Hi!

I'm guessing that you use full recovery model when the job is running, change recovery model to bulk logged and when it's done change back to full.

This will reduce the...