Creating Analysis Services Cubes - SQL Server Analysis Services 2012 Cube Development - page 117

<Attributes>
<Attribute>
<AttributeID>Calendar Year</AttributeID>
</Attribute>
</Attributes>
</Dimension>
<Dimension>
<CubeDimensionID>Promotion</CubeDimensionID>
</Dimension>
</Dimensions>
</Aggregation>
</Aggregations>
</AggregationDesign>
</ObjectDefinition>
</Create>
As you can see, the aggregation design contains only one aggregation, Aggrega-
tion 0 , which aggregates the Calendar Year attribute. We can easily extend this
aggregation design XMLA to include additional aggregations and additional attrib-
utes. When you assign the aggregation design to a given partition, you're simply
changing the partition's metadata so that Analysis Services knows which attributes
to aggregate when you process the partition. To create the actual aggregation files,
we must use the ProcessIndexes option, which rebuilds both index and aggreg-
ation files. Note that the partition must already contain data before you can use the
ProcessIndexes option to add indexes and aggregations.
Using BIDS Helper for customizing aggregations
The Analysis Services user community has published a number of very useful free
tools through www.codeplex.com . One of these tools is BIDS Helper. The tool is dis-
tributed with an installation wizard and takes only a few clicks to integrate it with
SSDT. BIDS Helper offers much useful functionality, and I recommend that you be-
come familiar with the tool. BIDS Helper allows you to customize aggregations using
Aggregation Manager.