You could also try to use this code. It should generate column definitions for your views. copy/paste them from results and wrap into CREATE TABLE statement.

select column_name + ' '...

The quick solution is to use select..into to create tables. Then you have to script the tables and adjust constraints and nullability.

Regards

Piotr

Are you going to create a new database for each user?

The installation process can copy database to any location that you want and then you can attach the database using...

Please have a look at

select * from information_schema.columns

Regards

Piotr

How about this

select 'alter table ' +

quotename(object_schema_name(parent_object_id)) +

'.' + quotename(object_name(parent_object_id)) +

' drop constraint ' + quotename(name)

from sys.objects where name in

(select constraint_name from information_schema.check_constraints)

Regards

Piotr

If you are using SSMS 2008 you can try to script the table containint the columns as DROP..CREATE script. SSMS generates separate drop and create statements for all constraints on...

Judging by the name, 'lmdcol1_ck' is a check constraint. You have to remove constraints and indexes referencing the column before you change its collation. If you have to do...

Have you considered DDL triggers? You could have a trigger that would log all DDL changes and could rollback any alterations to table schemas.

Regards

Piotr

The only way I see is to write a CLR stored procedure that will go to the url to fetch the XML. But the assembly will have to be configured...

You can use 'max server memory (MB)' setting to limit the amount of the memory used by SQL Server. Use it with care.

You said the server is 'slow'. What do...

I understand that servername is not your local machine?

You have to log on to the server and run SQL Server Surface Area Configuration tool. There in Surface Area Configuration for...

Having said that, wouldn't you be looking for something like

select #tmp.key1, #tmp.key2 FROM #tmp

inner join

(

SELECT key1 FROM #tmp

EXCEPT

SELECT key1 FROM #tmp2

) q on #tmp.key1 = q.key1

?

Regards

Piotr

I just wonder, how set based solution would scale for thousands of rows. I have impression that for data islands problem loop based solution might scale better.

Piotr

Steve's idea of increasing fill factor is quite interesting and it may decrease the frequency of index rebuilds. Maybe there is a way to analyze if the clustered index required...

It could be that an index was dropped at some stage or the amount of data changed execution plan and another index is used/not used. Lock escalations could probably lead...