Restore Mode - "Restore to new location, or with different settings"

[benny_rt2]

Veeam Backup & Replication 9.5 Update 4
Version: 9.5.4.2753
Community Edition

ESXi 6.7u1

Veeam is not restoring the VM virtual disks to the original datastores. If I override and manually tell Veeam the specific datastores to restore each of the VMs disks, Veeam was restoring the disks to those specified datastores but in random order.

For example:

If a VM has multiple virtual disks, and each disk of the VM is on different Datastores

VM BRTADEV3 -
vhdd 1 is on datastore DS01 (DS01 was the 1st/primary VHDD when first setting up the VM)
vhdd 2 is on datastore DS02
vhdd 3 is on datastore DS03

If I try to do a full restore using Restore Mode = "Restore to the original location" Veeam will try to restore all three virtual disks to the primary BRTA-DS01 datastore rather than to their original respective datastores. I can work around that by choosing "Restore to new location, or with different settings" when the Restore dialog box is asking for the Restore Mode. When choosing the later Restore Mode Veeam will ask me specifically which vhdd to restore to which datastore.

virtual disk 1 restore to datastore DS01
vhdd 2 restore to datastore DS02
vhdd 3 restore to datastore DS03

Veeam honors that the three datastores (DS01, DS02, DS03) are to be used as the designated three datastores, but it doesn't honor which vhdd goes to which datastore. In other words, what Veeam does is it restores the vhdds to those three datastores, but in seemingly random order:

vhdd 1 restores to datastore DS02
vhdd 2 restores to datastore DS03
vhdd 3 restores to datastore DS01

Thanks,

Benny

[Dima P.]

Hello Benny,

Can I ask you to raise a support case and share the case ID? Thank you in advance!

[PetrM]

Hello Benny,

Just an idea: mentioned datastores (DS01, DS02, DS03) are in a cluster? If it's true Storage DRS (if it's enabled) provides recommendations for virtual machine disk placement so it might be interesting to test the same scenario with another vm (just create vm with empty disks?) which has its disks on datastores that don't belong to a cluster. Thanks in advance!

[benny_rt2]

Hi again,

I was in error when describing what was occurring. I did another full restore yesterday (the first full restore since creating this thread) and this time I noticed better what is happening.

Veeam is restoring the vmdk files to their correct respective Datastores (good) which I specify by using the option: "Restore to new location, or with different settings".

However, in vCenter when I look at the guest VM "edit settings" after a full restore it looked weird which is why I was thinking in my original post that veeam placed the vmdk files on the wrong datastores.

What really is happening, the various vmdk files are on their correct Datastores and their respective SCSI controllers are correct... but the labels given by vCenter such as "Hard Disk #" are incorrect on a few:
Hard Disk 1 - [DS05] BRTADEV4/BRTADEV4_1.vmdk SCSI 1:0
Hard Disk 2 - [DS06] BRTADEV4/BRTADEV4_2.vmdk SCSI 2:0
Hard Disk 3 - [DS03] BRTADEV4/BRTADEV4.vmdk SCSI 0:0
Hard Disk 4 - [DS07] BRTADEV4/BRTADEV4_3.vmdk SCSI 3:0
Hard Disk 5 - [DS08] BRTADEV4/BRTADEV4_4.vmdk SCSI 3:1

should be showing it like this instead:
Hard Disk 1 - [DS03] BRTADEV4/BRTADEV4.vmdk SCSI 0:0
Hard Disk 2 - [DS05] BRTADEV4/BRTADEV4_1.vmdk SCSI 1:0
Hard Disk 3 - [DS06] BRTADEV4/BRTADEV4_2.vmdk SCSI 2:0
Hard Disk 4 - [DS07] BRTADEV4/BRTADEV4_3.vmdk SCSI 3:0
Hard Disk 5 - [DS08] BRTADEV4/BRTADEV4_4.vmdk SCSI 3:1

One caveat though is that my existing backup job for that VM of which I did a full restore, the restore the old guest VM of the same name (of course,because it was a full restore). But now that existing backup job for BRTADEV4 doesn't work because it says "Virtual Machine BRTADEV4 is unavailable and will be skipped from processing. Nothing to process. All machines were excluded from task list."

I'll need to create a new backup job with a full backup - I cannot continue incremental of the original backup job after the full restore? I guess that makes sense since the full restore was from days ago.

Thanks

Benny

[PetrM]

Hi Benny!

I'd suggest to open a support case and share case ID for our reference.

Thanks!

[benny_rt2]

Hi PetrM,

I think I'm good for now. I've probably made this thread sound confusing or convoluted, but basically the veeam backup and restore is working as it should. The vmdk files are restoring to their respective datastores correctly and the virtual controllers for those v.disks are good too. in vCenter when in "edit settings" of the guest it shows the labels such as "Hard Disk #" incorrectly, which doesn't really matter. To resume my scheduled backups of this particular VM all I had to do is edit the backup job and re-add the newly restored guest vm to that backup job. When I then started the backup for that VM it rebuilt the main full backup file on the repository.