Select Page

Source: Pixabay

I recently attended a course run by Hans Hultgren on Data Vault Modelling. I have a small confession to make at this point; Sorry Hans, I’ve never read your book. The good news for me was that the course doesn’t require you to have read the book first. It does assume some basic understanding of Data Vault concepts which the introductory videos included as part of the course materials takes care of it if you do not have that all ready.

The course was structured with a lot of practical team work as well as instructor lead material. Our class as a whole had a good grasp of the basic Data Vault concepts and data modelling in general. Hans took this into consideration when presenting the material and allowed plenty of time for deviations, modelling our real world issues instead of sticking religiously to any pre-baked script. It also helps that he is passionate about the subject and really knows his stuff, giving clear and complete answers to tough questions when asked. He is also still paying attention to what is going on in the rest of the Data Vault world, happy to ignore the printed material in his own course to discuss what is now best practice.

I personally knew just enough about Data Vault to be dangerous before I went on this course. After attending the course, I have a much better and correct understanding of Data Vault and its application. It also helped me understand where you could use Data Vault and how it fits into the much bigger picture of Agile Business Intelligence (BI). I won’t be getting technical about explaining what Data Vault is as there’s an internet full of information that could do a better job. However, Data Vault provides a design pattern to help take care of the boring stuff of data warehousing (history, audit trail etc.) and allows you to concentrate modelling the Business View. It does require a change in thinking, particularly that your presentation (dimensional) layer does not have to include everything. It only needs what is required to answer the user questions, now. In reality this means you need to run your Data Warehouse projects using the Agile methodology or the user will demand everything up front in fear of not getting another delivery. There is an overhead with Data Vault, but this is countered by being able to work better under Agile which should mean that your Data Warehouse has a better chance of real success by delivering what people actually want.

Ben

%d bloggers like this: