ETL Architectures – Concepts and Implementation

ETL Architectures is an in-depth, technical course that teaches the concepts for designing and implementing the appropriate architectures to use in managing the extraction, transformation and loading (ETL) of data for:

    High performance decision support environments (data warehouses, dimensional data marts, Operational Data Stores (ODS), etc)
    Master Data Management hubs (Customer Data Integration (CDI), Product Information Management (PIM), etc)
    General data integration (e.g. Service Oriented Architectures (SOA))

This course will review these architectures and concepts with the primary focus on the concepts and techniques that apply to various approaches to ETL. Participants will learn when to use certain techniques, based on their technical and business requirements. With hands-on workshops, attendees will study different ETL products and methodologies for implementation in today’s heterogeneous system environments.
Benefits To Your Company

By learning the best way to design ETL architectures, architects and ETL developers will be able to implement the appropriate tools and techniques to satisfy business requirements and relate them to the supporting data structures. They will:

    Understand the concepts of extraction, transformation and loading in decision support systems, master data management systems, SOA environments.
    Understand the various forms of data architectures and how to apply ETL techniques to these
    Understand sophisticated techniques for more complicated ETL solutions (real-time, high volume, etc)
    Construct ETL architectures that are flexible to support changing business and technical requirements
    Learn about the most common ETL products and their strengths and weaknesses.

Who Should Attend

    Data Warehouse Architects
    Enterprise Architects (Data, Technical)
    ETL Developers
    Data Architects
    Business Intelligence designers
    Database designers
    Database administrators (DBA)

What Makes This Certified Course Unique

This ICCP-certified course provides participants with practical, in-depth understanding of how to create appropriate ETL architectures for decision support and data integration solutions. Hands-on workshops throughout the course will reinforce the learning experience and provide the attendees with concrete results that can be utilized in their organizations.
Course Outline

    Review common system architectures
        Transaction Processing
        Decision Support
        Master Data Management
        Service Oriented Architecture
    ETL Concepts
        General principles
        Design and plan for reuse
        Design for error handling
        Design for performance
        Design for maintainability
        ETL Standards
        ETL and Meta Data
        ETL Tool Usage
    ETL for Decision Support
        ETL for the Data Warehouse
            Data Sourcing / Changed Data Capture
            Data Transport
            Data Staging
            Changed Data Determination
            Loading normalized warehouse structures
        ETL for the Data Mart
            Surrogate key lookup and assignment
            Slowly Changing Dimensions - Types 1,2, 3 & 6
            Denormalization and impact on ETL
            Populating “junk” dimensions using a Cartesian product
            Aggregation
        ETL for the ODS
            Real/near time approaches
            Data Modeling differences
        Row level security
        Closing the loop
    ETL for Master Data Management (MDM) and Service Oriented Architectures (SOA)
        Customer Data Integration (CDI)
        Product Information Management (PIM)
        Integrating ETL and SOA environments
        Integrating ETL with Data Quality tools
        Integration with OLTP systems
    ETL Tools
        Leading ETL tool vendors
        ETL tool strengths / weaknesses
        Choosing the correct ETL tool
    High performance ETL
        Indexing (b-tree, bitmap, join indexes, etc)
        Forms of Parallelism
        RDBMS tuning and ETL
        Massively Parallel Processing (MPP) platforms vs. Symmetrical Multiprocessing (SMP) platforms
        ETL query optimization
    Workshop conclusion
        Summary, additional exercises, sources for further reading, etc.

SOURCE:http://www.ewsolutions.com/education/data-warehouse-training/document.2008-04-23.8274387713