Database Reference
In-Depth Information
•
Data warehouses often use non-normalized or partially
normalized schemas to optimize query performance.
•
Data warehouses should conform to Inmon's twelve rules for data
warehouses.
•
Three common data warehouse architectures are the basic
data warehouse, data warehouse with a staging area, and data
warehouse with a staging area and data marts.
The field of data warehousing is a fascinating breakthrough and is the subject of
many contemporary researches. Since it is relatively new, vast opportunities for data
warehouse architecture and ETL transaction optimization still abound. Data warehousing
is studied as an advanced course in several undergraduate degree programs, as well as
graduate programs. The supporting technologies are provided via products from the three
leading software engineering companies — IBM's DB2, Oracle (from Oracle Corporation),
and Microsoft's SQL Server. Additionally, these companies provide readily available
documentation on the topic.
With the advancement of the WWW and Web-accessible databases, it is anticipated
that the fascination in data warehousing will continue into the foreseeable future. And
speaking of Web-accessible databases, the next chapter discusses this topic.
24.7 Review Questions
1.
Give the definition of a data warehouse. Clearly outline what
data warehousing entails.
2.
Provide a rationale for data warehousing.
3.
Describe a scenario that would warrant the use of a data
warehouse.
4.
Discuss the main characteristics of a data warehouse in terms of:
•
Definitive features
•
Nature of data stored
•
Processing requirements
•
Rules that govern the data warehouse
5.
State the three architectural approaches to data warehouse
design. For each approach, describe the basic architecture,
and provide a scenario that would warrant such an approach.
6.
Clearly explain the ETL process. Give examples of ETL tools.
