I guess that is good if you can be sure about that just a refresh would do it.

I've seen quite a few DWHs where some query runtime gets into the hour range (most time because of the AdHoc attack and under educated users). They schedule their queries somehow (like with a WAITFOR TIME/DELAY) and are pretty angry when they get back in the morning seeing their query being killed.. But then again. They would know and expect this behaviour with a well documented DB switching procedure.

Good article Leo!

Regards, Hanslindgren