When I saw this article on real time database virtualization, I was expecting something along the lines of a new way that we might provision resources for database instances or some new performance improvement using Hyper-V or VMWare. Instead it's something much larger, and potentially much more exciting.
The drive for real-time business intelligence is something Microsoft, Oracle, and many other companies are working on. They are looking to build products and sell you solutions that might somehow enable you to make quicker, more informed decisions. Or in the case of SeeWhy, perhaps your systems will make those decisions for you be reacting to data.
The database virtualization that's being talked about, however, seems more like something that we might see in a Hollywood movie, with databases being updated in real time, from a variety of sources, and providing only a limited set of information. Our applications in this real-time BI world will get data from a distributed set of databases that cache data in memory, yet somehow we are able to get a unified view of our data.
It sounds neat, but also unrealistic. The advantages of reacting quickly, of using information to your advantage, at least for any particular company, come from their ability to not only analyze and distill lots of information, but also to keep their competitors away from it. If I were to build the best retail sales analysis tool in the world, and then sell the analysis and not the tool, how do my customers get an advantage over each other. Best Buy and Circuit City don't want to buy analysis from the same company, they want to do things differently than each other. Information is power and sharing information across some grid like we do electricity isn't likely to happen. And without that sharing, I'm not sure how many of these big, grid connected systems smaller companies will build.
I do like the idea of more agile systems, of using column-oriented databases where appropriate and row-based relational databases where they make sense. Experienced BI professionals realize that it's not just indexing strategies that make a difference in an OLAP type system, but fundamentally different structures and even storage engine and indexing internals that provide better results.
At some point we'll likely have more agile, adaptive systems that will make decisions for themselves on various optimizations. Perhaps with some limited amount of autonomy to reconfigure themselves to match changing workloads. I still think, however, that there will be increasing numbers of DBAs to manage and monitor these systems. However intelligent we make them, we always seem to find situations that require a human to analyze and then help adapt the system to perform at its best.
Steve Jones
The Voice of the DBA Podcasts
The podcast feeds are now available at sqlservercentral.mevio.com to get better bandwidth and maybe a little more exposure :). Comments are definitely appreciated and wanted, and you can get feeds from there.
or now on iTunes!
- Windows Media Podcast - 30.2MB WMV
- iPod Video Podcast - 25.9MB MP4
- MP3 Audio Podcast - 5.3MB
Today's podcast features music by Everyday Jones. No relation, but I stumbled on to them and really like the music. Support this great duo at www.everydayjones.com.
I really appreciate and value feedback on the podcasts. Let us know what you like, don't like, or even send in ideas for the show. If you'd like to comment, post something here. The boss will be sure to read it.