Long ago and far away I used a simple database built on bsddb. It was fast, simple and designed for our specific application. The problem was that we wanted replication. We wanted to make multiple copies of our data.

IWADB (I Wrote a DataBase) is meant to be similar to that store with the biggest difference being you get replication. There isn't a crazy query language or anything. No specialized indexes. Just a dumb REST interface.

What is that? Why even use it?

Whoa pard'ner! You're not supposed to actually use IWADB! It is something I did for fun that I doubt has much practical value. Maybe someday, but that day isn't today.

The only reason I can imagine IWADB could ever turn out to be something valuable is that it uses some powerful systems.

Rather than use bsddb, we're using lmdb. It is a great key value store that has a nifty feature of ordering its keys. Lmdb is really fast and well tested.

For replication, we're using Kafka. Kafka is a really slick durable queue. You can configure it to keep as much stuff as you want and it is able to scale out to many instances without much trouble. IWADB was written to have an excuse to play with Kafka.

Last but not least, we're using CherryPy! CherryPy has the benefit of being a good citizen when it comes to processes and it has excellent support for HTTP.