Storing strings in SQL Server and Azure SQL DB
summary
Explains when to use UTF-8 versus UTF-16 in SQL Server and Azure SQL DB with concrete numbers and migration tactics: UTF-8 gives 50% storage savings for ASCII/Latin-heavy data and shows about 35% read/write I/O improvement in tests, while UTF-8 worsens performance ~25% for CJK data where UTF-16 is better; SQL 2019/Azure SQL add UTF-8 collations, conversion options are ALTER TABLE (blocking) or copy-and-replace (faster), and UTF-8 not supported with xml in-memory ltp or Always Encrypted enclaves.
AI-written summary · about
chapters (16)
- 0:00Introduction
- 1:12Challenges
- 2:10Contoso
- 2:41New challenges
- 4:17Unicode
- 6:21Encoding
- 7:54Encoding in SQL
- 10:45Data types
- 12:09Storage differences
- 13:05Demo
- 13:22Operational considerations
- 15:25Other performance considerations
- 15:53Other considerations
- 17:16String literals
- 18:35Conversion Methods
- 19:10Alter Table
related talks
45mThe wonderful world of collations, encodings and text storage
01h 02mThere's No Such Thing as Plain Text
48mMS SQL New functions, syntaxes, tips & tricks
01h 06mSQL ❤️ JSON: The Best of Both Worlds
27mWorking with Multilingual Data in SQL Databases through Erlang
49mSQL Server on Azure VM - Configuring for Price-Performance
51mThe 10 things every developer must absolutely know about Azure SQL
45mJSON - a first class citizen in Azure SQL DB