Just a small addition to nyulg's last point:-

-------------------------------------------------

-if you are modifying the totallity of your table it could be faster to copy all your data with modifications to a new...

If you aren't bothered about recovery, the fastest way to do it will be to use SELECT INTO (because SELECT INTO is not logged to the transaction log).

e.g.

SELECT 0 as...

To clear the transaction log just run 'BACKUP TRAN dbname WITH TRUNCATE_ONLY'

As to the actual database space, SQL Server doesn't always re-use space freed by a DELETE command (SQL 2000...

Afraid not. In most cases, a severity level of 16 or above is deemed a fatal error that won'r reach your test for @@Error.

However, some level 16 errors are not...

Not all errors can be handled inside your stored procedure.

It all depends on the severity level of your error. In a lot of cases you will never reach the...

Each time you run your import, SQL Server will write the transactions to your transaction log, and so your log will continue to grow until you back up either the...

That depends on your client.

If you are running the INSERT statement from Query Analyzer, then before you run the INSERT statement, just run the SET ANSI_NULLS ON and SET ANSI_WARNINGS...

Your error message is complaining about the connection that is issuing the INSERT into [scheme].[slcustm] table, not the connection that created the Trigger, or the options that the trigger is...

Drop and re-create the table

or

DBCC CHECKIDENT ('table name', RESEED, seed value)

e.g.

DBCC CHECKIDENT ('MyTable', RESEED, 0)

This is a SQL Server restriction. A table can have only one Timestamp column.

If you need more than one timestamp, then you will have to consider partitioning the table into...

You are probably inserting blank into the date (which defaults to Jan 01 1900) instead of NULL e.g.

INSERT INTO TableX (DateCol,Col1,Col2) values (NULL,'Col1Value','Col2Value')

I suspect you are doing something like

INSERT...

Have you tried specifying the domain username in the format DOMAIN\USERNAME?

Edited by - ianscarlett on 11/25/2002 03:50:19 AM

Bulk insert is supposed to be faster - up to twice as fast, because it interacts with SQL Server at a lower level than BCP.

Do you have an INSERT trigger that does an UPDATE?

If the column isn't indexed, there should be no degradation in performance as the index won't need updating.

By the way, each non-clustered index is a pointer to the clustered index,...