VM with libvirt

Requirements:

apt-get install qemu qemu-kvm libvirt-clients libvirt-daemon-system virtinst bridge-utils

And vmx compatibility: grep -E --color 'vmx|svm' /proc/cpuinfo

Build an image

with Virt-builder

Debian package: libguestfs-tools

virt-builder debian-11 -o test.img --size 100G --root-password password:XXX --hostname test.parmentier.io --install "net-tools" --append-line '/etc/resolv.conf:nameserver 213.186.33.99'

Notes:

• you can use --firstboot-command 'apt-get install -y ...'
• --append-line '/etc/hosts:192.168.225.1 server.ostechnix.local'

source

Install/Import an image

Debian package: virtinst

virt-install \
--name test.parmentier.io \
--ram 4096 \
--disk path=/var/lib/libvirt/images/test.img \
--vcpus 2 \
--os-variant debian10 \
--network bridge=br0,mac=xxx \
--graphics none \
--noautoconsole \
--boot hd \
--import 

Notes:

• there is no --os-variant debian11 yet, but debian10 should be fine?
• if you use a bridge network, ensure it is configured on host, and you might have to specify a virutal mac address (its the case for ovh vm within dedicated).
• If you are trying to install an iso from scratch, you will certainly need --cdrom=/path/to/file.iso and --graphics vnc see example

source

Start vm

virsh --connect qemu:///system start test.parmentier.io

Auto start

virsh autostart test.parmentier.io

Get into vm

virsh console test.parmentier.io

Edit configuration

virsh edit test.parmentier.io

Uninstall

virsh shutdown test.parmentier.io
virsh undefine test.parmentier.io

Reboot safely

virsh reboot test.parmentier.io

Hard reboot

virsh reset test.parmentier.io

Other useful commands

• suspend: froze on memoryy
• shutdown
• destroy: do uninstall first

Network (host)

On debian 11, ensure apt-get install -y bridge-utils:

iface eno3 inet manual

auto br0
iface br0 inet static
    address xx/24
    gateway yyy
    bridge-ports eno3
    bridge-stp off
    bridge-fd 0
    hwaddress 0a:xx:00

Note: on debian 11, without mac address, it did not work.

Also, ensure you have ip_forward=1, you cna check value with sysctl net.ipv4.ip_forward, and edit with sysctl net.ipv4.ip_forward=1. To make it persistant: edit /etc/sysctl.conf, find the line ip_forward=, then run sysctl -p /etc/sysctl.conf.

• source 1
• source2
• source3

Network (vm)

auto enp1s0
iface enp1s0 inet static
        address xx
        netmask 255.255.255.255
        broadcast xx
        post-up ip route add yy/32 dev enp1s0
        post-up ip route add default via yy
        pre-down ip route del yy/32 dev enp1s0
        pre-down ip route del default via yy

Note: replace xx and yy