Is someone knocking on your door pitching the second generation of the Oracle/Sun Exadata appliance, waving benchmarks and promising the best database performance ever?

I mean as Oracle CEO Larry Ellison declared in September 2008, when the first generation arrived built on HP hardware, it is, after all, on its way to being the most successful new product launch in Oracle’s 30 year history.

Hmmm. Remember your parents saying if something sounds too good to be true it probably isn’t? Well, as everyone in the IT world knows, that applies to technology as well as an enticing new car deal.

Then there’s the saying, if it ain’t broke don’t fix it. And that is exactly the thought that should pop into mind when a vendor is trying to convince you to buy something so new.

If you’re running an IBM database environment and it’s working well, what’s the enticement to spend millions on something not well-known beyond all the marketing hype?

In the first year of sales, Oracle has reportedly has shipped 20 Exadata machines, and many, even as Ellison noted, admitted were ‘test drive’ deals. While Ellison put some names to customers in a recent earnings discussion, the tally was under a dozen and no specific sales figures were provided in Oracle press materials.

Just as elusive is the actual cost for Exadata hardware, software, deployment, integration work and support services. As one tech writer noted in 2008, the software price tag was initially around $1.6 million alone, with HP hardware coming in around $650,000.

DBMS2 put an estimate of $5.5M for an HP Oracle Database Machine and noted lots of confusion around actual pricing given the software on the services side, number of extra Exadata cells that might be needed and how much data actually fits on a cell.

And if you do take time to review the benchmarks Oracle’s waving, you need to bring a magnifying glass to do the needed examination given parameters are, well, a bit unique. As one industry expert notes many of the Oracle initialization parameters used in this benchmark “don’t make much sense.”

Oracle is big on saying what Exadata can potentially do, but it hasn’t mentioned much what it can’t do (and every technology has its limitations), and there doesn’t even seem to be any sort of upgrade path for those few enterprises that may have jumped on the first generation.

But if Oracle’s marketing hype does have you wondering if your database environment is doing all it can do, you’ll be interested to hear why some IBM customers chose not to jump to Exadata when that knock came on the door.

Stay tuned in the following days as I’ll be providing some telling insight, directly from customer experiences, about why IBM customers closed the door on the Oracle/Sun machine.