoria Eastwood: that Infobright's Brighthouse is designed to counter "the rigidity and inflexibility of the data warehouse," that they're about "ever-evolving and changing analytics." Tuerk says that if you have predictable queries and transactions, a product like Netezza might be best for you. She also told me that they run up most often against Vertica and Greenplum in competitive situations.

What of Greenplum? Greenplum CTO Luke Lonergan was slated to appear on the panel but had to cancel. Greenplum does partner with Sun in an appliance deal so Lonergan would have fit right in.

The companies that were most missed however were Teradata and DATAllegro. Folks who follow data warehousing know Teradata, so I won't comment on the company or its products. I like DATAllegro too although I don't know them as well as I should. I will relay a comment made by a Netezza sales guy present at the TDWI chapter meeting: that DATAllegro doesn't sell true appliances because they integrate disparate components, not all their own. For example, DATAllegro uses the Ingres RDBMS rather than a custom DBMS like Netezza's heavily modified fork of PostgreSQL.

By the Netezza salesman's self-serving criterion, Netezza, IBM, and Teradata would have the appliance label all to themselves: whatever it takes I suppose, if you're a salesman and need to differentiate and class your company with the big guys. The TDWI-DC panel showed however that, fortunately, we have many more than three appliance options to meet the market's disparate data warehousing needs.Just as every presidential candidate this cycle is the candidate of Change, it seems that all the DBMS vendors offer the preferred data-warehouse appliance solution. That's the message I heard from appliance panelists at today's TDWI Washington DC chapter meeting. For a couple of them it was a real stretch, which in one case wasn't a bad thing. The net take-away is that we are seeing Change in the DBMS world, even if for the politicians that word is still only a promise.