This is what i would use in such a case.

select col1,col2  into #abc from tbl

group by col1, col2

having count(*) > 1

delete tbl

from #abc

where tbl.col1...

if you can not set deadlock priority it in the SQL Code fired by that session. then simply kill the other process causing deadlock ....

kill 376

-- Amit

kill 376

-- Amit

I made the following tables and view which are as per requirements of Partitioned View. Strangely the queries that reffer to partitioned columns using operators such as = & in  gives proper...

As per BOL,

There is nothing called PKM 

-- Amit

Don't remove option to answer old questions. this would be very much the case when someone comes back after few days (rom a vacation, leave etc.)

I suggest you remove...

I did came across simmilar situation. Just check the enviornment settings "path " (goto command prompt. and type PATH)

also test by giving command dtsrun from command prompt (from c:\ only)...

* shrink db / log files for simple recover mode database

* frequently recompiling procs

* fragmented indexes that needs

* identity gaps fix

* remove duplicate records from a table...

you need atleast one full backup that was taken before your deleted happened. and  subsequent transaction log backup(s) taken after your delete has happened. (assuming no one truncated tran log...

use following to view contents of page 15 :1 : 40595  ( db id : file id : page number)

dbcc traceon (3604)

dbcc page (15 ,1, 40595)

-- Amit

If your webserver and database were on same machine initially, they would have used named pipes for communications, which is a bit faster them TCPIP. try switching to named pipes...

As per your requirements, Tables requires ID, Name, CreationDate, Owner, FullTextCatalogName, FullTextIndex, FileGroup, ClusteredIndex, PrimaryKey, IsAnsiNullsOn, IsQuotedIdentOn

Just to check, You are thnking of a queries such as

February 11, 2004 at 2:56 am