Well, no, because that would be a violation of the definition of UNIQUE.

What you want to do is either clean out the duplicate values or re-define your index/constraint.

The big question...

In my Googling I found a few saying it was different, but for SQL 2000. Anyhow, even when invoked from a one-off schedule it still ran to completion just...

eccentricDBA:

This makes sense and fits with what I am seeing. Solution is bad, though (and what I'm trying to avoid) as the files are very large and getting down...

Steve,

This is the explanation I've always heard and understood to mean that the file size / space used would eventually even out, so that the smaller files with less data...

Nope.

As a minor point, I'm in SQL Server 2005, not 2000. (Though I'm fairly certain this behavior remains the same.)

But that's also not the question I am asking. ...

For stuff like that, you should investigate the (free) SQL Server Performance Dashboard tool. It should give you everything you need (and them some) in a usable format. ...

That depends upon how you set up the filegroups, etc. If you have only the PRIMARY filegroup and just files under that, the data will be loaded into all...

If you have an open query window to the DB from which you issue the ALTER DATABASE command, that open query window is the single user.

Check it out:

USE DBA_DB

ALTER...

You will need to do the detach from the same query window.

Make sure it returns to multi-user when you reattach.

Much better to set to single user mode.

ALTER DATABASE MyDB SET SINGLE_USER WITH ROLLBACK IMMEDIATE

No, users will not become "orphans" by kicking them off the DB. They just lose...

When is not so important (though it can impact index fragmentation). What's more important is how frequently are you growing the data files and by how much? Too...

Wahoo!

Sucks that you have to go through this but good that you can demonstrate the issue. Let's hope they fix it fast for you.

Only if you need to convert it. For rather silly business reasons/vendor application issues, I'm having to do this myself for one of my DB's. I don't care...

What Phil said. That's why I do the *+datavalue+* method quick check.

Some text editors have settings to show these return characters. I think WordPad does but I'm not...

That's it.

The last character (subloc 5) is a carriage return

(UNICODE 13)

see here: http://www.ssec.wisc.edu/~tomw/java/unicode.html#x0000

So what you have in that column is:

0.99[Carriage return]

Sounds like your parsing needs to account for the carriage...