A UNION statement is an aggregation, so, unless you're doing UNION ALL, it's not the sum of the three queries, but the aggregation of the three. Check the execution plan...

You can specify a stop time for a trace. So you just need the job to start it and then it will run until the @stoptime property. You can read...

Those are waits related primarily to disk. Are you backing up across the network? Otherwise, I'd focus right on the disk to see what issues there might be there.

WHOOP! Can't wait to use the robot for presentations.

"ASK A QUESTION! YOU HAVE 10 SECONDS TO COMPLY. 3.... 2.... 1...."

<snicker>

I've never seen SSMS by itself cause any issues and I've left it open for days at a time.

Ummm... You might not want to move to working as a DBA if network admin was stressful. DBA work becomes hyper-stressful. Think about it, one of the single most important...

This will put more load on the production system than using the backup and restore process.

Even if you do tests to determine an approximate run time, you can still hit resource contention or blocking issues that cause that time to change. This is a tough...

Partitioning is not primarily a performance tuning mechanism. It is primarily a data management mechanism. There are cases where you can get performance improvements, but you need to remember that...

The only suggestion I had for the higher versions of SQL Server is that you could use the ROW_NUMBER to break up the deletes into smaller groups. You still have...

Nope. You got it. If you're inserting to a table with a clustered index, it's inserting into the clustered index. That event, by itself, is absolutely not a performance problem.

A clustered index stores the data. So when you run a delete statement, it goes to the clustered index to delete the data. That's how it works.

As I said before, your choices in 2000 are more limited. Filter the data on additional criteria to reduce the batch size so you're deleting in smaller chunks and providing...

We're still working on that.