Data Virtualisation and DWH Virtualisation
I presented a new style of content today at the Sydney DAMA March event, and thought it may be worthwhile to post the transcript as a paper. It is all about virtualisation your entire Data Warehouse, what is needed to achieve this and what it means for ETL and what role emerging Data Virtualisation techniques play here. In a way it’s the overarching story that supports recent blog posts about NoETL.
The premise is that we basically could do without ETL at all, or at least we can architect the systems to support this. It’s no goal of its own but certainly interesting to think about, especially considering how we can use views (therefore plain SQL) to even stage data delta (CDC)! The software I have been posting about is very near completion; basically all views can be generated to support a complete virtual DWH.
The full transcript is available here.