I am trying to grasp this nosql thing since it sounds good for scalability.
I have read the Digg is now using CassandraDB. Can you help me understand how a digg-like model can be structured efficiently ?
If you're trying to quickly get a grasp on the major trade-offs involved in NoSQL systems, check out this blog post I wrote: Visual Guide to NoSQL Systems. Essentially, you must choose two of the following properties for you system to excel at: consistency, availability, or partition tolerance.