Data sources are an extremely important piece of any
BI solution. A scorecard or dashboard can only be as good as the data it
is presenting, so it is important to consider the quality,
accessibility, and format of the data before embarking on a BI solution.
Data sources serve as the
foundation for KPIs and analytic charts and grids. In addition, data
sources can be used for dashboard filters. Specifically, tabular data
sources can be used for the Custom Table filter, and the Analysis
Services data source can be used for the Multidimensional Expressions
(MDX) query, member selection, named set, time intelligence, and time
intelligence connection formula filters.
The following two
subsections discuss the two broad categories of data sources that PPS
supports: multidimensional data sources and tabular data sources.
Multidimensional Data Sources
Multidimensional data
sources are all variants of SQL Server Analysis Services. They feature
dimensions, fact data, and support for the MDX language, as opposed to
columns and values that are used in tabular data sources.
Multidimensional data sources
are the primary use case for PPS, and many features such as analytic
charts and grids and the decomposition tree in PPS require
multidimensional data to operate. If you do not have any data in
multidimensional format, consider porting some of your data to a
multidimensional format to take full advantage of the features of PPS.
Analysis Services (2005, 2008, and 2008 R2)
SQL Server 2008 R2 PowerPivot
Note
Online
analytical processing (OLAP) data sources, which were available in PPS
2007, have been removed from PPS 2010. This is largely due to functional
gaps caused by variations in features and support for the MDX language
between other OLAP-supporting data sources and true Analysis Services
data sources. If you have a multidimensional data source in a format
other than SQL Server Analysis Services, it will not be accessible
directly by PPS 2010.
Tabular Data Sources
Tabular data sources come in a
wide variety of formats. Tabular data sources all feature columns and
values and conceptually are similar to a spreadsheet.
Tabular data sources
have limited functionality. You can represent them as KPIs on scorecards
or have them appear as data values within filters to interact with
various non-analytic report types. Generally, this is the extent of
their functionality. Listed here are the tabular data sources supported
by PPS:
Excel Services (Excel Services 2007 or 2010)
Import from Excel workbook (Excel 2007 or 2010)
SharePoint list (SharePoint 2007 or 2010)
SQL Server table (SQL Server 2005, 2008 or 2008 R2)
Custom data source
Conceptually, tabular data is turned into microcubes
within Dashboard Designer. Each tabular data source can define
dimension and fact data types in the data source definition editor.
Dimension values
are populated through members that are currently available in the data
column. For instance, take the example of a dimension column that can
contain Yes or No values. If the data only contains No values, it will
not be possible to select a Yes value when adding a dimension filter to a
KPI until the data contains at least one Yes value. In addition,
dimensions created from tabular data are also always flat. Therefore, it
is not possible to create a parent/child relationship and hierarchies
between dimension values.
Fact data types are
determined by the contents of the list. If all the data values are
numbers, the data type is considered a number and can be aggregated as
numbers. If just one value is text, the entire list will be considered
as text fact data.
The Data Source
template allows you to select the appropriate data source for your KPIs.
For KPIs, you can use all types of PPS data sources. For analytic chart
and grid reports, a multidimensional data source is required.
Data Source Security and Trusted Locations
All
data connections must be stored in a trusted data connection library.
This is true for all data connections that you want to reference from
SharePoint. The reason for this is to ensure that all external data
connections are approved before they can be used. In other words, a user
cannot upload a data connection file to a random document library and
use it in PPS; the user needs proper permissions to add the data
connection file to the proper library.
By default, all data source
locations are trusted and, if desired, need to be locked down by the
farm administrator.
