LXD

From ArchWiki

This article or section needs expansion.

Reason: Possibility to improve introduction by quoting upstream? (Discuss in Talk:LXD)

LXD is a manager/hypervisor by Canonical for containers (via LXC) and virtual-machines (via QEMU).

Installation

Install the lxd package, then enable the lxd.socket.

Alternatively, you can enable/start the lxd.service directly, in case you want instances to autostart for example.

Configuration

Unprivileged containers

It is recommended to use unprivileged containers (See Linux Containers#Privileged containers or unprivileged containers for an explanation of the difference).

For this, modify both /etc/subuid and /etc/subgid (if these files are not present, create them) to contain the mapping to the containerized uid/gid pairs for each user who shall be able to run the containers. The example below is simply for the root user (and systemd system unit):

You can either use usermod as follows:

usermod -v 1000000-1000999999 -w 1000000-1000999999 root

Or modify the above mentioned files directly as follows:

/etc/subuid
root:1000000:1000000000
/etc/subgid
root:1000000:1000000000

Now, every container will be started unprivileged by default.

For the alternative, see howto set up privileged containers.

Configure LXD

On the first start, LXD needs to be configured.

Run as root:

# lxd init

This will start an interactive configuration guide in the terminal, that covers different topics like storages, networks etc.
You can find an overview in the official Getting Started Guide.

Accessing LXD as an unprivileged user

By default, the LXD daemon allows users in the lxd group access, so add your user to the group:

# usermod -a -G lxd username
Note: Do not forget to relogin to apply user's group membership. You can show your effective groups using id command.
Warning: Anyone added to the lxd group is root equivalent. For more information, see [1] and [2].

Usage

LXD consists of two parts:

  • the daemon (the lxd binary)
  • the client (the lxc binary)
Note: LXD is not LXC; the naming is a bit confusing, you can read the forum post on comparing LXD vs LXC regarding the difference.

The client is used to control one or multiple daemon(s).

The client can also be used to control remote LXD servers.

Overview of commands

You can get an overview of all available commands by typing:

$ lxc

Create a container

You can create a container with lxc launch, for example:

$ lxc launch ubuntu:20.04

Container are based on images, that are downloaded from image servers or remote LXD servers.

You can see the list of already added servers with:

$ lxc remote list

You can list all images on a server with lxc image list, for example:

$ lxc image list images:

This will show you all images on one of the default servers: images.linuxcontainers.org

You can also search for images by adding terms like the distribution name:

$ lxc image list images:debian

Launch a container with an image from a specific server with:

$ lxc launch servername:imagename

For example:

$ lxc launch images:centos/8/amd64 centos

To create an amd64 Arch container:

$ lxc launch images:archlinux/current/amd64 arch

Create a virtual machine

Just add the flag --vm to lxc launch:

$ lxc launch ubuntu:20.04 --vm
Note:
  • For now, virtual machines support less features than containers (see Details on virtual machines for example).
  • Only cloud variants of the official images enable the lxd-agent out-of-the-box (which is needed for the usual lxc commands like lxc exec).
    You can search for cloud images with lxc image list images: cloud or lxc image list images: distribution-name cloud.
    If you use other images or encounter problems, take a look at #lxd-agent inside a virtual machine.

Use and manage a container or VM

See "Manage instances" in the official Getting Started Guide of LXD.

Container/VM configuration

You can add various options to instances (containers and VMs).
See Configuration of instances in the official Advanced Guide of LXD for details.

Tips and tricks

Access the containers by name on the host

This assumes that you are using the default bridge that it is named lxdbr0 and that you are using systemd-resolved.

# systemd-resolve --interface lxdbr0 --set-domain '~lxd' --set-dns $(lxc network get lxdbr0 ipv4.address | cut -d / -f 1)

You can now access the containers by name:

$ ping containername.lxd

Other solution

The factual accuracy of this article or section is disputed.

Reason: When does systemd-resolve stop working? Is there a bug report or some other reference? (Discuss in Talk:LXD)

It seems that the systemd-resolve solution stops working after some time.

Another solution is to use systemd-networkd with the following lxd.network (replace x and y to match your bridge IP):

/etc/systemd/network/lxd.network
[Match]
Name=lxdbr0

[Network]
DNS=10.x.y.1
Domains=~lxd
IgnoreCarrierLoss=yes

[Address]
Address=10.x.y.1/24
Gateway=10.x.y.1

Use Wayland and Xorg applications

Note: Always consider security implications, as some of the described methods may weaken the seperation between container and host.

There are multiple methods to use GUI applications inside containers, you can find an overview in the official LXD forum.

The following method grants containers access to the host's sockets of Wayland (+Xwayland) or Xorg.

Note: Using Xorg might weaken the seperation between container and host, because Xorg allows applications to access other applications windows. So container applications might have access to host applications windows. Use Wayland instead (but be aware that Xorg downsides also apply to Xwayland).

Add the following devices to a containers profile

See also LXD documentation regarding devices.

General device for the GPU:

mygpu:
   type: gpu
Note: The path under "listen" is different, because /run and /tmp directories might be overridden, see https://github.com/lxc/lxd/issues/4540[dead link 2023-09-16 ⓘ].

