Using dedup on different jobs for similar machines

[ceez]

So we have a few 2008r2 servers, 2012 servers, and Win7 machines that we are backing up.

At this time to take advantage of dedup all these systems are bundled together in a single backup job. This job runs every 2 hours. There are some systems like the win7 what we want to backup just once a week or once a day as well as some servers.

If I create different jobs I lose taking advantage of deduping, correct?

i.e.:
Win7 backup job (3 vm's) - e:\backup\win7backup\
Win08r2 backup job (4 vm's) - e:\backup\win08r2\
Win12 backup job (3 vm's) - e:\backup\win12\

The vrb and vbk files are stored in different folders so there wont be any deduping going on not to mention that the file names will be different based on the job itself.

Any way of being able to split my backups enjoying the benefits of dedup or is this just the way of the beast?

Thanks!
ceez

[Vitaliy S.]

Dedupe is performed within the job, so splitting your VMs into different jobs will make all three jobs use their own cache for dedupe.

If you need to split your job, I would still go this way. The most efficient dedupe can be achieved from using the same OS in the backed up VMs, so make sure you group similar Guest OS into the same job.

p.s. btw, you may want to consider using Windows Server 2012 R2 with dedupe option as a target repository for your backups too.