There are a large number of independent
vendors such as Ascential, Informatica, and so on, as well as several database
vendors such as Oracle, IBM, and Microsoft in this market.
Enterprise data can be of very different kinds ranging from flat files to data stored
in relational databases, with the more recent trend of data being stored in XML data
stores. The extraordinary number of database related products, and their historic
evolution, makes this task exacting.
SSIS Basics
[ 8 ]
It is not very frequent that extracted data is usable as it is, and may need some kind
of transformation. Often, the extracted data has inconsistencies. There are many
reasons for inconsistencies, as computer applications may be affected by changes in
technology (version changes, new methodologies, etc.), poor or no validation, which
are the norms in legacy data software changes (for example, the date and time will
be new data types in SQL 2008 to address several issues of awareness of time zones,
higher precision needed in financial applications, address compatibility with third
party vendors, etc.), changes in the way applications interact with backend data,
localization, etc.
Some inconsistencies can be understood quite easily, such as the same information
stored in different data formats, while others could be more difficult.
The efficiency and agility of an enterprise would depend on the quality of data in its
databases. Hence it is not only necessary to cleanse or scrub data, but also modify it
to be internally consistent, and have the correct data quality.
Pages:
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27