Before you copy the KVM image you should remove the system-id field from the arango-database, and any uls identifiers.
Use virt-sparsify to copy the KVM image into the images directory, this can take a while, this prevents the image from taking up unnecessary space despite being thin provisioned.
sudo virt-sparsify /var/lib/libvirt/images/<kvm_name.qcow2> ./images/<kvm_name.qcow2>
N.B. The name of the original KVM should match the copy made by virt-sparsify
Once there use 'virsh dumpxml' to place the xml of the KVM into this directory
virsh dumpxml <vmname> > <vmname>.xml
You must also use virsh net-dumpxml default > ./network/default.xml to dump the network xml to a file.
Then if your system has a northbound and a southbound network add two lines as demonstrated below On a single NIC system just add one line
<host mac='<guest_NIC_mac>' name='<kvm_name>' ip='<guest_ip>'/>
Replace the variables in the above line with the information found in the kvm guest xml file. This defines the KVM to have a static IP address.
The final thing before copying this directory is to update the variables in the hook script. Hopefully it is self-explanatory.
Copy this directory to the new baremetal machine and run the .kvm_import.sh (recommend copying using,
rsync -rhPS --append-verify ./<this directory> ./<destination directory>
This should install the KVM on the new machine and restart the machine.
Next check that the KVM starts on boot, and also check that the Iptables forwarding is set.
virsh list --all, sudo iptables -L