We keep enhancing our dedupe features. We started version just to make sure we high light prominent feature, which came in same service pack.
V5- Major feature are Garbage colletion and journal pruning.
Garbage collection : our dedupe prunign works based on reference count for given signature,so when given job get purne we have to go back and touch our primary table to decrement reference count. Primary table is activly used but backup job too, so our purning share time with backup jobs, with this change we stop counting reference count for pruning and instead we keep bitmap to keep track what is valid and not valid. this help improve performance of backup.
one afile per secondary file: there concern that DDB folder size keep increasing over the period of time. the reason is secondary file not getting trucated. current implementation we have sixteen archive file per one secondary file, if extrended retention is mixed with normal rentention those secondary file sizes dont go down. we have enhance this in V5 now we have 1 afile per secondary file so DDB folder size keep in check.
Journal pruning during reconstruciton: we keep track of prunign activity in journal as we no longer keep the track of reference count, so during reconstruction those journal entries with get played back which helps in recostrution time.
if you are creating new DDB in SP16 or later you can enable these check box on DDB property sheet before running first backup.
how can you get the V5 : for current DDB you need to run compact operation to convert 16 archive file to 1 archive file. Open support ticket they can help you on this.
Above are high light feature with bug fixes for V5