I think we should be careful not to overstate the problem here. Worst case scenario is where dedupe is enabled on a system that already contains a substantial amount of data. In this case you're going to get a lot of blocks written to during the dedupe process that will be subsequently marked as deleted but not zeroed out. Sdelete will fix those for you. However on a new volume, with data added over time, the dedupe process will probably be able re-use those deleted blocks rather than using new blocks, so the issue should be less evident. That's my theory anyway :-)
Clik here to view.
