Database Reference
In-Depth Information
• A project sponsor who is deeply involved and invested in the project's success,
using all of his considerable organizational clout as needed.
• A project manager who is engaged and knows common project management
tools and techniques.
• An It department that is amenable to bring Essbase into its quiver of tools, and
that is ready to support the project.
• one or more SmEs who will actively engage to assist with business requirements.
• And, finally, enthusiastic users who are more than mere Essbase consumers, but
instead are active participants in the development and refinement of the system.
Does this list resemble your company's Essbase project environment? Where it
does not, you need to review why your company is not able to assemble a project team
that looks like this and fix it. Essbase projects are difficult enough without adding the
complication of missing key players.
11.3.4 Source Data (or, At Least I Think That Is Where This Report Came From)
you have received your mandate. you have obtained your project sponsor and your cast
of characters are all lined up. But, stop for a moment. Before you get too excited, how is
your data? unless your project is not requiring any new sources or if data will be com-
pletely supplied by user input, bad source data can bring your project to a screeching
halt and often after you have spent an exorbitant amount of money.
The issue of bad source data is magnified with Essbase because, unlike SQL
(structured query language) data repositories that can bury bad hierarchies or incor-
rect data assumptions in the data not selected, the inherent dimensionality of Essbase
assumes that every hierarchy and data element must be correct because it is exposed.
Who wants a system that is reporting erroneous data? Implementation of any business
intelligence tool will expose and highlight your bad data. many companies “mask” bad
data by reporting with Excel
®
and other end-user applications. These tools often allow
users to fix data in their reports; consequently, they are not forced to be disciplined
about data quality. It is up to you as the Essbase administrator to make sure that your It
department provides good data; if they do not, it is still your responsibility, but the job
just got harder. Chapter 2 of this topic, Slay Bad Data in Essbase, gives both an overview
and a concrete method to fix bad data in Essbase. This section will only review at a high
level the various data source models and offer strategies to deal with the issues that often
are seen in these data sources.
11.3.4.1 Data Warehouse
An enterprise data warehouse is, because of its validated data
and metadata, relational schema designed for retrieval, and data definitions, the most
desirable source for Essbase data. While there is no guarantee that the data you need for
your Essbase implementation is there or that it is stored at the right level, because of the
data investment the business has already made in the warehouse, there is a very good
chance what you need is there. make the corporate warehouse your first and preferred
stop for data and, remember, warehouses are relationally based and that means you can
use SQL to extract, transform, and load into Essbase.
11.3.4.2 Data Mart
The problem with data warehouses is that they were in all prob-
ability never designed for Essbase. Sometimes this does not matter and the fact data
and metadata that you need can be quickly accessed through traditional Essbase
