Don't unterestimate the negative impact of compression on deduplication. Usually you will see better dedupe results if you store more copies (of the same backup) on the dedupe target. If you look at the job, you should look at the per object and overall dedupe ratio. Personally I think the overall dedupe ratio is not calculated correctly. Here is a per object result where I keep 14 copies on disk.
[Normal] From: BSM@CM.domain.com "MSX_BACKUP" Time: 08.09.2014 20:20:22
Target-side Deduplication Statistics for MSX.domain.com:/Microsoft Exchange Writer(Exchange Information Store)/Microsoft Information Store/0a4df680-5d47-4801-ae4f-4b7398d24f58/File "MSVSSW-APP".
Using device: "B2D_CM [GW 1584:1:6136511032400778809]@CM.domain.com":
Mbytes Total: ................. 74042 MB
Mbytes Written to Disk: ....... 169 MB
Deduplication Ratio: .......... 438.1 : 1
While the Overall Deduplication Ratio is 4.8 : 1 for the job. This must be due to some very small VSS objects that have a dedupe ratio of 1:1. You can share a full session report and details on how many full/incr you keep on your store.
Regards,
Sebastian
