Often times, we put our data into a relational model out of habit more than anything else. These projects force me to look at my data in alternative ways. (No worries: This isn’t a case of a solution looking for a problem)

• Tokyo Cabinet
• Project Voldemort : Replication, clustered data, java library
• Flared
• Oracle Berkeley db : On top of which SVN sits
• MemcacheDB
• CouchDB

Articles:

• Is the relational database doomed

I haven’t found much in the way of articles that talk of the best practices and patterns to use with these products, but if I find any, I’ll be sure to putthem on my delicious feed under the database tag.