There is currently a lot of talk around high performance, web-oriented, databases. This blog from hosting vendor, Rackspace is a good introduction. Other informative sources of material about the so-called “NoSQL” movement can be found at Vertica and Amazon’s Simple DB.

However one high performance engine that never seems to be mentioned in this movement is Apache’s Lucene. Extremely fast, more mature and with a bigger developer community than most other engines, with the right modeling, Lucene can be an extremely effective “dimensional” database.

We will post some examples of dimensional modeling soon but if you want a quick look at how Flint can help to load XML into Lucene, see this page on Google Code.