Currently handling cloud to cloud migration using a custom toolkit - under the hood we're running volume transfers with single host instances attached to each cloud that serializes/deserializes resource files. Problem is on a 10Gibs network running a 300Gb disk it takes 30 minutes but 700GB disk takes 7 hours......... Any technical folks have any high level optimizations? Not sure of network topology on the 700GB machine (still in the dark here) but my assumption is the 300GB is barely used. I thought about simple snapshots and streaming content to a block device on the destination cloud but there is some overhead with this approach.