We could also add data virtualisation but that would also cloud the water in what could be simplified as decision to automate or not. Recently, the additional possibilities arising from the big data echo system and the concepts of a data lake or swamp, the possibilities of schema on read (where data doesn’t necessary have to conform to a model on load) will add a layer of complexity to this discussion which we just don’t need. That is, a relational engine that stores and delivers agreed data. What I hope to achieve with this post is to materialise some of my thoughts on the matter and present a few ideas that may be considered by others (if they were considering utilising this technology).īefore we get into the grit of discussion, it is probably worth noting that I consider the data warehouse in the more traditional sense. I’ve come to realise that the last point (my reaction) plays on my mind because the subject is filled more with emotion than rational decisions. That’s just a long enough time for me to see the pros and cons, examine the state of affairs, consider the consequences of both camps and lastly consider my reaction to those consequences. I have been thinking about data warehouse automation for at least 2 years now.
0 Comments
Leave a Reply. |