http://www.sqlservercentral.com/blogs/sql_awesomesauce/2010/11/11/pass-day-3_3A00_-keynote-live-blog/

Printed 2014/12/20 05:30AM

PASS Day 3: Keynote Live Blog

By Jen McCown, 2010/11/11

 I’m live blogging the PASS keynote…it’s 8:44 local time.   

Here’s the PASS website description of the day 2 keynote …this is the one everyone’s been talking about, the DeWitt keynote.  Remember to follow along with the livestream: http://www.sqlpass.org/summit/na2010/livekeynotes.aspx!

Rick Heigis:

Heh, we just applauded the internet.

PASS board of directors is a volunteer board.  They keep saying “best ___ EVER”, and I always hear “Best ___ EVAR!!!!”  Board Directors are now Douglas McDowell, Andy Warren, and Allen Kinsel.

New event in 2011: The PASS SQLRally, with precons in BI, DBA, De, Prof, and Dev tracks (including Grant Fritchey [http://scarydba.wordpress.com] , Kevin Kline, Patrick LeBlanc, Devin Knight, and more). #SQLRally is now open for you to register. Precon $299, conference $199, or both together for $449 before Dec 1. See www.SQLRally.com

Wow, next year Summit is October 11-14, with two precon days instaed of one pre and one postcon. Earlybird reg is under $1k. Cool, maybe we’ll be able to hit SQL Connections in November for the first time ever…maybe…

Send questions for today’s keynote using ASKDrDeWitt@SQLPASS.org #sqlpass Kids at home, play along!

 

Dr David DeWitt: Technical Fellow, Data and Storage Platform Division

We’re getting art by the SQL Server query optimizer. Every color in the pic represents different parameters in the TPC-H benchmark.  Today we’re talking about  SQL query optimization (another power lecture).  Fundamentals, leaving with an understanding of why optimizers sometimes produce bad plans. Query optimization is harder than rocket science.

100,000 ft view has “magic happens”‘ as the key component.

Query 8 of the TPC-H benchmark, there are 22 million alternative ways of executing this query.  Cost-based query optimization was invented by Pat selinger in the 70s. This remains the hardest part of building a DBMS. QO transfers queries into an efficient execution plan: parse [into logical operator tree], QO [physical operator], execution. Physical operators = how they do it, e.g. nested loop join, etc.

Example…


Copyright © 2002-2014 Simple Talk Publishing. All Rights Reserved. Privacy Policy. Terms of Use. Report Abuse.