Thursday, 21 December 2017

DWH Life cycle

DWH is a process of building a data warehouse
1) Requirement gathering
  • It is done by business analysts, Onsite technical lead and client
  • In  this  phase  a  Business  Analyst  prepares  business  requirement specification(BRS)Document
  • 80% of requirement collection takes place at clients place and it takes 3-4 months for collecting the requirements
2) Analysis:
  • After collecting the requirements data modeler starts identifying dimensions, facts & aggregation depending on the requirements
  • An ETL Lead & BA create ETL specification document which contains how each target table to be populated from source
3) System Requirement Specification (SRS)
  • After collection of onsite knowledge transfer offshore team will prepare the SRS
  • A SRS document includes software,hardware,operating system requirements
4) Data Modeling
  • It’s a process of designing the database by fulfilling the use requirements
  • A data modeler is responsible for creating DWH/Data marts with the following kinds of schema

  • Star schema
  • Snowflake schema
5) ETL Development
  • Designing ETL applications to fulfill the specifications documents which are prepared in analysis phase
6) ETL Code review:
Code review will be done by developer
The following activities take place
  • Check the naming standards
  • Check the business logic
  • Check the mapping of source to target
7) Peer Review:
Code will be reviewed by a team member
  • Validation of code but not data
8) ETL Testing:
Following tests will be carried out for each ETL Application
  • Unit testing
  • Business Functionality testing
  • Performance testing
  • User acceptance testing
9) Report development environment:
  • Design the reports to fulfill report requirement templates/Report data workbook(RDW)
10) Deployment:
  • Process of migrating the ETL Code & Reports to a pre production environment for stabilization
  • It is also known as pilot phase/stabilization phase


No comments:

Post a Comment