Data Modelling in MarkLogic, and how my mate Norm can help you!…

Data modelling in an aggregate/document NoSQL database can be unfamiliar to most. In this post I mention a couple of techniques and how they can help…

In both relational database systems and NoSQL key-value and columnar stores you have to take a logical application ‘object’ and shred it in to one or more structures that are flat. They are typically table based, with some NoSQL options supported one or more values per column, or some sort of built in hash mapping function.

Shredding poses a few problems for application developers. These mainly centre around them having to do work just to store and retrieve their information. As a lazy bunch (I used to be one, so I can get away with that comment!) they don’t like spending time doing this – they prefer to spend time on the interesting and unique stuff. This is normally an issue because organisations want these storage layers to be fast and reliable – so they force their developers (quite rightly) to spend time on this layer too.

Data Modelling in MarkLogic, and how my mate Norm can help you!…