We use KVM with libvirt on a scientific instrument designed to supports experiment-driven research in all areas of computer science related to large-scale parallel or distributed systems.
Our users do not have root access on the production environment. A script based on "sudo" is available to create tap devices (tap0...tapn). So they can run their guest VMs with "kvm" or "virsh" commands using the defined tap devices and a defined range of MAC adresses (we have an ACL for DHCP requests from all VMs).
Now, we want to allow the users to use "virt-install" with a tap device already created.
So we want to evaluate if adding that option to "virt-install" is difficult or not?
Thanks for your help,
-- Tina Rakotoarivelo