I'm currently working around it by bind-mounting the upstream source over the Ubuntu package thus:
mkdir -p /home/all/SourceCode/vagrant cd /home/all/Sourcecode/vagrant git clone https://github.com/vagrant-libvirt/vagrant-libvirt.git sudo mount --bind /home/all/SourceCode/vagrant/vagrant-libvirt /usr/share/rubygems-integration/all/gems/vagrant-libvirt-0.0.43 After this I can use management_network_device in the Vagrantfile. E.g.: Vagrant.configure("2") do |config| config.vm.define :tactyle do |tactyle| tactyle.vm.box = "ubuntu/bionic" tactyle.vm.network :public_network, :dev => "virbr0", :mode => "bridge", :type => "bridge" # tactyle.vm.synced_folder "../data", "/vagrant_data" tactyle.ssh.insert_key = false tactyle.vm.hostname = "tactyle" tactyle.vm.provider :libvirt do |libvirt| # volatile memory in MegaBytes libvirt.memory = 1024 # virtual CPU qty libvirt.cpus = 4 # headless system (no video output) libvirt.graphics_type = "none" # UK keyboard libvirt.keymap = "en-gb" # VM architecture requires 64-bit libvirt.machine_arch = "x86_64" # pass the host's /dev/random device to the guest libvirt.random :model => "random" # set the management interface libvirt.management_network_device = 'virbr1' end Alas, although the VM now starts, it fails in the initramfs and drops to the shell due to the root fs block device missing! -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/1784229 Title: 18.04: VM fails to start due to management_network assuming non- existent interface eth0 To manage notifications about this bug go to: https://bugs.launchpad.net/vagrant-libvirt/+bug/1784229/+subscriptions -- ubuntu-bugs mailing list [email protected] https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
