Posts Tagged conformed dimensions
Yesterday, was the first day in the new job. There as a small bit of the “new boy at the new school” about it, but it was not a big part of the day. Most of the day was consumed with meeting people (I hope I start remembering all those names), and a couple of external activities (a physiotherapy appointment to continue sorting out my back, and another with my GP to puck up some more repeats on some of the prescriptions).
The major achievement for the day was getting a work pass. The feeling that I’ve now established my working identity in the new organisation as definitely there. Which follows on from the “Betwixt and Between” blog post.
The new job, to start with, looks like being a data warehousing project. The design and implementation of data warehouse is a class of IT problem which I’m very well equipped to tackle. class of The replacement of a SuperWeb solution, with a Microsoft SQL Server Analysis Services solution also greatly appeals to me. I’ve not previously had an opportunity to implement a Microsoft solution (my previous organisation being heavily an Oracle shop). Also, the appeal is doubly satisfying as removing SuperWeb is very satisfying (the previous organisation did use it, and it was a choice I never supported, for multiple reasons).
The data warehousing project exposes a number of big question in my data warehousing style, approach and understanding. Which I’ll have to work through with the new organisation and the analytical clients of the data warehouse. These include:
- Data management principles. I’m not sure how mature this organisation is in understanding the benefits of good data management principles. Some of those principles data management principles start to manifest themselves in the design decision which data warehousing solution must make.
- A Variables Dimension. Stylistically, I “like having one of these”. From an data management perspective, it allow the identification of the same measure across all analysis outcomes. The question is “Are the users of the analysis cubes ready for this?”. Another question I have is “How do Analysis Services, and Excel Pivot Table, react to having this form of construct”. I’ll have to wait and see on this one.
- Conformed dimension. These are cornerstone of the “Kimball Warehouse Bus”. They, from an analysis perspective, should allow for the pulling together of multiple cubes, and the standardisation of the label sets which also enables analysis. Another question is “How well does Excel Pivot Table, work with this construct?”.
I’ll continue to blog about these concepts, and the progress of the implementation project. I’ve a bit of evangelising and training to to to spread some understanding of the approaches which I’ve detailed above.
Today, I’m off to a Scrum Project management training course. This training will prove to be interesting. I’ve had an interest in this style of project management for a while. So, some formal training on “How to do scrum” will be useful.