• EN
Question

Recovering a backup virtual machines from a dead host

  • 21 December 2022
  • 9 comments
  • 498 views
S
Userlevel 2

Hi all,

Recently, my ESXi home lab host died, taking with it the main data store that contained all my working VMs including a Veeam Replication and Backup Community Edition VM. This VM was backed up several of the VMs on the ESXi host to an alternate location, so perhaps there is light at the end of the tunnel.

In the first instance, I will need to create a new VM for Veeam, then reinstall Replication and Backup and re-attach the backup store. However, when looking at https://helpcenter.veeam.com/docs/backup/vsphere/importing_backups.html?ver=110, there is a pre-requisite requirement that states:

  • The server from which you plan to import backups must be added to the backup infrastructure. Otherwise you will not be able to access backup files.

 

Does this mean I need to re-create to lost VMs in ESXi first, before I can import any backups? Do then need to be specific configurations that match the original VM (such as vHDDs)? Do they even need be fully recreated with OS? Of is it simply that I need to re-add the new ESXi host into the new Veeam backup console, then restoring from the backup will allow for the creation of the VMs? 

This is unclear to me as to what this means.

Thanks for any insight. 

 

9 comments

S
Userlevel 2

Hmm, reading this for the nth time, perhaps when the text refers to “server”, it means the store/share (in my case using NFS) when the backups where created. 

JMeixner
Userlevel 7
Badge +17

The new VM has to be access to your backup files from the previous VBR server. Then you can import the “old” backups into the new server

If you have a configuration backup from the previous VBR server, you can recreate the configuration of your previous VBR server with this.

Jochen (Joe) Meixner | Veeam Vanguard 2024 | Veeam Legend 2021 - 2023 | Veeam Certified Architect (VMCA) 2022 | Twitter: @JoMeix | BlueSky: @jmeixner.bsky.social
Mildur
Userlevel 7
Badge +12

To do a restore, you just have to re-add your virtual environment and the NFS share to a new VBR server, rescan the NFS share within Veeam and do the restore.

You can follow the blog article from Hannes:
Restoring infrastructure from scratch with Veeam

 

Best,

Fabian

 

Product Management Analyst @ Veeam Software
S
Userlevel 2

Stupidity on that front I think means that the configuration was lost. Veeam was only up and running for a short while before this failure, and initially the configuration was stored in the default location of the local Veeam VM drive (or course, now gone). I honestly can’t remember if I moved it to the share, but certainly this is something I will do better for next time.  

S
Userlevel 2

@Mildur - nice - thanks.

MicoolPaul
Userlevel 7
Badge +20

Stupidity on that front I think means that the configuration was lost. Veeam was only up and running for a short while before this failure, and initially the configuration was stored in the default location of the local Veeam VM drive (or course, now gone). I honestly can’t remember if I moved it to the share, but certainly this is something I will do better for next time.  

Don’t worry about that, without the configuration you won’t have the configuration of your backup jobs, or the details of the Veeam topology (proxy servers, repository servers & directories etc), or the wider topology (your vSphere environment).

 

For recover you just need to configure your vsphere environment within a new Veeam install to then restore to, that’ll be sufficient. As for the recovery itself, you could do an instant VM recovery if you care about having it back online ASAP, otherwise you can just restore the entire VMs to your host to bring online as/when they’re restored. As it’s a lab, it’s down to your preference there!

Michael Paul - #VeeamLegend | #VeeamVanguard | Veeam Certified Architect | VMware vExpert | https://micoolpaul.com | Twitter: @MicoolPaul | Mastodon: @micoolpaul@masto.nu | Bluesky: @micoolpaul.bsky.social
dloseke
Userlevel 7
Badge +6

The config is a “nice to have” item that speeds things along.  But it’s not necessary.  You can start a new backup server from scratch, attach to your repository, and then rescan the backups in and it’ll start rebuilding the restore points into the database.  And you’ll of course need to add your new host that you want to restore too.  The config database just keeps all of these things, plus the associated passwords (assuming that the config database backup is encrypted) so it just makes the recovery a whole lot easier.

I’m assuming your backups are current, but my homelab host failed a few months ago (motherboard let out the magic smoke) and I had good luck in grabbing a similar server and swapping in the RAID card and drives and recovering my VM’s from the disk and then migrating it to a “new server” using Veeam QuickMigration.  That was great because I didn’t then have to rebuild my Veeam servers.

 

Derek M. Loseke, Senior Systems Engineer | Veeam Legend 2022-2023 | VMSP/VMTSP | VCP6-DCV | VSP/VTSP | CCNA | https://technotesanddadjokes.com | @dloseke
S
Userlevel 2

A bit late, but I would like to report back this this all went pretty well.

One of the VM had multiple VMDKs of which I only wanted to restore one. This was fine, although i then needed to re-import the VM into ESXi, which lead to a slight issue in that the network VLAN had changed and ESXi would not let me select the new one. I have to remove the vNIC, and re-add a  new vNIC (thinking about it, I might have been able to hand modify the VMX file, but this worked as well). 

The other VM were restored as full VMs so I go the option to change the VLAN network ID during restoration. 

Thanks everyone for their input.

Scott
Userlevel 7
Badge +8

As long as Veeam can see the new ESXI and the backup files on a volume/datastore you are good to import them. No need to recreate the VM’s.

Comment


Terms & Conditions