I have 5 attributes and need to arrange them in different ways for
different analysis. They all connect to the fact using the same key.

Is it ok to do this in a single cube or should I have different cubes for
this?

The hierarchies I want are something like

Hierarchy 1
===========
attribute1
attribute2
attribute3
attribute4
attribute5
key

Hierarchy 2
===========
attribute1
attribute4
attribute5
key where key is hidden


Hierarchy 3
===========
attribute3
attribute5
attribute4
key where key is hidden


thanks very much

--
Message posted via http://www.sqlmonster.com

Theres no reason why you can't do this on one cube. However, there are
several ways to do it, a virtual dimension, a single dimension with operands
and MDX or a set of multiple-heirarchy dimensions.

Alternativly if you are using a front end like Proclarity the user may be
able to define their own calculated member groups.

Not sure whether using SQL Server 2005 (Yukon) is an option, but
multiple hierarchies can be defined for the attributes of a dimension.
From SQL Server 2005 BOL:

Attributes are the building blocks of dimensions, and are used in
Microsoft® SQL Server™ 2005 Analysis Services to organize measures
contained in a cube. Dimensions are collections of attributes, with each
attribute corresponding to one or more columns in the dimension table.
In a dimension, attributes are arranged in hierarchies that provide the
aggregation drill-down paths for measures contained in a cube. In client
applications, attributes are used to provide grouping and constraint
information.

Hierarchies are used in Microsoft® SQL Server™ 2005 Analysis Services to
hierarchically organize the members of a dimension. Attributes in a
dimension can be organized as levels in hierarchies, which provide
navigation paths in a cube. Hierarchies have one or more levels. The
members in the bottommost level of a hierarchy are also called leaf
members, while the members in other levels in the hierarchy are called
nonleaf members.
...

- Deepak

Deepak Puri
Microsoft MVP - SQL Server

*** Sent via Developersdex http://www.developersdex.com ***