At last someone we understand …

It is rare with all the background reading and research that I have to do that I read a contemporary book and say “Yes, this guy gets it” – notice that I say contemporary. There are plenty of good books about the day-to-day business intelligence problems that I see which were written over 20 years ago (The Mythical Man Month by Frederick P Brooks [1975] and Peopleware: Productive Projects & Teams by Tom DeMarco and Timothy Lister [1987] to name but two). So it was with immense pleasure that I read the opening chapters of Netezza Underground: The Unauthorized Tales of Derring-do and Adventures in Resilient Data Warehousing Solutions by David Birmingham[2008]. This book describes, in a hugely readable way, many of the same concepts and ideas that I have been espousing for the last 15 years about how to deliver a successsful data warehouse solution.

David describes the building of data warehouses as the need to think of terabytes, not transactional systems and a discussion of very-large-scale data systems. He says that the rules are different here and yet oddly the same:

  1. Everything is requirements driven
  2. Simplify and clarify
  3. Use correctly powered and scalable systems
  4. Governance
  5. Data management
  6. Strong architectural approach
  7. Building the environment with the expectation of change
  8. Testability
  9. Go Parallel
  10. Never do bulk inside a traditional RDBMS

I would love to drill into each of these in more detail but doing so would simply reprise an excellent book and probably breach copyright! And if the mention of Netezza as a vendor puts you off, then read the book anyway, because whilst the author is an avid enthusiast of the product (and some reviewers dislike the consequentially irreverent style), the approach, techniques and philosophy he describes provide much of the same advice that we have been giving our customers for so long and that can be applied to other technical solutions. Of course if you are interested in implementing a system we are here to help too!

This article was originally published on BIonRails, another Data Management & Warehousing website

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.