A Function by Any Other Name is a Function
0 likes660 views
This document discusses user defined functions in SQL Server and their performance impacts. It identifies purposes for creating user defined functions, discusses the different types including scalar, inline, and multi-statement functions. It demonstrates the performance impact of each type of function through examples and benchmarks. Scalar functions have linear performance effects while inline and multi-statement functions allow for more complex logic but can introduce duplication of access. In summary, functions provide benefits like consolidation and reuse but their performance impact depends on the type and how they are used.
A Function by Any Other Name is a Function
- 1. Performance Impacts of User Defined Functions
- 2. Introduction Jason e: [email protected] Strate e: [email protected] b: www.jasonstrate.com t: StrateSQL Resources jasonstrate.com/go/Waits MAKING BUSINESS INTELLIGENT www.pragmaticworks.com
- 3. Functions http://www.flickr.com/photos/protocol/3244887521/
- 4. What Functions? To use functions or not to use • User Defined Functions functions, that’s today’s question. – Scalar – Inline • Aka Table-Valued – Multi-statement • Aka Table-Valued CLR Functions Out of Scope http://www.flickr.com/photos/mshades/3154733910/
- 5. Goals 1. Identify purposes for creating User-Defined Functions 2. Discuss the types of User-Defined Functions 3. Demonstrate performance impact in selecting different types of functions
- 6. Session Warranty • Not a deep dive – Inform those that don’t know – Arm those that do know • Not included – Secret sauce – Pixie dust – Bacon http://www.flickr.com/photos/shawnzam/31302636/
- 7. A Function is a Function • Benefits – Consolidation – Modularization – Maintenance – Reuse • Performance – Looking at impact versus gain http://www.flickr.com/photos/vincentsl/3543888150/
- 8. A Function is a Function Scalar Functions Multi- statement Inline
- 9. A Function is a Function Scalar Functions Multi- statement Inline
- 10. Scalar Functions Body Parameter(s) Value Statement(s)
- 11. Scalar Functions • Function – Single or multiple statement(s) – Returns single value • Usage – SELECT, WHERE, ORDER BY – CHECK Constraints – Computed columns
- 12. Scalar Functions • Cursor like behavior – Linear performance effect • Data Access • Executes on – SELECT – WHERE – ORDER BY http://www.flickr.com/photos/blockpartypress/2625954377/
- 13. Scalar Demo Demo 001 – Scalar Functions
- 14. Scalar Performance Basic Function - Basic Function - CPU Basic Function – Duration 9,000 IO 12,000 10,000 8,000 9,000 10,000 7,000 8,000 6,000 7,000 8,000 5,000 6,000 6,000 5,000 4,000 4,000 3,000 4,000 3,000 2,000 2,000 2,000 1,000 1,000 - - - Native Scalar Native Scalar Native Scalar
- 15. Scalar Performance Data Access - Data Access- Data Access - Duration CPU IO 300 250 40000 35000 250 200 30000 200 25000 150 150 20000 100 15000 100 10000 50 50 5000 0 0 0 1 10 100 1000 10000 1 10 100 1000 10000 1 10 100 1000 10000 Native Scalar Native Scalar Native Scalar
- 16. A Function is a Function Scalar Functions Multi- statement Inline
- 17. Inline Functions Body - Parameter(s) Table Statement
- 18. Inline Functions • Returns table parameter – Columns defined in RETURN statement • Parameterized View • DML Operations allowed – INSERT – UPDATE – DELETE • Query-able results
- 19. Inline Functions • Limitations – Simple operations – Duplication of access
- 20. Joining Functions • APPLY – CROSS = INNER – OUTER = LEFT OUTER • Pass in columns as parameters – Replaces ON clause
- 21. Inline Demo Demo 002 – Inline Functions
- 22. Inline Performance Basic Function - Duration Basic Function – Basic Function – 12,000 CPU IO 9,000 10,000 10,000 8,000 9,000 8,000 7,000 8,000 7,000 6,000 6,000 5,000 6,000 5,000 4,000 4,000 4,000 3,000 3,000 2,000 2,000 2,000 1,000 1,000 - - - Native Scalar Inline Native Scalar Inline Native Scalar Inline
- 23. Inline Performance Data Access - Data Access- Data Access - Duration CPU IO 300 250 40000 35000 250 200 30000 200 25000 150 150 20000 100 15000 100 10000 50 50 5000 0 0 0 1 10 100 1000 10000 1 10 100 1000 10000 1 10 100 1000 10000 Native Scalar Inline Native Scalar Inline Native Scalar Inline
- 24. A Function is a Function Scalar Functions Multi- statement Inline
- 25. Multi-Statement Performance Parameter(s) Function Table Didn’t we already see this?!?
- 26. Multi-Statement Performance • Returns table parameter – Columns defined in DDL • Flexibility of Scalar function • Flexibility of Inline function • Query-able results
- 27. Multi-Statement Demo Demo 003 – Multi-Statement Functions
- 28. Multi-Statement Performance Basic Function - Basic Function - CPU Basic Function – Duration 9,000 IO 12,000 10,000 8,000 9,000 10,000 7,000 8,000 6,000 7,000 8,000 5,000 6,000 6,000 5,000 4,000 4,000 3,000 4,000 3,000 2,000 2,000 2,000 1,000 1,000 - - - Native Scalar Inline Multi Native Scalar Inline Multi Native Scalar Inline Multi
- 29. Multi-Statement Performance Data Access - Data Access- Data Access - Duration CPU IO 1200 1000 140000 900 120000 1000 800 700 100000 800 600 80000 600 500 60000 400 400 300 40000 200 200 20000 100 0 0 0 1 10 100 1000 10000 1 10 100 1000 10000 1 10 100 1000 10000 Native Scalar Inline Multi Native Scalar Inline Multi Native Scalar Inline Multi
- 30. Summary • Performance is important – Other factors can be just as important • Know the limitations • Functions have a place
- 31. Resources • http://msdn.microsoft.com/en-us/library/ms189294.aspx • http://www.sqlteam.com/article/intro-to-user-defined- functions-updated • http://www.amazon.com/Transact-SQL-User-Defined- Functions/dp/1556220790/ref=sr_1_1?ie=UTF8&s=books&qi d=1268026139&sr=8-1
- 32. For more information… Name: Jason Strate Email: [email protected] Blog: www.jasonstrate.com Resource: jasonstrate.com/go/functions Products Services Foundation BI products to covert to a Microsoft BI Speed development through training, and Helping those who do not have the platform and simplify development on rapid development services from means to get into information technology the platform. Pragmatic Works. achieve their dreams.