I’m working on a new project, and I am using Multipass on an Ubuntu machine to provision some virtual machines on my local machine using cloudinit files. All good so far!
I wanted to expose one of the services I’ve created to the bridged network (so I can run avahi-daemon), and did this by running multipass launch -n vm01 --network enp3s0 when, what should I see but: launch failed: The bridging feature is not implemented on this backend. OH NO!
By chance, I found a random Stack Overflow answer, which said:
Currently only the LXD driver supports the networks command on Linux.
So, let’s make multipass on Ubuntu use LXD! (Be prepared for entering your password a few times!)
Firstly, we need to install LXD. Dead simple:
snap install lxdNext, we need to tell snap that it’s allowed to connect LXD to multipass:
snap connect multipass:lxd lxdAnd lastly, we tell multipass to use lxd:
multipass set local.driver=lxdResult?
user@host:~$ multipass networks
Name Type Description
enp3s0 ethernet Ethernet device
mpbr0 bridge Network bridge for MultipassAnd when I brought my machine up with avahi-daemon installed and configured to broadcast it’s hostname?
user@host:~$ ip -4 addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
37: br-enp3s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
inet 192.0.2.33/24 brd 192.0.2.255 scope global dynamic noprefixroute br-enp3s0
valid_lft 6455sec preferred_lft 6455sec
user@host:~$ multipass list
Name State IPv4 Image
vm01 Running 203.0.113.15 Ubuntu 22.04 LTS
192.0.2.101
user@host:~$ ping vm01.local
PING vm01.local (192.0.2.101) 56(84) bytes of data.Tada!
Featured image is “Apoptosis Network (alternate)” by “Simon Cockell” on Flickr and is released under a CC-BY license.
Hello! Did the same. At first it worked. But after that, the IP address disappeared from my main interfaces.
And now when I build a new instance, it does not want to receive an IP address by dhcp.
enp2s0 UP
enp3s0 UP
br-enp3s0 UP 10.24.64.27/24 fe80::bfc:7cba:9742:838e/64
br-enp2s0 UP 192.168.50.68/24 fe80::accd:f15c:e206:1206/64
Hmm, this is strange. I must confess, I’ve stopped using LXD and Multipass, and have gone back to using Vagrant and Virtualbox.
However, it might be worth running tshark, tcpdump, or something similar against the bridged interface, and then try running `dhclient` on that interface while you’re running the packet capture.
It might reveal whether the DHCP request is getting a response. If you are, and you get an IP address from that, then you need to look at something inside the VM – does `journalctl -xe` show you anything of use? Does running `netplan apply` do anything useful?
When you say that when your build a new instance it doesn’t get an IP address, do you mean you have two VM instances?
Also, as a random question, are you bridged to a WiFi interface or a wired one… As this might make a difference too.
There are lots of options here. Let me know what you try and how it gets on!