April 18, 2015 at 2:02 pm
Comments posted to this topic are about the item NONclustered 2 clustered
April 18, 2015 at 2:11 pm
Thanks for this question which seems very simple and evident ( at least it is my own opinion ).
I am surprised that only half of the voters have found the good choice.
And thanks also for the link to the sp_help procedure for which I have forgotten the columns returned for indexes.
April 18, 2015 at 6:50 pm
patricklambin (4/18/2015)
Thanks for this question which seems very simple and evident ( at least it is my own opinion ).I am surprised that only half of the voters have found the good choice.
And thanks also for the link to the sp_help procedure for which I have forgotten the columns returned for indexes.
Well, the numbers were pretty small when you answered - hardly a significant sample.
Nice simple quesion - but I'm not sure the right answer is evident; yes, it's clear that this is what the engine should do - but that's no guarantee at all that that's what it will do.
Tom
April 19, 2015 at 12:03 am
Very nice question, had to test it before answering 😉
And the shortcut 4 exec sp_help dsa
is highlight the table and press Alt + F1
Thanks & Best Regards,
Hany Helmy
SQL Server Database Consultant
April 20, 2015 at 12:17 am
Nice Question.. Index remains same irrespective of altering the table by adding a constraint and then by adding an unique clustered index.. I believe the index should have been dropped and recreated as unique clustered primary key in the background. Please correct me if wrong or throw some light on its internal.
Note: 10 points in 5 days!! 😀
______________________________________________________________Every Problem has a Solution; Every Solution has a Problem: 🙂
April 20, 2015 at 12:51 am
karthik babu (4/20/2015)
Nice Question.. Index remains same irrespective of altering the table by adding a constraint and then by adding an unique clustered index.. I believe the index should have been dropped and recreated as unique clustered primary key in the background. Please correct me if wrong or throw some light on its internal.Note: 10 points in 5 days!! 😀
From BOL:
If the index enforces a PRIMARY KEY or UNIQUE constraint and the index definition is not altered in any way, the index is dropped and re-created preserving the existing constraint. However, if the index definition is altered the statement fails. To change the definition of a PRIMARY KEY or UNIQUE constraint, drop the constraint and add a constraint with the new definition.
April 20, 2015 at 3:51 am
This was removed by the editor as SPAM
April 20, 2015 at 5:28 am
What am I missing here? The BOL page for CREATE INDEX [url=https://msdn.microsoft.com/en-us/library/ms188783.aspx]https://msdn.microsoft.com/en-us/library/ms188783.aspx[/url] clearly mentions that:
The index type cannot be changed by using DROP_EXISTING.
Thanks & Regards,
Nakul Vachhrajani.
http://nakulvachhrajani.com
Follow me on
Twitter: @sqltwins
April 20, 2015 at 6:50 am
Thanks, Carlo, for a good question. It gets the brain moving on a Monday.
April 20, 2015 at 9:17 am
Thanks for the question. Had to think a bit about possible answer two.
April 20, 2015 at 9:23 am
Nice one, off the beaten path. Thanks, Carlo!
April 21, 2015 at 3:07 am
Nice question, thanks.
Need an answer? No, you need a question
My blog at https://sqlkover.com.
MCSE Business Intelligence - Microsoft Data Platform MVP
Viewing 12 posts - 1 through 11 (of 11 total)
You must be logged in to reply to this topic. Login to reply