Understand entity-attribute-value query costs, lost types, indexing limits, and alternatives using real columns, sparse columns, or JSON.
JSON or Real Columns: Deciding Where Attributes Live
Putting everything in a JSON column is tempting. Here is what it costs in reads, type checks and null handling, and when to promote a property.
Vertical Partitioning: Moving Rarely Used Columns Off a Hot Table
Moving rarely used columns to a side table can cut reads, or change nothing. Here is how to measure it and what it costs your writes.
Polymorphic Associations: Why One Column Cannot Point at Many Tables
One column that points at many tables cannot be enforced by SQL Server. Here is the demo, the orphans it allows, and a design that says no to bad rows.
Evolution of Big Data: From Flat Files to the Lakehouse
Evolution of Big Data follows one pattern: data grew, the old tool strained, and someone built the next one. This post walks from flat files and the relational model through warehouses and the cloud to the lakehouse.
Many-to-Many Bridge Table in SQL Server: Model and Query It
A many-to-many bridge table links two tables that each have many partners in the other, such as orders and plants. This post builds one with a composite key and two foreign keys, then joins it three ways. It also shows what the keys block and why an inner join hides plants nobody ordered.
Turning a Comma-Separated Column Into a Proper Child Table
A list stuffed in one column makes every search a guess. Split it, clean it and move it into a child table, then check the result before you drop anything.