Slashdot Mirror

I thought the whole reason why NoSQL is "better" than SQL is it's based on column based storage, while most SQL databased are row based storage. Couldn't you make a column-based database that uses SQL as a query language? There is nothing wrong with SQL as a language, there are just some workloads where column based storage is faster (mostly data analytics).

There are a number of column-oriented SQL databases. Another commenter pointed out Sybase IQ; I'll point out Vertica, and leave you to Google others if you want.

Trash SQL in favour of coding all your data access needs. Welcome back to 1973, guys.

It's not like we could do parallel SQL in the 1980's. Or that you can't do parallel SQL in a compute cloud today.

No, It basically seems like they don't want to pay software vendors any money for database technology. That's mostly what the arguments boil down to. Oracle RAC is very scalable, arguably easier to do at massive scale than MySQL - but you have to pay Oracle money. For an Internet startup, I can understand why you'd take your chances with "roll your own". For an enterprise... I think not.

> I don't know why this article is so harshly critical of MapReduce.
> Are these guys just trying to stake a reputation based on being critical of Google?

Um... yes?

The Database Column is being coy about being a corporate blog for Vertica, a high performance database database product, but in fact it is. Vertica is a commercial implementation of C-Store and was founded by Michael Stonebraker, the most prominent proponent of column based databases (get it? the database column). So yes, they have a very good reason to be hostile to Google.

http://www.vertica.com/company/leadership
http://en.wikipedia.org/wiki/C-Store
http://en.wikipedia.org/wiki/Michael_Stonebraker
http://www.databasecolumn.com/2007/09/contributors.html

1) They don't look like hammers,
2) They don't work like hammers,
3) You can already drive in a screw with a hammer,
4) They aren't good at ripping out nails, and
5) They aren't good at driving nails.

Brought to you by The Hammer Column, a blog written by experts in the hammer industry, and launched by Hammertron, makers of a revolutionary new kind of hammer.

Vertica launches database-focused blog: The Database Column.

The column was copyright by Vertica. Wouldn't they be concerned about the type of competition that MapReduce presents?

Actually **his** article is fine - he is simply saying the current horozontal approach of RDMBS is not as efficient as vertical for data wharehousing applications. He is saying the concept "one size fits all" is obsolete as data wharehouses provide different stresses than OLTP applications. Nowhere does he say that RDBMS as a concept is obsolete or anything like that.

It's the moron at computerworld that misread his article and thought that he's saying that vertical is superior to RDMBS, which complete nonsense (that's like arguing TCP is better than Ethernet - meaningless!). Stonebraker is arguing about the on-disk layout used by common RDBMSs and saying what they currently use is less than ideal for some applications, which I can appreciate. He argues that his product, Vertica, outperforms conventional RDBMSs for datawharehousing.

As proof that the computerworld monkey is the moron, Vertica itself is described as "a brand-new column-oriented RDBMS". Yeap, Vertica is a RELATIONAL DATABASE. Sheesh.

Also, I don't think it's news that Michael Stonebraker (a great name, by the way), co-founder and CEO of a company that (surprise!) happens to develop column store database software, thinks that column store databases are going to be the Next Big Thing. Right or wrong, his opinion can't exactly be considered unbiased...

It's actually worse than that. From the linked article:

NDOVER, Mass., Sept. 5, 2007 - Vertica Systems today launched a unique, multi-author blog that offers database architects and administrators, CIOs and other IT professionals expert insight and opinion about database evolution and emerging technologies. The Database Column (www.databasecolumn.com) features weekly contributions from six well-known database visionaries discussing enhancements that improve database performance and scalability, decrease database administration overhead and system costs, and help organizations make more informed decisions.

The entire blog on which it appeared is a corporate shill for Vertica, the makers of a column store database. Which isn't to say that what's said there is incorrect. It may not be. But there should be absolutely no expectation of lack of bias.

There is an article, and it has many references. How is a 'Captain Obvious' sort of comment labeled Insightful? The insightful part is in the article. The first author, Michael Stonebraker, architected Ingres and Postgres. He looked at OLAP databases, which is a market that is much larger than a special case. He proposed storing the data in columns rather than in rows. He tested this, it works. In fact it works so well that he can clobber a $300,000 server cluster with a $800 dollar PC. I know that I would be pretty happy to spend a year porting to his database if I could pocket half of that annual hardware cost savings. The savings in electricty would be enough to pay for several pretty serious Starbucks addictions. His key insight seems to be that he can vastly improve OLAP performance by storing the data in columns rather than in rows. This change could be quite transparent to the end users & developers, except for the massive speed-up and cost savings, of course. This paper describes a general solution for a common problem. Stonebraker has developed Vertica , which is still support ad-hoc querries in SQL. This seems like a pretty general purpose solution for OLAP.