Improving Performance with Intelligent Query Processing
summary
Demonstrates specific toggles and metrics: enable lightweight profiling (trace flag 7412 on 2017 or DB scoped configs on 2019 compat level 150) and see real effects — table variable deferred compilation shifts an estimate from 1 row to 228,000 rows and changes plan shape, memory grant feedback shrinks a 731MB grant to ~8.5MB and prevents tempdb spills on subsequent runs, and scalar UDF inlining permits parallelized plans; uses new DMVs (query_plan_stats, in‑flight plans) and shows measurable, working improvements.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 1:20Contact Information
- 2:25Agenda
- 3:07What is Intelligent Query Processing
- 3:43Performance Tuning
- 4:12Lightweight Query Profiling
- 6:35Compatibility Level
- 8:06Lightweight Query Profiling Features
- 10:16Table Variables
- 11:07Temporary Tables
- 11:52Table Variable Compilation
- 13:37ScalarUFD Inlining
- 17:32Memory Grant Feedback
- 20:53Batch Mode on Rowstore
- 22:54Demo
- 31:39Inline Scalar Functions
- 34:35Compatibility Levels
- 36:29Memory Grants
- 39:34Batch Mode
related talks
53mBanishing Slow Queries with Intelligent Query Processing
53mBanishing Slow Queries with Intelligent Query Processing
22mEmpowering SQL: Mastering Intelligent Query Processing in 20 Minutes
53mBanishing Slow Queries with Intelligent Query Processing
01h 00m7 Tips to Performance Tuning, Optimization, and Everything
22mEmpowering SQL: Mastering Intelligent Query Processing in 20 Minutes
22mEmpowering SQL: Mastering Intelligent Query Processing in 20 Minutes
50mBuilt-in Intelligence with Azure SQL and SQL Server 2022
