Abstract

The project cirkulates around the idea of a storage format, shaped like a table, that's not bound to a specific platform or a specific filesystem and that has capabilities to work as a cloud service. This storage format is also a bit more advanced than a regular table because it has three dimensions.

The idea came from a lecture about Google app engine where it was explained that google uses a self made system called BigTable for storing large amounts of data. This project is similar to BigTable but is built from scratch and has key features such as being open source and is based on Java for platform independency.

The purpose of this project is to create a simple platform for data storage that can be used as a cloud service and where the underlying technology is able to cluster over a wide variety of systems and hardware. Using small means you are supposed to be able to create a data storage where the traditional file server is phased out and the modern cloud concept is used in its place. It would in theory be possible to create a global cluster with almost unlimited storage capacity though in practice it may be more suitable for private clouds. Also it doesn't matter if you only have one server or several hundreds of them, to the user it will function just the same.

This project takes on the problem of solving data storage in a economical, scalable and usable way. It also solves the issue of versioning data, something that similar services don't offer, take good money for or only offer in a limited form.