[SOLVED] Issues converting VirtualBox VM to Proxmox

scilent

New Member
Aug 16, 2023
3
0
1
Hi everyone,

I'm facing a challenge with running a VirtualBox VM in Proxmox, and I'm hoping to get some guidance.
I've gone through related threads and followed the guide on pve.proxmox.com/wiki, but unfortunately, I haven't been able to find any answers to my problem.

To give you some context, I initially had a .vmdk file exported from VirtualBox. I converted it to qcow2 file using the following command:

Code:
$ qemu-img convert -f vmdk -O qcow2 /home/newVM/PG2_1.vmdk /home/newVM/PG3_1.qcow2

The conversion was successful. Next, I created a new VM using the "Create VM" button. I removed the Hard Disk and then imported the .qcow2 file using the command:

Code:
$ qm importdisk 100 /home/newVM/PG3_1.qcow2 scadaLV-dir

I encountered no issues, and I can see the disk listed in the scadaLV-dir storage under the 'VM Disks' tab.
A detached disk also appeared in the Hardware section of the VM, so I went ahead and attached it.
Here's what the VM configuration looks like at the moment:

Code:
$ qm config 100
bios: seabio
boot: order=ide2;ide0;net0
cores: 2
ide0: scadaLV-dir:100/vm-100-disk-0.raw,size=500G
ide2: none,media=cdrom
machine: pc-i440fx-7.2
memory: 4096
meta: creation-eqmu=7.2.0,ctime=1691748860
name: testPG3
net0: e1000=9E:EE:9A:3F:BD:5C,bridge=vmbr1,firewall=1
numa: 0
ostype: win7
scsihw: virtio-scsi-pci
smbios1: uuid=aa0fabc7-052f-4594-8eb0-77c760c6e246
sockets: 1
vmgenid: b7d6a1ad-204c-4a9f-b099-83296588d59e

However, when starting the VM, something isn't right - the RAM memory is extremely low at only 0.35%, while the CPU fluctuates from 0% to 50%.
Furthermore, noVNC returns the error message "failed to connect to the server."

I tried setting bios to OVMF, with the only difference being a slightly higher RAM usage at 7% and CPU running at 70%.

Any suggestions would be greatly appreciated.
 
Last edited:
ide0: scadaLV-dir:100/vm-100-disk-0.raw,size=500G
The IDE Bus is slow, and it may cause the performance issue. I would change it to SCSI instead.

EDIT: to do that, you can go to the Hardware and first detach it and then add the detached disk as SCSI.
 
The IDE Bus is slow, and it may cause the performance issue. I would change it to SCSI instead.

EDIT: to do that, you can go to the Hardware and first detach it and then add the detached disk as SCSI.

Thank you for your response. I followed your suggestion and re-attached the disk using the SCSI bus. This led in a change of memory usage - now over 80%.
However, noVNC still remains unresponsive as VE does not boot.

I'm starting to suspect that the issue might originate from the source image.
I re-tried the disk import with a slightly modified command: I added the optional --format parameter, without luck:
Code:
qm importdisk 100 /home/newVM/PG3_1.qcow2 scadaLV-dir --format qcow2

I have two VM images so it's unlikely that both are corrupted, but I'm going to run the same process with a new image to see if corruption is the root cause.
 
Hi Moayad,

Thanks for your reply.

The ultimate solution is the code below that I need to run every time I add a new VM:

Code:
pvecm updatecerts -F

Code:
systemctl restart pvedaemon pveproxy

Solved.