Big stuff coming from DATAllegro
In the literal sense, that is. While the details on what I wrote about this a few weeks ago* are still embargoed, I’m at liberty to drop a few more hints.
*Please also see DATAllegro CEO Stuart Frost’s two comments added today to that thread.
DATAllegro systems these days basically consist of Dell servers talking to EMC disk arrays, with Cisco Infiniband to provide fast inter-server communication without significant CPU load. Well, if you decrease the number of Dell servers per EMC box, and increase the number of disks per EMC box, you can slash your per-terabyte price (possibly at the cost of lowering performance).
Note: Actual per-terabyte numbers, when available, are likely to refer to “user data.” I.e., they are calculated after compression. So any appliance vendor that has a software release improving compression can automatically claim a corresponding decrease in its per-terabyte price.
There are at least two obvious reasons for dividing your data warehouse/data mart contents into segments. One is that you might keep them on different boxes with different performance characteristics. Second, even assuming identical hardware, you might want to keep different kinds of workloads from clashing with each other.
The downside to such segmentation is pretty obvious too. Data movement latency can be a big problem. Even beyond that — sometimes you really do want to run a query that spans different data marts.
So wouldn’t it be nice if you could get the benefits of data warehouse/data mart segmentation, while mitigating the drawbacks of such an architecture? Well, suppose that nodes in different appliances could interoperate in some of the same high-speed ways that nodes in a single appliance do.
More next week.
Keep getting great research about data management and related technologies. Get a FREE subscription by RSS/Atom or e-mail!
Technorati Tags: DATAllegro, data warehouse appliance, compression, data mart
Comments
Leave a Reply