Urgent - recovery/transfer from a proxmox server that no longer starts

[mfmconsulting]

Greetings,
I need urgent support...
The Proxmox server that was in production has a compromised boot disk.
Internally there are two Western Pro Digital HDD disks in Raid 1, where the images and disks are still present (Example vm-108-disk-0.qcow2 or the .vma files of the VMs)
I have an urgent need for the company to transfer these files in order to restore the primary virtual machine (domain controller) and therefore recover all the files inside it (one of the .qcow2 disks).

N.B. Attached image of some disks already transferred from the internal WD HDD disks to an external USB HDD connected to a Desktop PC in the same LAN

Gretings
Thank you
Manuele

 

[Kingneutron]

https://github.com/kneutron/ansitest/tree/master/proxmox


mkdir -pv /root/bin/boojum

Setup the BKPDEST.mrg in that directory and point it to separate storage (disk or NAS)


Run the bkpcrit script and it (among other things) will make a backup of /etc for you on that backup destination


The script has comments on some critical files:

lxc and VM definition files are in:

/etc/pve/nodes/proxmox

If you don't have backups, you will need these definition files to move the VM to another server.
If you don't have the VM/LXC config files, you will need to recreate them.

https://github.com/kneutron/ansitest/blob/master/proxmox/proxmox-recover-vm-disks-without-backup.sh



Best thing to do would be to make an immediate backup of your VMs and containers with proxmox integrated backup, to separate media or NAS, and restore onto another pve server.

Have an assistant start going through these scripts and working out a plan to use them (or something like them) going forward.

Come back later when the environment is back to a sane state and read the next part.

====================

Where are the backups? It sounds like this place needs to setup a proper backup regimen and document - and TEST - a DR process.

You should have daily, weekly, monthly backups and should be sending some of them offsite to a professional agency.

And you need to test your restores every so often. You cannot wait until a production server breaks to implement proper Disaster Recovery practices, and you need to talk to your bosses about this.

Best of luck on the recovery

P.S. look into clustering -- a production domain controller VM should not be on a single point of failure. Cluster should have been able to migrate the VM to another working node.

 

[UdoB]

Please take a look at https://shop.proxmox.com/index.php?rp=/store/proxmox-ve-premium for "Response time 2 hours within a business day (on critical support requests)".

Sorry, if this sounds rude, this is a community driven forum with (mostly) volunteers...

---
Edit: Yes, it sounds rude, after reading my first sentence a few minutes later. It was meant to be a helpful link in case you didn't know the shop yet. Good luck!