NTILE deals 100 customers into eight batches in one query. The catch is keeping those batches stable while new rows arrive.
How to Read a Stored Procedure You Did Not Write
Learn to read stored procedure code by tracing inputs, writes, transactions, errors, and dependencies before trying to understand every line.
Interview Questions That Actually Reveal SQL Skill
Use SQL Server interview questions that reveal reasoning about joins, NULLs, performance, and recovery instead of rewarding memorised trivia.
CODEPAGE 65001: Importing UTF-8 Text Files Without Garbled Accents
Your import turned Élodie into odd symbols. Here is how CODEPAGE 65001 fixes the reading, and why the column type still matters.
Relational, Document, Graph and Vector: Choosing the Right Data Model
Choosing the Right Data Model means matching the shape of your data to the questions you ask. This post stores one bookshop four ways in SQL Server 2025, with relational tables, JSON documents, graph tables and a vector column, and runs the question each one answers best. It ends with a rule for choosing.
Searching Every Error Log for Slow I/O Warnings
The warning you need may be in an archived log. Here is a short loop that searches every retained error log and tells you which ones it could not read.
What to Check When You Are Woken at Three in the Morning
Use a DBA on call checklist to establish impact, inspect blocking and disk space, review failed jobs, and choose a safe next step during incidents.