A data warehouse is never static for long. There are always new requirements to meet, new projects to deliver, new reports to develop, new business problems to solve.

Add new developers joining the team, contractors coming and going, various vendors responsible for different projects and the result can be an increasingly fractured and diluted ETL architecture.

A disciplined, cohesive ETL architecture and framework is an essential foundation for the ongoing health of a data warehouse investment.

If development streams are allowed to diverge from this, the result can be a data warehouse that is difficult and costly to maintain and troubleshoot, increasingly hard to enhance and develop while being almost impossible to audit.

The solution

Analytics8 has many years of experience in reviewing and assessing data warehouse and ETL architectures and can provide an assessment and recommendations covering the following areas of ETL design and implementation:

• ETL tool selection and configuration, scalability and upgrade paths
• The ETL framework and standards
• Consistent application of standards and conventions
• The effective and efficient use of staging areas
• Development and testing repositories and promotion methodologies
• Exception handling, scheduling and error trapping
• Component design and reusability
• Logical and physical data model design, effectiveness and best practice
• Management of hierarchical and reference data (master data management)
• Metadata management and data dictionaries

Each is evaluated against known industry best practice, and our own real-world experience to provide balanced, realistic, business-driven recommendations that can be implemented over a realistic timeframe.