Note: You must be registered in order to post a reply.
T O P I C R E V I E W
Posted - 06/11/2013 : 01:07:18 This is a general OLAP query I'd appreciate any ideas on. It'd vastly increase uptake of OLAP reporting in our production system.
In order to improve performance of cubes with lots of textual attributes, those attributes are being enclosed in hierarchies and removed from selection as row attributes. They appear as columnar attributes instead and so aren't considered for sort candidates. This works very well until a report is designed with column attributes such as Month. All column attributes are repeated for each unique Month, cluttering up the report with duplicate static attributes.
For an example see "Product Sales Amount by Country Month and Year" on the test server. The hierarchy attribute 'Product Line' should report once for each Product Model Category, however being a columnar attribute it repeats for each Month shown. The only measure required is Internet Sales Amount.
If Product Line is shown rowwise it becomes a sort candidate, once several such attributes are selected the report's performance degrades significantly. Showing Product Line as a non-repeating column attribute would be very useful if it's feasible..