3

I am working with Analytics app which can build and process DM and Cube. User can select dimensions/attributes/measures/hierarchies they want to see in Cube.

My cube loading/processing flow when I first time create and load cube

1) Create DS, DSV, dimensions, cube structures 2) Process Full dimensions 3) Process Full cube

After that user may change dimension structure (add/delete attributes) and I have to update and reload cube second time, flow:

1) Update dimensions, cube structures 2) Process Full dimensions 3) Process Full cube

I have performance issue when I do Process Full second time on dimensions because it reprocess related measure groups and partitions.

Any suggestion I can change my flow structure will be appreciated

Andriy Kuzmych
  • 193
  • 2
  • 11

1 Answers1

2

Simple yet efficient change, as per Microsoft recommends, always "process data" then "process indexes" instead of "full process". Reference : http://msdn.microsoft.com/en-us/library/cc966525.aspx

Dominic Goulet
  • 7,983
  • 7
  • 28
  • 56