I do it with a "FULLSCAN" every time a I migrate a DB from 2k to 2k5

By the way "1204" is OK; "1205" is overkill !!!!

Without indexes your table could be *very* fragmented ... so clustered index will put it under control. The fact the a reboot "fixed" the problem is something to be really...

sp_spaceused does *not* reports the actual values unless you update the ussage of the the database ( dbcc updateusage (0) ) because of that sp_spaceused gives you an extra parameter...

compatibility level does *NOT* introduces performance gain/losses basically compatibility "issues" or lack of them 😀

Note that pausing distribution agents is not exactly pausing "replication" is just pausing the subscriber activity; the publisher is still inserting in the distrigution database (logreader).

rami.atedgi (12/4/2007)

---

Hello everybody,

My query is like this:

CREATE TABLE MFGINSTRUCTIONS

([APPROVEDAT] datetime ,[APPROVEDBY] varchar (15) ,

[CREATED_BY] varchar (15) ,[EXPDATE] datetime ,

[INSTRUCTIONCODE] int ...

Backup and then restore it with a different name ( and different filenames )

No matter how you slice it 42 million rows will KILL anything ( Client or Server). You should check is that is really all needed. All I can...

You don't have to be an SQL Server Ranger to know that there is something terribly wrong with 34 indexes in one table.

By the way a quick way to determine...

Deletes will definitely benefit "AS LONG AS" partitioned colums are specified on the WHERE clause, Inserts will take advantange of the partitioning too. Seriously doubt that Truncate will have any...

I guess the Question poster just went strictly by what was "written" in BOL 😉

ms-help://MS.SQLCC.v9/MS.SQLSVR.v9.en/sqlexprbol/html/279fe698-5107-4891-b669-b04217ee421e.htm

But the description clearly says:

Shared Memory

Lets you connect to an instance of SQL Server...

But Regardless the answer that an sql trace...

U (12/3/2007)

---

noeld (12/3/2007)

---

U (12/3/2007)

---

I don't mind running Profiler in a Dev environment but when there is a problem in Production environment I would rather prefer a tool that will quickly...

if you can make that job into a set of stored procedures you can publish the "EXECUTION" of the stored procedure and it will effectively bypass the distribution database activity.