Help: Recover proxmox VE environment?

[jmcdermo]

Hello everyone!

Thanks in advance for the help.

I’m in the middle of trying to help a friend who’s husband recently passed away. His service is a week from Saturday. He was quite a tinkerer and had many/most of the lights around their house configured with various IOT devices.

This whole environment is controlled by an NUC model NUC7i5BNK with a 512GB M.2 SSD
The NUC is running Proxmox Virtual Environment 4.15.18-10
In that Linux environment the Home Assistant.io software was controlling everything.

When I got here a couple of days ago, the NUC was hung at the startup of the logical volumes:
"A start job is running for Activation of LVM2 logical volumes" and it never completed.

Eventually my attempts to scan and repair what I thought was a damaged SSD drive rendered the system unbootable. /dev/sda1 no longer contains bootable information - bad superblock error.

/dev/sd3 still shows as being an LVM2_member with two partitions? pve-swap and pve-root and I can still successfully mount /dev/sd3 if needed.

If I went to get another M2.SSD to install in the NUC along with a USB harness that allowed me to attach the old M2.SSD at the same time. Would I be able to do a clean proxmox installation onto the new SSD and then somehow mount the old SSD's sd3 partition to migrate/move the contents over onto the new clean environment?

This is for a friend's family, trying to give them back control of all their IOT devices. As I said at the beginning, unfortunately, the person who set this all up passed away so I can't ask him if this system was backed up and if so how/where.

Any help will be greatly appreciated.

 

[oguz]

hi,

my condolences for your loss.

This whole environment is controlled by an NUC model NUC7i5BNK with a 512GB M.2 SSD
The NUC is running Proxmox Virtual Environment 4.15.18-10

that's a pretty old version to work with.

/dev/sd3 still shows as being an LVM2_member with two partitions? pve-swap and pve-root and I can still successfully mount /dev/sd3 if needed.

okay, hopefully the VM/CT disks are stored on that.

does the disk have the file /var/lib/pve-cluster/config.db? that database contains all the VM configurations -- make a copy of that if you want to restore them.

Would I be able to do a clean proxmox installation onto the new SSD and then somehow mount the old SSD's sd3 partition to migrate/move the contents over onto the new clean environment?

yes, if you have the VM disks and the configuration database, then you should be able to recreate the VMs.

other useful files might be:

* /etc/network/interfaces
* /etc/hosts
* anything else that seems out of the ordinary, maybe files in /root/ or in /home/ if there's anything there

i'd suggest you to make a new PVE installation (using the latest PVE7) and import the disks and the configs there.

 

[jmcdermo]

Oguz,
Thank you for the reply. Fortunately, the damaged SSD was recovered by using ddrescue to create an image of the damaged SSD and then dd to copy the image to a new SSD. So, I was able to bring up the old Proxmox environment. As you pointed out, that leaves me with an old version of Proxmox. Now that I have the original environment running again I now need to learn how to work with Proxmox and one of my next tasks will be to create a disaster recover backup system and then upgrade to the latest version. Wish me luck.

 

[oguz]

Thank you for the reply. Fortunately, the damaged SSD was recovered by using ddrescue to create an image of the damaged SSD and then dd to copy the image to a new SSD. So, I was able to bring up the old Proxmox environment.

ah, great!

one of my next tasks will be to create a disaster recover backup system and then upgrade to the latest version. Wish me luck.

you can take hints from my previous post about things you should backup in that case!