I'm creating a data mart in SQL Server 2008 using SSIS for load, and SSAS for an OLAP cube. So far, everything is working great. However, I haven't created any indexes on the source database other than the default clustering on primary key.
I'm pretty comfortable with designing indexes on the application databases, but since this database is intended primary to be the source for a cube, I'm not sure what sort of indexing, if any, will be beneficial.
Is there any sort of indexing I should be doing to improve the processing of the dimensions and cube? I'm using your regular Molap storage.