Large Storage Requirement

[thecompnerd]

All backup data must reside on a single volume, correct? If that's the case, let's suppose that my Veeam solution is virtualized. ESXi has a max VMDK size of 2TB. If I use all 2TB for my backups and need more, what do I do? Would I create another 2TB VMDK, and span (combine) the drives through Windows to get effectively 4TB of storage? Is that the best way to handle it? If I installed Veeam on a physical box, I wouldn't need to worry about it, other than the max disk size allowed by the OS.

[Gostev]

Correct. However, backup to VMFS is considered a bad practice, because it may significantly complicate recovery in certain disasters scenario, and make it simply impossible in other (there are existing discussions about this which you can refer to for more info). We highly recommend using proper backup targets instead, and when running Veeam in VM, the best choice is NTFS formatted LUN on iSCSI storage. This will provide you with up to 16TB storage per LUN. Thanks.

[larry]

you can back up to a iscsi or nfs connected to your veeam vm. I connect to a 9tb iscsi lun for some backups. The data from one job needs to go to one volume, other jobs can go some where else. I use a phyiscal box but send data to local raid, iscsi and nfs.

[thecompnerd]

Thank you both for the reply. I think I will stick go with a physical box as my Veeam backup server.

Larry,

If I want to get the most out of deduplication, wouldn't I want all of my backups to go to a single volume? Assuming that all of my VM's for that job/destination are similar in OS or contain other files that would be similar.

[larry]

I use a couple of jobs for a few reasons
how many copies ( restore points ) I save depends on the VMs.
Some VMs also go to tape. I send VMs backups to one Lun which goes to tape nightly. Some VMs have some disks in a job go to tape but other disks in a second job with no tape.
Cant fit all VMs to one LUN.
Some jobs run a few times a day others only at night.
Timing of replication and backups of some VMs. The fewer the VMs the easier the timing.
Splitting of jobs for Surebackup labs.
Vms that I might want to run a backup as a group manualy once in a while I keep togeter.
All replication I do one VM per job. I see no reason to do otherwise but a few reasons for one vm per job.

I also agree with "If I want to get the most out of deduplication, wouldn't I want all of my backups to go to a single volume? Assuming that all of my VM's for that job/destination are similar in OS or contain other files that would be similar"

[withanh]

I think I will stick go with a physical box as my Veeam backup server.

I use a virtual Veeam server and connect an iSCSI LUN formatted with NTFS to it as my backup target, bypassing VMware for this LUN. It's about 7.5TiB and works great. An advantage of using a virtual Veeam server is you can leverage the Virtual Appliance mode for backups.