Database growth excessive when changing varchar length from 50 to 100.

Hi all,

I'm trying to get an understanding of a serious problem I have with a large DB in production. This is going to be obvious to someone (everyone probably) <bg>

I have a table which consists of numerous varchars and ints but also a Text type field. This table resides in a SQL 2000 Database. This DB currently has a data file size of 16Gb and a Transaction Log size of 17Gb. When I edit the table and increase the size of a Varchar field from 50 to 100 these files grow to more than double their size!

Why is this happening and how can I prevent this?

TIA

NozFx

Hi all,

Ok made two huge mistakes here. One I have large text data along side other fields that when changes are made the structure changes dramatically. Secondly I have full transaction logging in which all changes are logged. Going to have to make sure I learn my lessons here. Got all the sizes down now but I needed double to do that.

NozFx

sql

database creation durations

Trying to bring up a server at a DR site and running into what I think are
excessive durations for .mdf / .ndf file creation. It takes SQL server 1.5
hours to grab each 50 GB file. Disk subsystem at the site is a local
Ultra-Wide SCSI. RAID5 array.
Is 1.5 hours for a 50 GB excessive? What metrics can I use to monitor?
Oscar ElleseffIs this a 3 disk Raid 5? If so that is the worst performing Raid array you
can have for write intensive operations. Depending on the processors and
disks this may not be out of the question. If you upgrade to SQL2005 you
can take advantage of the zero-initialization feature for data files and
this will be cut down to probably a minute or less.
--
Andrew J. Kelly SQL MVP
"Oscar" <Oscar@.discussions.microsoft.com> wrote in message
news:B9334E0E-8AF1-4C79-9E0D-5E909C541E54@.microsoft.com...
> Trying to bring up a server at a DR site and running into what I think are
> excessive durations for .mdf / .ndf file creation. It takes SQL server
> 1.5
> hours to grab each 50 GB file. Disk subsystem at the site is a local
> Ultra-Wide SCSI. RAID5 array.
> Is 1.5 hours for a 50 GB excessive? What metrics can I use to monitor?
>
> Oscar Elleseff
>
>
>