Since this issue isn’t really related to FOG, I’ll mark it as solved.

@Tom-Elliott I guess I have to use one of the physical machines as my master then until the new server arrives.

@Tom-Elliott

Okay, it’s a hardware problem. It’s working fine on a proxmox server with a Xeon L5640 but not on one with a L5420.
No clue what do do now, since the working one is in use for something different.

@Tom-Elliott

The guide i followed was wrong, the files are called efi-XXX.rom, not pxe-XXX.rom.

I updated iPXE and, after that failed, also tried using the updated roms and seabios from qemu’s git repo.
With both attempts however, it get’s stuck at a way earlier stage.

Is there a way to avoid the more recent iPXE versions?

@Tom-Elliott @george1421

Even tried it on different hardware and with the latest proxmox update now, still no luck.
Any ideas how to get it working?

Thanks for your help, much appreciated

I followed this https://gist.github.com/papamoose/e0b18d002f8686aaf767 and replaced all three roms, however nothing (not even the version number) has changed, still stuck - maybe they were already up to date?

For the questions:

1. Legacy
2. 66 - 192.168.162.55 (FOGs IP), 67 - undionly.kpxe
3. I followed the CentOS 7 guide from the Wiki, so I’m using the Git trunk.

Hey,

I’ve just set up my FOG server using CentOS 7 on ProxMox KVM and tried PXE booting another VM on the same host, connected via bridge.

However no matter what I do (trying different virtual network cards: e1000e, virtio, realtek, CPUs, …), the VM always get’s stuck at the same stage in PXE boot, Proxmox/KVM just shows the VM’s state as “internal-error”.

This is where it get’s stuck:

Edit:
If I use pxelinux.0 it shows some boot menu for about half a second and then only shows a black screen with an underscore at the top and get’s stuck in the same internal-error state.