If you prefer to keep ad hoc querying available, you could use some form of source control for the queries.

I have to ask what leads from an interest in journalism to an interest in doing SQL, especially since you mention that you aren't familiar with what SQL is. ...

I'd be more inclined towards a dimension. Views don't do anything but hide code.

Bob Hovious (8/26/2009)

---

Since Flo brought up the TVF approach, I want to ask. Can anyone point me to documentation about the performance of TVFs v stored procedures? ...

That works too. And if you include extended properties and query those, you can have definitions in there that are business-meaningful.

It's a common enough requirement. But the only duplicate-safe solutions are also performance killers, or require a more complex implementation, of the sort that I outlined for you.

If you...

I'll second the CLR function call idea on this one.

You can also use OpenQuery (or one of its cousins) for this.

Titles are like certs and degrees. Some people have them because they reflect a degree of skill, knowledge, experience and wisdom. Some people have them because they passed...

Have you checked to make sure the metadata for that file matches the actual file?

Like, if it has the right drive and directory in sys.files, or not?

I know what you mean.

With that, it sounds like setting the fill factor low and rebuilding indexes frequently is your best bet.

That's certainly one way to do it.

I'd take a look at the documentation programs available from RedGate and ApexSQL. Might be more useful/efficient at it. Those are what...

A better solution would replace that whole proc with this:

declare @Client_ID varchar(10);

declare Deletes cursor local fast_forward for

select ClientID

from DeleteClientID

where RunQ = 0;

open Deletes;

fetch next from Deletes

into @Client_ID;

while @@fetch_status = 0

begin

update...

You'd want to use:

SET @MyCount = (SELECT count(*) FROM DeleteClientID WHERE RunQ = 0);

Use that at the beginning instead of setting to @@ROWCOUNT. That should give you what you...

I'm assuming the clustered index isn't something that just sequentially increments, or you wouldn't get that level of fragmentation. Is changing to something that increments possible?

Ideas for that would...