I'm wondering if there's such thing as an "incremental restore"?
Say I restore a folder from a good backup to a separate drive, then continue to backup the original folder. Something changes in the original folder and gets backed up. If I then restore *that* backup over the top of the folder I restored to the seperate drive, will I restore only the changes? Or the entire folder?
Here's the "long" version, with the actual scenario I'm looking at.
ServerA is a Windows 2008 VM on a VMWare 5.1 host. Within Windows 2008, an iscsi connection was made to a NAS and the storage there is now the N: drive within ServerA. (Yes, that's real. Rather than attaching the NAS to the ESXi host as external storage, it was done this way. It was long ago, before we were experts.
) Files on the N: drive take up about 5Tb of space.
Not only does VBR *not* backup iscsi attached drives, but that NAS is old and needs to be retired. (Veeam agent for windows backs up iscsi attached drives, but that's not a long term solution.)
So I'm spinning up a new wizbang VMWare 6.0U3 datacenter stack with gobs of VSAN space and moving servers to it. I of course want to move ServerA to the new datacenter and VSAN, and move everything from the iscsi NAS space to a VMDK on that VSAN.
I can only have the N: drive "down" for a night, or maybe a weekend.
So here's my thought on how to accomplish this:
1. Backup the N: drive via Veeam Agent for Windows
2. Restore that backup to a VMDK on the new VSAN
3. Rename the N: drive to Z:
4. Attach the restored VMDK to ServerA and assign it to N:
Problems with this scenario:
Step 2 is going to take about 3 days, maybe more. In the mean time, folks are still busily using the N: drive. It can't be down for the amount of time it will take to restore it to a VMDK. How am I going to capture the changed files?
While the restore (N: drive to a vmdk, taking about 3 days) is happening, I'm still doing Agent backups on that N: drive and capturing the incrementals. This only takes an average of 45 minutes.
Once the restore is complete, I'm hoping I can:
1. Mount the VMDK to ServerA and give it a drive letter; say Y:
2. Kick everybody off the N: drive for the evening.
3. Do one last incremental.
4. Then restore *that* incremental to the Y: drive.
5. Change drive letters
If I do this, will I only be restoring the delta files? Will it only take as long as restoring the changed files? Or is Veeam going to restore the *entire* 5TB again, taking another 3 days?
My apologies for the wall of text, but this is the scenario we're in, and I need to find a solution.
Thanks for your help!