Comments on: Functions, IO statistics and the Execution plan http://sqlinthewild.co.za/index.php/2009/04/29/functions-io-statistics-and-the-execution-plan/ A discussion on SQL Server Mon, 14 May 2012 23:51:41 +0000 hourly 1 http://wordpress.org/?v=3.3.1 By: Andy Y http://sqlinthewild.co.za/index.php/2009/04/29/functions-io-statistics-and-the-execution-plan/comment-page-1/#comment-1777 Andy Y Tue, 16 Aug 2011 18:17:14 +0000 http://sqlinthewild.co.za/?p=231#comment-1777 Just wanted to say thanks for this great blog post. I've always known the perf drawbacks of UDFs, but then a skeptical dev here asked me to prove it. It was only then that I realized that Exec Plans don't show everything. Dug around via Google to try and find the correct Profiler settings, and found your blog post which answered my questions! Thanks! Just wanted to say thanks for this great blog post. I’ve always known the perf drawbacks of UDFs, but then a skeptical dev here asked me to prove it. It was only then that I realized that Exec Plans don’t show everything. Dug around via Google to try and find the correct Profiler settings, and found your blog post which answered my questions! Thanks!

]]> By: Gail http://sqlinthewild.co.za/index.php/2009/04/29/functions-io-statistics-and-the-execution-plan/comment-page-1/#comment-411 Gail Tue, 12 May 2009 17:04:10 +0000 http://sqlinthewild.co.za/?p=231#comment-411 Yeah, I'm specifically and only talking about the scalar UDFs here. The inline table valued functions are, in fact, the only ones that perform well Yeah, I’m specifically and only talking about the scalar UDFs here. The inline table valued functions are, in fact, the only ones that perform well

]]> By: Alex Kuznetsov http://sqlinthewild.co.za/index.php/2009/04/29/functions-io-statistics-and-the-execution-plan/comment-page-1/#comment-410 Alex Kuznetsov Tue, 12 May 2009 15:50:28 +0000 http://sqlinthewild.co.za/?p=231#comment-410 While scalar UDFs may be a huge drag, the inline ones usually aren't. Just wanted to emphasize the distintion between scalar and inline UDFs: http://sqlblog.com/blogs/alexander_kuznetsov/archive/2008/04/21/not-all-udfs-are-bad-for-performance.aspx While scalar UDFs may be a huge drag, the inline ones usually aren’t. Just wanted to emphasize the distintion between scalar and inline UDFs:

http://sqlblog.com/blogs/alexander_kuznetsov/archive/2008/04/21/not-all-udfs-are-bad-for-performance.aspx

]]> By: Gail http://sqlinthewild.co.za/index.php/2009/04/29/functions-io-statistics-and-the-execution-plan/comment-page-1/#comment-409 Gail Sat, 02 May 2009 12:54:26 +0000 http://sqlinthewild.co.za/?p=231#comment-409 Looking forward to it. Looking forward to it.

]]> By: Mladen http://sqlinthewild.co.za/index.php/2009/04/29/functions-io-statistics-and-the-execution-plan/comment-page-1/#comment-408 Mladen Fri, 01 May 2009 23:34:23 +0000 http://sqlinthewild.co.za/?p=231#comment-408 i'll post a blog about it in a very near future... a bit busy at the moment. i’ll post a blog about it in a very near future… a bit busy at the moment.

]]> By: Gail http://sqlinthewild.co.za/index.php/2009/04/29/functions-io-statistics-and-the-execution-plan/comment-page-1/#comment-407 Gail Fri, 01 May 2009 21:35:00 +0000 http://sqlinthewild.co.za/?p=231#comment-407 That's not good. Normal CTE's are *supposed* to behave like views, not like this. That’s not good. Normal CTE’s are *supposed* to behave like views, not like this.

]]> By: Mladen http://sqlinthewild.co.za/index.php/2009/04/29/functions-io-statistics-and-the-execution-plan/comment-page-1/#comment-406 Mladen Fri, 01 May 2009 20:12:59 +0000 http://sqlinthewild.co.za/?p=231#comment-406 no, they're not recursive. i do however have calls from cte6 to cte2 and cte3 etc... no, they’re not recursive. i do however have calls from cte6 to cte2 and cte3 etc…

]]> By: Gail http://sqlinthewild.co.za/index.php/2009/04/29/functions-io-statistics-and-the-execution-plan/comment-page-1/#comment-405 Gail Fri, 01 May 2009 19:33:56 +0000 http://sqlinthewild.co.za/?p=231#comment-405 Recursive CTE's by any chance? From what I've heard, and I haven't played with them much myself, with a recursive CTE the exec plan and Stats IO are for the first (anchor) query only, not the recursion. Recursive CTE’s by any chance?

From what I’ve heard, and I haven’t played with them much myself, with a recursive CTE the exec plan and Stats IO are for the first (anchor) query only, not the recursion.

]]> By: SQL Server and Cloud Links for the Week | Brent Ozar - SQL Server DBA http://sqlinthewild.co.za/index.php/2009/04/29/functions-io-statistics-and-the-execution-plan/comment-page-1/#comment-404 SQL Server and Cloud Links for the Week | Brent Ozar - SQL Server DBA Fri, 01 May 2009 15:45:23 +0000 http://sqlinthewild.co.za/?p=231#comment-404 [...] Functions, IO Statistics and Execution Plans - Gail Shaw does another awesome blog with T-SQL scripts and screenshots. [...] [...] Functions, IO Statistics and Execution Plans – Gail Shaw does another awesome blog with T-SQL scripts and screenshots. [...]

]]> By: Mladen http://sqlinthewild.co.za/index.php/2009/04/29/functions-io-statistics-and-the-execution-plan/comment-page-1/#comment-403 Mladen Fri, 01 May 2009 15:27:25 +0000 http://sqlinthewild.co.za/?p=231#comment-403 i knew about this for functions but i've noticed this behaviour with some CTE's too. all this totally wrecked my confidence in set statistics io. i knew about this for functions but i’ve noticed this behaviour with some CTE’s too. all this totally wrecked my confidence in set statistics io.

]]>