At the Code Camp last week, Eric, Eric, and Josh did a live podcast of their weekly show that had a panel of some talented people working with SQL Server here in Colorado. It's an interesting debate and discussion and if you have time, kick it off and listen in the background.
One of the interesting things that was asked was regarding the scale out capabilities of SQL Server. These are rather limited, but there are a few, distributed partitioned views being the main one, that can let you add machines.
I have heard fairly often, and complained myself, why doesn't SQL Server scale out more like Oracle's RAC solution? I've heard various answers, but the answer given by one of the panelists, that there hasn't been a widespread adoption of RAC, so why should Microsoft invest in it, is one that I wasn't expecting. It does make sense as SQL Server exists to make money, not necessarily be the best product. If there isn't a large demand, then why do it?
I'm not sure that RAC is the best model for building a scale-out technology. And it's incredibly expensive, which is probably the main reason why I think it's not widely adopted. The complexity of managing an even more complex Oracle technology is probably another reason.
However I've heard quite often, and from many companies, how can we scale out SQL Server? There are many, many companies and people I've worked with that would love to add another 2 or 4 way server to their instance and have additional scalability for their SQL Servers. If I run Dynamics or some other COTS application on a 4 way box and run into performance issues, tuning is hard, and moving to a larger architecture, such as an 8 or 16 CPU box, is very expensive. If I could just add another 4 way server, that would be idea.
I'm sure there are any number of ways to build a strong grid database. ParAccel seems to have one idea, though they suffer from the same price tag issues as Oracle. If they, or Microsoft, would build a less expensive solution, I think you'd start to see much wider adoption.
My vote is that Microsoft should look at this like they did their early cluster solutions. Go for 2 nodes, work out the bugs in a few versions, and then expand this to larger nodes. This is a great time to begin deploying this type of technology, as data sizes get larger and larger, and we have more and more people working with our databases. We definitely need some way to serve all these people in an easy to deploy, understand, and manage solution.
And if you don't think there's a demand, think about the demand for Service Pack 3. I heard that no one was calling for that about 500 votes ago. Vote for Service Pack 3!
Steve Jones
The Voice of the DBA Podcasts
The podcast feeds are now available at sqlservercentral.podshow.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 - 34.8MB WMV
- iPod Video Podcast - 26.2MB MP4
- MP3 Audio Podcast - 5.4MB
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.