
Dimensional Modeling for Data Warehouses
Practical Techniques for Star Schema Design, ETL, and Managing the Data Warehouse LifecycleBy Sofia PetrovaLength15h 17m
About this audiobook
A warehouse fails one grain statement at a time. This book makes the method explicit.
Most dimensional models do not collapse in a dramatic moment. They erode through vague grains, department-shaped requirements, and dimensions that cannot answer the next question. Dimensional Modeling for Data Warehouses gives you a repeatable method for building star schemas that survive real analytical pressure. You will learn to identify business processes rather than departments, build a bus matrix that sequences delivery, and declare a grain in one sentence that holds up when the business asks something new.
The book moves from architecture and requirements through the four-step dimensional design method, then into fact table types and additivity. You will choose between transaction, periodic snapshot, accumulating snapshot, and factless fact tables, handle semi-additive balances correctly, and design dimensions with surrogate keys, hierarchies, and the slowly changing approach the business actually needs. Advanced chapters cover role-playing, junk, outrigger, and mini-dimensions, bridge tables for multi-valued relationships, and complete models for retail sales, inventory, order management, and finance. Later chapters cover change data capture, quality screens, surrogate key pipelines, late-arriving data, partitioning, aggregates, semantic layers, governance, and testing on modern cloud platforms.
What you will learn:
• Apply the four-step dimensional design method to any business process
• Build a bus matrix that sequences delivery across subject areas
• Declare a grain in one sentence and defend it against real questions
• Choose between transaction, periodic snapshot, accumulating snapshot, and factless fact tables
• Handle semi-additive balances and additivity correctly
• Design dimensions with surrogate keys, hierarchies, and slowly changing dimensions
• Use advanced patterns: role-playing, junk, outrigger, mini-dimensions, and bridge tables
• Model retail sales, inventory, order management, and finance end to end
• Implement ETL subsystems, CDC, quality screens, surrogate key pipelines, and late-arriving data handling
• Plan physical design, partitioning, aggregates, semantic layers, governance, and testing
Written for data engineers, analytics engineers, and warehouse architects, this book is for practitioners who need models that stay coherent as platforms change and questions multiply. If you are responsible for the warehouse lifecycle, from requirements to evolution, this is your practical guide.
Audiobook details
GenreTechnology, Business and Economics
Length15 hrs 17 mins
Narrated byListen with 1,000+ voices
FormateBook with Audio
Publish dateSep 27, 2026
LanguageEnglish
Table of contents
1Dimensional Modeling for Data Warehouses
2Foreword
3Preface
4About This Book
5Chapter 1: Warehouse and Analytics Architecture
Show all chaptersShow less
61.1 The Reporting Problem That Starts Every Warehouse
71.2 Dimensional Bus Architecture Versus Normalised Enterprise Models
81.3 Data Lakes, Lakehouses and Where Dimensional Models Live
91.4 The Roles of Staging, Presentation and Semantic Layers
101.5 Where Dimensional Models Earn Their Place
11Chapter 2: Requirements, Business Processes and the Bus Matrix
122.1 Interviewing Business Users for Dimensional Requirements
132.2 Profiling Source Data Before You Model It
142.3 Identifying Business Processes Rather Than Departments
152.4 Building the Bus Matrix
162.5 Sequencing Releases and Setting Scope
17Chapter 3: The Four-Step Dimensional Design Method
183.1 Step One: Select the Business Process
193.2 Step Two: Declare the Grain in One Sentence
203.3 Step Three: Identify the Dimensions
213.4 Step Four: Identify the Facts
223.5 Testing the Model Against Real Questions
23Chapter 4: Fact Table Types and Additivity
244.1 Transaction Fact Tables
254.2 Periodic Snapshot Fact Tables
264.3 Accumulating Snapshot Fact Tables
274.4 Factless Fact Tables
284.5 Additive, Semi-Additive and Non-Additive Measures
294.6 Degenerate Dimensions, Allocations and Multi-Currency Facts
30Chapter 5: Dimension Design and Surrogate Keys
315.1 Surrogate, Natural and Durable Keys
325.2 Attribute Completeness and Hierarchies
335.3 Date and Time Dimensions
345.4 Flags, Grouping Attributes and Null Handling
355.5 Denormalisation Versus Snowflaking
36Chapter 6: Slowly Changing Dimensions
376.1 The Business Question Behind Every SCD Choice
386.2 Type 1: Overwrite
396.3 Type 2: Row Versioning
406.4 Type 3 and Hybrid Techniques
416.5 Choosing a Type from How the Business Asks Historical Questions
42Chapter 7: Advanced Dimension Patterns
437.1 Role-Playing Dimensions
447.2 Junk Dimensions
457.3 Outrigger Dimensions
467.4 Mini-Dimensions for Rapidly Changing Attributes
477.5 Bridge Tables for Multi-Valued and Ragged Relationships
48Chapter 8: Modelling Common Business Processes
498.1 Retail Sales and Promotions
508.2 Inventory Positions and Movement