Analysis Services - 07 Dimension Attribute Relationships скачать в хорошем качестве

Analysis Services - 07 Dimension Attribute Relationships

So, let get this straight... we have gone to the trouble of defining our Dimension Hierarchies to ensure that they are nested correctly and display correctly but they still wont work correctly for aggregation? YEP! This little tid bit is missed by so many self-taught people! What are Dimension Attribute Relationships ? They are the roadmap as to how aggregations are structured in cubes. Without Attribute Relationships, Analysis services would not know how to perform fast and accurate aggregations. For example, using the Date Hiearchy we have been using... How does Analysis services know to obtain the total year sales it only needs to add two numbers together Semester 1 & Semester 2. How does Analysis Services know to add the first 6 months into semester 1 and the second lot of 6 into semester 2? It doesn't! That's the whole point of Attribute Relationships this tells Analysis Services specifically the levels of aggregation and also it can then work out how best to calculate the values above. Remember from earlier we are wanting to view data from the most generic to most specific and to get that kind of optimisation we need to define the "roadmap" of how the aggregations get stored. That's what Attribute Relationships is all about!