Query tuning optimization
•Download as PPTX, PDF•
0 likes•322 views
This document discusses query tuning and optimization techniques: 1. It outlines common approaches consultants take such as proactively leveraging metadata to identify missing, duplicate, or unused indexes and fetch top resource-intensive queries, and reactively identifying bottlenecks and top queries from metadata. 2. It explains how to fetch actual execution plans and compare them to estimated plans to understand performance issues. 3. It reviews common execution plan patterns like index seeks, scans, lookups, and warnings, and how to delve deeper into topics like cardinality estimation, residual predicates, data distribution impacts, and sorts. 4. It promises a demonstration of these techniques.
Recommended
More Related Content
Similar to Query tuning optimization
Similar to Query tuning optimization (20)
More from Harsh Chawla
Recently uploaded
Recently uploaded (20)
Query tuning optimization
- 1. © 2012 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION Presented by Harsh Chawla SQL Community Session – Delhi NCR DBA: Query Tuning and Optimization - Expert
- 2. 1. DB Consultant Approaches 2. Fetching Execution Plans 3. Common execution plan patterns 4. Understand the data distribution 5. Q&A Welcome. Microsoft Services helps businesses around the world maximize their return on investment in Microsoft products and technologies. 12/28/2015 2
- 3. Proactive Approach 12/28/2015 3 • Leverage MDW • Have a Performance Baseline Understand the load • Missing Indexes • Duplicate Indexes • Unused Indexes Understand the Indexing • Leverage Plan hash and Query Hash to find top resource intensive queries • Leverage MDW to identify the top queries Fetch Top Queries for Tuning
- 4. Reactive Approach 12/28/2015 4 • Leverage MDW /SQLNexus to identify the resource contention Identify the Top Bottlenecks • Leverage Plan hash and Query Hash to find top resource intensive queries • Leverage MDW to identify the top queries Identify the Top Resource intensive queries • Set Statistics IO ON • Set Statistics Time on • Set Statistics Profile on Explore the execution plans
- 5. Fetching Execution plans 12/28/2015 5 • Estimated Vs Actual • DMVs show only estimated plans • Extract the variables from the execution plan • Execute the query to get the actual execution plan
- 6. Common Execution Plan Patterns 12/28/2015 6 • Index Seek • Index Scan • Key/RID Lookups • Implicit Conversions • Missing Statistics Warning • Sort Warning • Hash Warning
- 7. Let’s Delve Deeper 12/28/2015 7 • Cardinality Estimation • Residual Predicates • Impact of Data distribution • Sorts
- 8. Cardinality Estimation 12/28/2015 8 • Beauty lies in the data • There is never a coincidence • Wrong Estimation cause Hash warnings/sort warning or almost every mess up • It’s the biggest gotcha while query tuning • Try to investigate the numbers
- 10. Residual Predicates 12/28/2015 10 • What’s a Predicate • If the predicate is not SARGable, Residual Predicates will be seen
- 11. Demo 12/28/2015 11 Impact of data distribution
Editor's Notes
- Add all the scripts to identify top queries
- Show Demo on basic SQL Execution Estimated Vs Actual Show the Parametrized query and explain how they can get the parameters out of the execution plan Get the query hash and plan hash Only zero cost plans are not shown here Get the execution plan from the sys.dm_exec_query_plans and sys.dm_exec_query_stats
- Hash warning sort warning - everything happens due to wrong cardinality estimation
- Show Demo for Basic Query Execution.sql Improving cardinality estimation Also tell everything is dependant on statistics 4. What if we don’t have statistics , how many rows will be estimated 5. Always keep your statistics updated 6. Auto create stats and auto update stats enable
- Demo