joshua13
07-31-2007, 12:00 PM
I had to copy c. 1 terrabyte drive (c 50gig free) to a newly formatted drive. Block copy was not suitable as it was not likely I'd be able to do this in one session. I consider SD to be one of the best and fastest ways to do this but I use other tools. I decided to run a little speed test. I ran SD for a while and then erased and started over with "sudo rsync -auxv --extended-attributes --delete /Volumes/d1/ /Volumes/d2" and ran that for the same time to see who copied more data. rsync won the race but as it copies in a different directory order it wasn't the same data so that was useless. I decided to drop that as I thought the results were likely to be uninteresting even if i set up a better test. I still wanted to do something else (i.e. kill two birds) with my big copy job so, so I decided to do the whole drive with rsync and then smart update with SD (without touching the target in between) to check on something I'd observed before. - SD copies a lot of things rsync just copied in such a scenario or which rsync thought unnecessary to copy by its checks in a case where the drives are already similar in contents. The results were that SD copied at least 200 gigabytes of data even though the drives were already identical to all appearances after rsync. So what criteria is SD using different than rsync - or just what's going on here?