Architecture of a Data warehouse

Architecture of a Data warehouse:

The architecture of a data warehouse is shown above.

Since normally a data warehouse is used for analyzing trends and preparing reports out of it, the common architecture (as shown above) will have a layer for reporting purpose.

1) Source Systems: Source Systems are systems that provide data to the data warehouse. Source systems , typically , can be OLTP systems or legacy systems that are used for operational purposes. There are 4 broad categories of source data:

a) Production data : Production data comes from operational systems. Any organization can have multiple operational systems, which may or may not be secluded from each other. The data format in each system may vary one from the other.

b) Internal Data: This data is internal to the organization or maybe a department in the organization.

c) Archived data: Most of the organization stores the data in archives which can be in large storage media. This data may be needed for reporting. This data can be fed into the warehouse ( temporarily ) , so that reporting for any trends can take place.

d) External data: Some data come from external sources. Eg: Weather reports , Base interest rates as announced by the Central Bank ( like Reserve Bank of India )or also can be the news reports etc.

(Next session will be centered around the next layers i.e. staging area and datawarehouse ).