Server 2012 Setup

[notesguru99]

Hi All,

I have a new server with Windows 2012 installed and 13TB of attached storage. So before I initialize the disks I wondered if there is an optimum config or setup?

I was planning to simply setup a 13TB D drive if you like as my backup repository and then enable the Windows Dedupe task. I'm aware that there is a 2TB file limit so I was going to create individual jobs for all of my VM's to keep the file size under 2TBs and then simply daisy chain them all one after another. Forward incremental too I think with a full backup once a week?

How does this all sound? Would you do anything different?

TIA
Stuart

[foggy]

Hi Stuart, using dedupe on a primary target for your backups is against our reference architecture. It is recommended to have some raw storage in-between for some shorter retention of backups.

[notesguru99]

Hi,
Thanks for the reply. I'm not sure what you mean with regards some 'raw storage' in between? Maybe I didn't explain fully, but I will also replicate my backups to an offsite storage and then we also drop some of the critical VM's on to tape for DR. So I have multiple copies of my backups, or do you mean something else?

TIA
Stuart

[foggy]

Ok, seems you're good in terms of 3-2-1 rule for multiple copies, just keep in mind that operational restores from a dedupe repository might lack performance.

[notesguru99]

Thanks, I will bear that in mind. As I understand I can turn 2012 dedupe on and off without too much fuss...

[skrause]

Just so you know, turning dedupe off does not automatically re-hydrate the data so the read penalties on restores will still be there.

[DaveWatkins]

You don't need to create individual jobs, just enable per VM chains on the repo. Also make suire you format it with /L and /A:64k even if you don't plan on using dedupe right now. It won't hurt and it means it's an option later (say after 2016 comes out)