Mastering OLAP Reporting: Meet Business Needs with Matrix Dynamics, Part 1

About the Series …

This
article is a member of the series MSSQL Server Reporting Services.
The series is designed to introduce MSSQL Server Reporting Services ("Reporting
Services"), with the objective of presenting an overview of its features,
together with tips and techniques for real-world use. For more information on
the series, please see my initial Database Journal article, A New
Paradigm for Enterprise Reporting.

As I
have stated since the charter article of the series, published about the time Reporting
Services was first publicly released, my conviction is that Reporting
Services will commoditize business intelligence, particularly in its role
as a presentation component within an integrated Microsoft BI solution.
Having been impressed from my first exposure to this exciting application, when
it was in early beta, my certainty in its destiny grows stronger by the day, as
I convert formerly dominant enterprise business intelligence systems, such as Cognos,
Business Objects, MicroStrategy, Crystal, and others, to the Reporting Services
architecture. I receive constant requests to conduct strategy sessions about
these conversions with large organizations in a diverse range of industries –
the interest grows daily as awareness of the solution becomes pervasive. Indeed,
the five- to six-plus figures that many can shave from their annual IT budgets
represent a compelling sweetener to examining this incredible toolset.

Note: To follow along with the steps we
undertake within the articles of this series, the following components, samples
and tools are recommended, and should be installed / accessible, according to
the respective documentation that accompanies MSSQL Server 2005:

Server Requirements

• Microsoft SQL
  Server 2005 Reporting Services

• Microsoft SQL
  Server 2005 Database Services

• The
  AdventureWorks sample databases

• Microsoft SQL
  Server 2005 Analysis Services

• The AdventureWorks
  OLAP cube

Client Requirements

• Microsoft
  Internet Explorer 6.0 with scripting enabled

• Business
  Intelligence Development Studio (optional)

Sample Files

We will be using one of the AdventureWorks sample
reports in the practice section, to save time and focus for the subject matter
of the article. The AdventureWorks sample reports are a set of
prefabricated report definition files that use the AdventureWorks
databases (both relational and Analysis Services) as data sources. The
sample reports are highly useful to many new report authors and other
practitioners, for whom they serve as a tool to assist in learning the
capabilities of Reporting Services, as well as templates for designing
new reports. For this reason, we typically make a copy of any report(s) we
modify within our lessons.

The samples are not
automatically installed. Before we
can install the Reporting Services samples, we must have already copied
the sample installation program to the PC with which we are working, in
accordance with the instructions found in the SQL Server 2005 Books
Online and elsewhere. We then run the sample installation program
to extract and copy the reports (and other) samples to the computer. The sample
installation program also installs the AdventureWorks databases (both Database
Engine and Analysis Services varieties)..

The samples come packaged within a Report Server
project file, which we will open and use in many lessons, rather than creating
a new project file. Please make sure that the samples and the project file are
installed before beginning the practice section of this article, so as to
provide an environment in which to complete the exercises effectively.

About the Mastering OLAP Reporting
Articles …

One of
the first things that become clear to "early adopters" of Reporting
Services is that the "knowledgebase" for OLAP reporting with this
tool is, to say the least, sparse. As I stated in my article, Mastering OLAP
Reporting: Cascading Prompts, the
purpose of the Mastering OLAP Reporting subset of my Reporting
Services series is
to focus on techniques for using Reporting Services for OLAP
reporting. In many
cases, which I try to outline in my articles at appropriate junctures, the
functionality of the reporting solutions of well-established, but expensive,
solutions, such as Cognos PowerPlay, can be met in most respects by Reporting
Services – at a tiny fraction of the cost.

The
vacuum of documentation in this arena, to date, represents a serious "undersell"
of Reporting Services, from an OLAP reporting perspective. I
hope to contribute to making this arena more accessible to everyone, and to
share my implementation and conversion experiences as the series evolves. In
the meantime, rest assured that the OLAP potential in Reporting
Services will be yet another reason that the application commoditizes
business intelligence.

