Cardinality Regression

Question

Cardinality Regression

Viewing 10 posts - 16 through 24 (of 24 total)

You must be logged in to reply to this topic. Login to reply

SQLRNNR SSC Guru Points: 281334 More actions · Answer 1

free_mascot (6/1/2014)
Thank you, Steve. Easy One.

I am curious why you think this was an easy question.

Jason...AKA CirqueDeSQLeil
_______________________________________________
I have given a name to my pain...MCM SQL Server, MVP
SQL RNNR
Posting Performance Based Questions - Gail Shaw[/url]
Learn Extended Events

Thomas Abraham SSChampion Points: 10761 More actions · Answer 2

Got it right only because I've been paying attention. And because I answered it without consulting BOL!! 😉

[font="Verdana"]Please don't go. The drones need you. They look up to you.[/font]
Connect to me on LinkedIn

richardd Hall of Fame Points: 3965 More actions · Answer 3

SQLRNNR (6/2/2014)I am curious why you think this was an easy question.

The fact that it's almost identical to the question posted on 15th May helps! 🙂

http://www.sqlservercentral.com/questions/SQL+Server+2014/108889/[/url]

The date on that question says 2nd June, so I assume there's a technical glitch somewhere.

SQLRNNR SSC Guru Points: 281334 More actions · Answer 4

richardd (6/2/2014)
SQLRNNR (6/2/2014)I am curious why you think this was an easy question.
The fact that it's almost identical to the question posted on 15th May helps! 🙂
http://www.sqlservercentral.com/questions/SQL+Server+2014/108889/[/url]
The date on that question says 2nd June, so I assume there's a technical glitch somewhere.

There is that, but since that question the documentation on BOL has been fixed.;-)

Jason...AKA CirqueDeSQLeil
_______________________________________________
I have given a name to my pain...MCM SQL Server, MVP
SQL RNNR
Posting Performance Based Questions - Gail Shaw[/url]
Learn Extended Events

batgirl SSCarpal Tunnel Points: 4979 More actions · Answer 5

SQLRNNR (6/2/2014)
richardd (6/2/2014)
SQLRNNR (6/2/2014)I am curious why you think this was an easy question.
The fact that it's almost identical to the question posted on 15th May helps! 🙂
http://www.sqlservercentral.com/questions/SQL+Server+2014/108889/[/url]
The date on that question says 2nd June, so I assume there's a technical glitch somewhere.
There is that, but since that question the documentation on BOL has been fixed.;-)

The two questions are actually opposite.

5/15 said I'm running the "old" version (by virtue of compatibility mode), how can I test the "new" version.

6/2 said I'm running the "new" version and it doesn't work, how can I run the "old" version.

wonderchic SSC Enthusiast Points: 197 More actions · Answer 6

On May 15th the question of the day was similar, but it was about forcing the new cardinality on a db w/ the compat mode set at 100, and the correct answer was to use the trace flag 2312, not sure how now the correct answer can be that's the flag used for forcing the old cardinality?

Kevin SSCarpal Tunnel Points: 4962 More actions · Answer 7

Please update this QOTD

The documentation now states.

you can run the query with trace flag 2312 to use version 120 (the new version) of the cardinality estimator

PChiragS SSCarpal Tunnel Points: 4965 More actions · Answer 8

PChiragS

SSCarpal Tunnel

Points: 4965

June 9, 2014 at 10:01 am

#1720152

good question thanks steve.

Michael Richter-564003 SSC Veteran Points: 237 More actions · Answer 9

In my opinion, Its not a Feature, ist a bug and fixed asap

Kevin SSCarpal Tunnel Points: 4962 More actions · Answer 10

I just tested this on my 2014 server and this is what I observed

Running a query with trace flag 2312 will use version 120 (the new version) of the cardinality estimator when the database compatibility is 110 (SQL 2012)

Running a query with trace flag 9481 will use version 70 (the old version) of the cardinality estimator when the database compatibility is 120 (SQL 2014)