This section covers RDBMS fundamentals — relational theory, schema patterns, and MySQL window functions — alongside the Kimball dimensional-modeling approach to data warehousing, including star, snowflake, and galaxy schemas, conformed dimensions, and bus matrices.
The relational model, ACID guarantees, and normalization — and where RDBMS still anchors the modern data stack alongside warehouses and lakehouses.
Comparison of star, snowflake, galaxy, hierarchical, network, and flat schema patterns — when each fits and the trade-offs in joins versus redundancy.
Window functions like LAG() for row-over-row analytics, time-series differences, and trend calculations without self-joins.
Central fact table surrounded by denormalized dimensions — the warehouse default for fast analytical queries and BI tooling.
Normalized variant of the star schema; reduces storage redundancy at the cost of additional joins and query complexity.
Bottom-up data-warehouse approach built from data marts, conformed dimensions, and a shared bus matrix spanning business processes.
Two-dimensional matrix mapping business processes to conformed dimensions — the planning artifact that anchors an enterprise data warehouse.
Worked end-to-end dimensional model for an e-commerce sales fact — grain, dimensions, surrogate keys, and slowly changing attributes.
ERDs as the conceptual layer above the schema — entities, attributes, cardinalities, and how an ERD becomes a physical design.
Bridging conceptual ERDs and physical schemas — the platform-agnostic logical model that survives database migrations.
Schema-design fundamentals: naming, keys, relationships, and how design choices ripple into query performance.
Cloud warehouse architecture: separation of storage and compute, virtual warehouses, and how it differs from classic on-prem RDBMS.
Bronze/Silver/Gold layered design on a lakehouse — the modern dimensional-modeling parallel for streaming-first warehouses.
ACID-on-the-lake: how Delta brings RDBMS guarantees (transactions, schema enforcement, time travel) to a Parquet data lake.
About this section. Row 1 covers RDBMS basics — relational theory, schema taxonomy, and MySQL window-function patterns. Row 2 dives into the dominant warehouse schema patterns (star, snowflake) and the broader Kimball architecture they sit inside. Row 3 walks through applied dimensional modeling, from the bus-matrix planning step to a worked sales fact for an e-commerce warehouse.
Start with relational fundamentals, then the schema overview, then the star and snowflake pages. The Kimball pages build on those to show how many star schemas share conformed dimensions across an enterprise, and the sales model applies everything to one realistic case. The MySQL LAG Function page is a standalone SQL technique useful for period-over-period analysis on any of these models.