#datavault vs #starschema for EDW purposing

Once again my friends, someone has decided to put an interesting entry up asking the age old question: why is Data Vault Model so important? Along with several other questions, they admitted they cannot see the value.  However, the author posted his words not in an attacking type of manner (as others have done in the past), but rather in a manner which leads to open discussion.  For this, I applaud him.  I have responded by commenting on his blog – which is currently awaiting approval.

However, you can find his entry & my comment (once approved) here:

https://thebibackend.wordpress.com/2012/06/05/thoughts-on-data-vault-vs-star-schemas/

So what’s the Data Vault Advantage?

I will do my best to just list the advantages (again) as I have so many times before:

  • Combining the Data Vault Model + Methodology leads to a workable, and flexible enterprise data warehouse
  • The Data Vault Model is flexible, and highly scalable
  • The Data Vault Model is easily adaptable to change with little to no impact to existing history
  • The Data Vault Model is pattern based
  • The Data Vault Model is “modeled” after over-simplified node based neural network
  • The Data Vault Model separates Keys from Descriptors From relationships
  • The Data Vault Model relies on Tuple and Set based Mathematics for near-linear scalability (because of the separation)
  • The Data Vault Methodology is based in SEI/CMMI Level 5, contains parts of TQM, and Six Sigma
  • The Data Vault Methodology is measurable, and quantifiable in terms of risk, and level of effort to build
  • The Data Vault Methodology is tool agnostic
  • The Data Vault Methodology is infrastructure agnostic
  • The Data Vault Methodology is componentized/modularized
  • The Data Vault Methodology is repeatable
  • The Data Vault Methodology is automatable
  • The Data Vault Methodology is risk averse
  • The Data Vault Methodology is independent parallel team-based
  • The Data Vault Methodology is consistent – what you do for one warehouse component, you do for ALL
  • The Data Vault Methodology is self-documenting
  • The Data Vault Model & Methodology are proven to work!

The long list of huge world-wide corporations using the Data Vault model and methodology today, are testament to these facts.

Why o why is not “BIGGER” outside of parts of Europe?

Well, like I said before: it has to grow in popularity.  Kimball wasn’t “bigger” when he started either, and no-one wanted to even hear the term Data Warehouse when Bill Inmon started talking about it in the 1970’s – they all thought it was hogwash.  Give the Data Vault time, you’ll see it’s popularity grow.

In the US, it is just now catching on with consulting firms and large corporate clients.  It’s also catching on in Canada, Mexico, Australia, China, and Indonesia.

Do a job search around the world, and you can and will find companies looking for Data Vault Modeler and Methodology skill sets.

Anyone have thoughts to add?  I’d love to hear your comments.

Thank-you kindly,
Dan Linstedt

Tags: , , , , , , , , , , ,

No comments yet.

Leave a Reply

*