Windows Server, two repositories on one parition.

[markcboffins]

Hi,
Any advice/best practice on the following would be much appreciated.

I have a single server Veeam setup, Backup and Replication 9.5 update 4. A single data partition is formatted REFS 64kb. The first backup repository is on this partition, it is configured: use per-vm files and align backup file data blocks.

I have one backup job backing up 9 VMs from a Hyper-V cluster, this is configured: incremental, inline data dedup, compression optimal, storage optimisation: local target, keep 7 restore points. I am happy with this job, it is working well, the VMs actual selection size is 2.88TB and the backup files on disk are about 2TB.

What I am wanting to do is create a backup copy job to a second repository on the same data partition, and keep some weekly/monthly/quarterly/yearly recovery points.

I have done some testing, creating a second repository on the same partition(same settings as first), and also from what I've read, any refs dedup benefits are only applied/kept per job/repository. So in affect, my first copy job to the second repository will be the same size (2TB), and this will only grow as I keep more GFS recovery points.

I am looking to make the most of my (somewhat limited) storage on our single Veeam backup server, and keep a longer period of retention. Having done some more reading, I believe having two repositories on one data volume, the best way to go might be Windows Deduplication?

Cheers

Mark

[HannesK]

Hello,
the best practice is to follow the 3-2-1 rule and have one backup copy off-site. Especially do not keep it on the same machine. That's why we do not allow GFS on the primary backup job right now (well, that will change with V10 for other reasons...)

Everything else you have recognized is correct and "as expected".

According to the feedback in the forum the performance of ReFS block cloning + ReFS deduplication is not good. So I would not go that way.

If you are aware of the potential issues (single point of failure), then I would keep it as it is right now and change to GFS backup jobs with V10

Best regards,
Hannes

[soncscy]

Mark,

Just my 2 cents, but convince your accounting dept. that the $200 for an extra 8 TB drive on newegg is worth it, and shove it in some other server somewhere. Peace of mind is worth $200

[markcboffins]

Hi Hannes and Harvey,
Thank you for your replies. Absolutely agree with you to have backups elsewhere, we do have offsite backups done using a separate method. Just exploring options with Veeam and the benefits of REFS / De-duplication, again trying to extend the recovery points.

Hannes, GFS on the primary backup job in V10 sounds interesting, this sounds like it would do the job. I would then need only one repository local with one GFS backup job (keep many restore points), and then use a Veeam job to copy offsite to another repository / cloud.

Cheers

Mark