Database Reference
In-Depth Information
2.6
Appendix: The Code
56
2.6.1
oDI objects
56
2.6.2 Essbase
64
2.6.3
SQL objects
69
2.1 introDuCtion
It seems self-evident that an Essbase data must be complete and correct every time it is
loaded, for what is the purpose of an inaccurate analytical tool—bad analysis? Without
good data, Essbase is just a collection of metadata and reports, of little worth, except
possibly to the consulting company that was paid to develop it. With good data, Essbase
is a powerful tool that empowers users to analyze reliable, accurate, and validated data
to make informed decisions.
And yet, most Essbase databases are at least temporarily incorrect after data is loaded.
Why this bad data is tolerated is a mystery; perhaps it is insufficient budget, or the belief
that figuring out how to eliminate bad data is difficult, or maybe just that no one has
tried. regardless of cause, the mission of this chapter is to give you the tools to Slay Bad
Data with minimal manual intervention.
2.1.1 Assumptions about You, Dear Reader
Who are you? What, you don't know your own name? This sounds like a case for a psy-
chiatrist and a strong dose of … whoops, wrong book, you need to go to the self-help
section and put this one back where it belongs.
Are you still here? If so, you are likely an Essbase developer who has experienced the
pain of bad data, knows all too well the consequences, and does not ever want to repeat
the experience. If you have somehow blissfully skated through your professional Essbase
life not experiencing the pain that is bad data, you have been lucky. Since you are read-
ing this chapter, you are to be commended for realizing that you have been dodging fate.
regardless of experience, you are searching for a way to slay the bad data that is storm-
ing the beautiful design of your Essbase outline.
2.1.1.1 Scope one of the many great aspects about Essbase is that it is not tied to any
single data source; universality and adaptability have been key ingredients to success
with Essbase. given the wide variety of data sources, be it a general ledger, sales fore-
casting system, or a series of spreadsheets, it is beyond the scope of this chapter to cover
data extraction . you or your Information technology (It) Department must master
that, not inconsiderable, challenge. In the end, however, the user never sees that part
of the process, it is you and your cube who will be blamed if there are data errors.
The concepts and techniques in this chapter will help guarantee that the load of this
extracted data to Essbase is complete and internally consistent, with bad data if not
actually slain then highlighted, and the effects measureable and mitigated and not sim-
ply hidden in an error file somewhere.
While every effort to fully explain this chapter's theories and processes will be made,
space considerations require that you have some familiarity with data-loading tech-
niques. table  2.1 lists the concepts you should be familiar with before you read this
chapter. The table is grouped by technology type.
Each technology type will have advanced content beyond these prerequisites, which
will be fully explained in this chapter.
Search WWH ::




Custom Search