 |
Data Warehouse
Architecture
Need to define data warehouse architecture
for an information management project and want an overview of key
architectural components?
Objective
The objective of the architecture and design phase is to create a
technical
blueprint in
sufficient detail for the development team to commence building the
information management solution.
This phase will define the technical architecture including:
- Data
warehouse and business intelligence architecture;
- Database architecture, system architecture and
infrastructure;
- Database sizing and performance;
- Mechanisms to extract, transform and load data;
and
- Data backup/recovery processes.
Prerequisites
This phase should not start without an:
- Approved business case; and
- Approved requirements specification
Tasks
This phase requires a team of highly specialized data architects, data
movement designers and business intelligence designers who will:
- Review hardware architecture including capacity
planning;
- Analyze data storage structures including,
staging area design,
operational data stores (ODS), data warehouse design, and data mart
design;
- Recommend data security processes including all
backup and recovery procedures;
- Consider what is required to satisfy data
quality requirements
including any specialized processes to cleanse data and report data
quality issues; and
- Design data usage, reporting and analytical
mechanisms including any business intelligence tool design
specifications
How long
should this take?
This is dependent upon the complexity of the requirements, the number
of source systems and interfaces, and how much the architecture and
design team has been involved in requirements specifications.
Information management best practices recommend that the data architect
and data analyst play a key role in specifying requirements. This will
reduce the time required to complete design tasks.
Data warehouse architecture phase is a highly focused effort that
should be completed by an experienced team in a four to six week period.
Deliverables
This phase delivers a technical architecture document, which
includes:
- Relational database design and data model
including;
- Data movement design including:
- Interface design;
- Data integration specifications;
- Physical data mapping, source to target map,
including business rules to be applied to extracted data;
- Design specifications;
- Mechanisms for data transfer and load; and
- Data load strategy.
- Data security design, including:
- Backup/recovery specifications.
- Data quality design specifications;
- Query profiles, including
- Design specifications for pre-built reports
or queries
- Data conversion solutions
Standards
and
best practices
- Data model standards;
- Data
base management standards;
- Data movement best practices;
- Architecture document template;
- Templates for:
- Interface specification;
- Data integration specification;
- Data quality specification;
- Load manager and warehouse manager
specification;
- Operations/infrastructure specifications;
- Business intelligence reporting
specification; and
- Metadata design specification.
Tools
Data warehouse design and development requires database
development software.
Summary...
The objective of the data warehouse architecture phase is to
create
a
technical blueprint in sufficient detail for the development team to
commence building the information management solution.
A highly specialized team of experienced data architects, data movement
designers and business intelligence designers should complete
this phase in a four to six week period.
|
|
|
|
