Query Store: A Retrospective
summary
Query Store originates around 2008, faces three cancellations, and ships in SQL Server 2016; it records query text, semantic context, and stores show plans plus runtime stats in hourly buckets by default. It supports plan forcing and introduces auto-forces regressions (2017), memory grant feedback and DOP feedback (2022), and readable secondary capture. It runs on by default in Azure and SQL 2022 and helps detect and revert plan regressions and de-risk major version upgrades.
AI-written summary · about
chapters (12)
- 0:00Intro
- 2:23(Mostly) Fun Facts
- 6:38What was it like before Query Store?
- 12:07Table Layout
- 17:40Architecture - Observer Pattern
- 20:15Query Store UX - SSMS
- 22:55Why does the Query Store "work"?
- 25:28Binary Persistence
- 29:58Major Version Upgrades (SQL Server)
- 33:21Query Store Features Over Time
- 38:55Biggest Blindspot - The Cloud
- 43:08Azure SQL Database
related talks
22mA Beginners guide to Query Store- 22mA Beginners guide to Query Store
50mQuery Store Best Practices- 50mQuery Store Best Practices
26mUsing Azure Query Store to Understand PostgreSQL Performance
19mStabilizing Performance with Query Store
51mMaking the SQL Query Processor Work for you- 19mStabilizing Performance with Query Store