For
more information about the Mastering OLAP Reporting articles, see
the section entitled "About the Mastering OLAP Reporting
Articles" in my
article Ad Hoc
TopCount and BottomCount Parameters.

Overview

As I have shown in many
past articles, the Microsoft Integrated BI Solution, consisting of the MSSQL
Server 2005 Database Engine, Analysis Services 2005 and Reporting
Services 2005, provides unprecedented flexibility in helping
implementers and developers to meet client and employer needs. As I convert
existing enterprise business intelligence to the Microsoft BI solution, I come
across opportunities to meet diverse needs on a virtually daily basis. Reporting
Services is particularly strong in the flexibility department: Presentation
nuances are legion, and one discovers, with constant use of Reporting
Services to meet a wide range of reporting needs, how truly flexible the
application can be.

Foremost in the
challenges that I find in the field are those that arise based upon the need to
replicate features that exist in previously created reports, often in tandem
with a requirement to enhance the reports to provide new capabilities and
presentation options. These replication efforts often come as a part of
converting the reports of a predecessor enterprise reporting system (for example,
converting Cognos PowerPlay or Cognos Impromptu reports to Reporting
Services). They also come about through a need to duplicate early efforts
within Reporting Services itself into enhanced reports that offer more
flexibility, additional features, and so forth, to information consumers, as we
do within the Practice session of this article.

As most who have worked
with Reporting Services know, authoring reports in the application
consists largely of associating controls within the report body with fields
that are created within one or more DataSets within the report file
(.rdl). Sometimes we can associate controls with other controls to
achieve results that might otherwise be difficult. One of my favorite controls
in Reporting Services, for OLAP reporting in general, and for its
synergistic qualities when working with other objects within Reporting
Services, is the matrix data region.

One certainly finds that
the table data region seems the clear favorite, among
the few Reporting Services books on the market as of this
writing. This is perhaps because the table
data region closely
resembles the functionality found in the those (largely) relational reporting
applications that were dominant before Reporting
Services appeared. Indeed,
the vast majority of current writing continues to surround relational reporting
(see my comments in the About
the Mastering OLAP Reporting Articles … section above), mostly because the
majority of the current Reporting
Services
writers have previously written books on the relational enterprise reporting
tools to which I refer, and have adapted much of their teaching approach with
those applications to Reporting
Services.
Even among the sample reports that ship with MSSQL Server 2005 Reporting Services, one finds the
vast majority to surround relational reporting of this sort.

Among many attributes
that I like within the matrix data region, is the fact that its rows and columns
can be dynamic, unlike those of a table
data region, which are
static. We can define matrices with either or both of static
and dynamic rows and columns, and thereby support meeting business
requirements in many environments where static columns and rows might limit
consumers. When we couple the capability for dynamic columns with the behavior
we can obtain when nesting other data regions within the matrix, we begin to see how we can produce a far more robust
presentation for our employers and clients.

In
this article, we will examine a scenario where the dynamic nature of the matrix
helps us to meet the expressed needs of a hypothetical group of information
consumers, both directly and (to some extent, a bit less intuitively) from a
more indirect perspective. In
this two-part session, we will:

• Prepare for the steps of our
  practice session by accessing the Sample Report Server Project that is
  available with an MSSQL Server 2005 installation (Part 1);

• Ascertain connectivity of the Shared
  Data Source, and open an existing sample OLAP report in Reporting
  Services that is based upon a table data region (Part 1);

• "Convert" an existing
  sample OLAP report (based upon a
  sample SQL Server Analysis Services cube), using a matrix data region, with which to perform our
  practice exercise (Part 1);

• Add parameterization (with
  multivalue input capabilities) for territorial regions to the new report,
  using a multivalue parameter (Part
  2);

• Make structural changes to the report,
  to meet the business requirements of a hypothetical group of information
  consumers for presenting independent matrices based upon a geographical parameter
  they select at runtime (Part 2);

• Incrementally preview
  the report to ascertain the effectiveness of our solution (Parts 1 and 2);

• Discuss, at appropriate
  junctures, the results obtained within the development techniques that we
  exploit throughout our practice session (Parts 1 and 2).