"for all servers" -- once you get it working on one server, create a monitoring template using that one as the source, then apply to the remaining.
As for the 100% for a period of 3 hours -
Show all metrics
Click on disk busy 3 pencils icon.
Click edit (this is KEY)
Then enter the warning/critical *and* occurrences.
So if 15 minute collection intervals you would say 12 occurrences.