Read bitmap filters in parallel hash join plans, compare fact scan output, and distinguish early row rejection from guaranteed savings.
What Is NewSQL? Distributed SQL Databases Explained
What is NewSQL? It’s the name for relational databases that spread data over many servers and still give you SQL and ACID transactions. This post walks one orders table through key ranges, replicas and a majority vote.
DevReach 2013: Presenting in Sofia, Bulgaria
I presented my Database Worst Practices session at DevReach 2013 in Sofia, Bulgaria, on October 1, 2013. It was my first time presenting in Europe. Here is my recap of the conference and my day in Sofia, with photographs.
OPTION (FORCE ORDER): When a Join Order Hint Helps and Hurts
Test OPTION (FORCE ORDER) against real join plans, see when fixed table order helps or hurts, and apply a reversible Query Store hint.
Relational Database in Big Data: Still at the Center
The relational database in big data projects is still the system of record, and most analytics platforms copy their core business data from it. This post shows what a relational database does well, where it sits in a modern platform, and how the SQL Server 2025 json type adds a flexible schema next to ordinary tables.
Columnar File Formats: Why Parquet Changed Big Data
Columnar file formats store a table one column at a time, so a report reads and compresses only the columns it needs. This post compares rows and columns on a tiny table, explains how a Parquet file is laid out, and measures the size difference in SQL Server 2025.
MapReduce Explained: Map, Shuffle and Reduce Step by Step
MapReduce splits a big job into many small map tasks, moves the results by key in a shuffle, and adds them up in reduce tasks. This post walks one word count through every step, with a recreated diagram and the same idea in T-SQL.