Device for the Wayland socket:

Note:
  • Adjust the Display (wayland-0) accordingly.
  • Add the directories in /mnt and /tmp inside the container, if they do not already exist.
Waylandsocket:
    bind: container
    connect: unix:/run/user/1000/wayland-0
    listen: unix:/mnt/wayland1/wayland-0
    uid: "1000"
    gid: "1000"
    security.gid: "1000"
    security.uid: "1000"
    mode: "0777"
    type: proxy

Device for the Xorg (or Xwayland) Socket:

Note: Adjust the Display Number accordingly (for example X1 instead of X0).
Xsocket:
    bind: container
    connect: unix:/tmp/.X11-unix/X0
    listen: unix:/mnt/xorg1/X0
    uid: "1000"
    gid: "1000"
    security.gid: "1000"
    security.uid: "1000"
    mode: "0777"
    type: proxy

Link the sockets to the right location inside the container

Note: These scripts need to be run after each start of the container; you can automate this with systemd for example.

Shell script to link the Wayland socket:

#!/bin/sh
mkdir /run/user/1000
ln -s /mnt/wayland1/wayland-0 /run/user/1000/wayland-0

Link the Xorg (or Xwayland) socket:

#!/bin/sh
ln -s /mnt/xorg1/X0 /tmp/.X11-unix/X0

Add environment variables to the users config inside the container

Note: Adjust the Display Numbers and/or the filename (.profile) accordingly.

For Wayland:

$ echo "export XDG_RUNTIME_DIR=/run/user/1000" >> ~/.profile
$ echo "export WAYLAND_DISPLAY=wayland-0" >> ~/.profile
$ echo "export QT_QPA_PLATFORM=wayland" >> ~/.profile

For Xorg (or Xwayland):

$ echo "export DISPLAY=:0" >> ~/.profile

Reload the ~/.profile:

$ source ~/.profile

Install necessary software in the container

Necessary software needs to be added. For now, you can install an example GUI application; this will probably install all necessary packages as well.

Start GUI applications

Now, you should be able to start GUI applications inside the container (via terminal for example) and make them appear as a window on your hosts display.

You can try out glxgears for example.

Privileged containers

Note:
  • Privileged containers are not isolated from the host!
  • The root user in the container is the root user on the host.
  • Use unprivileged containers instead whenever possible.

If you want to set up a privileged container, you must provide the config key security.privileged=true.

Either during container creation:

$ lxc launch ubuntu:20.04 ubuntu -c security.privileged=true

Or for an already existing container, you may edit the configuration:

$ lxc config edit ubuntu
name: ubuntu
profiles:
- default
config:
  ...
  security.privileged: "true"
  ...

Add a disk device

Read-Only

If you want to share a disk device from the host to a container, all you need to do is add a disk device to your container. The virtual disk device needs a name (only used internally in the LXC configuration file), a path on the host's filesystem pointing to the disk you want to mount, as well as a desired mountpoint on the container's filesystem.

$ lxc config device add containername virtualdiskname disk source=/path/to/host/disk/ path=/path/to/mountpoint/on/container

Read-Write (unprivileged container)

The preferred method for read/write access is to use the "shift" method included in LXD.

shift is based on Linux kernel functionality and available in two different versions:

  • the most recent version is called "idmapped mounts" and is included in all upstream kernels >5.12 by default. So it is also included in the regular Arch Linux kernel (linux).
  • the old version is called "shiftfs" and needs to be added manually to most kernels as a kernel module. It is available as a legacy version to support older kernels. You can take a look at this GitHub repo that uses the shiftfs kernel module from Ubuntu kernels: https://github.com/toby63/shiftfs-dkms

Shift should be available and activated by default on Arch with the regular Arch Linux kernel (linux) and the lxd package.

1. To check whether shift is available on your system, run lxc info

The first part of the output shows you:

 kernel_features:
    idmapped_mounts: "true"
    shiftfs: "false"

If either idmapped_mounts or shiftfs is true, then your kernel includes it already and you can use shift. If it is not true, you should check your kernel version and might try the "shiftfs" legacy version mentioned above.

The second part of the output shows you either:

  lxc_features:
    idmapped_mounts_v2: "true"

or:

  lxc_features:
    shiftfs: "true"

If either idmapped_mounts or shiftfs is true, then LXD has already enabled it. If it is not enabled, you must enable it first.

2. Usage

Then you can simply set the "shift" config key to "true" in the disk device options. See: LXD Documentation on disk devices

See also: tutorial in the LXD forums

Bash completion doesn't work

This workaround may fix the issue:

# ln -s /usr/share/bash-completion/completions/lxd /usr/share/bash-completion/completions/lxc

Troubleshooting

lxd-agent inside a virtual machine

Inside some virtual machine images, the lxd-agent is not enabled by default.

In this case, you have to enable it manually, for example by mounting a 9p network share. This requires console access with a valid user.

1. Login with lxc console and replace virtualmachine-name accordingly.

$ lxc console virtualmachine-name

Login as root:

Note: On some systems, you have to setup a root password first to be able to login as root, you can use cloud-init for this for example.
$ su root

Mount the network share:

$ mount -t 9p config /mnt/

Go into the folder and run the install script (this will enable the lxd-agent inside the VM):

$ cd /mnt/
$ ./install.sh 

After a successful install, reboot with:

$ reboot

Afterwards, the lxd-agent is available and lxc exec should work.

Check kernel config

By default, the Arch Linux kernel is compiled correctly for Linux Containers and its frontend LXD. However, if you are using a custom kernel or changed the kernel options, the kernel might be configured incorrectly. Verify that your kernel is properly configured:

$ lxc-checkconfig

Resource limits are not applied when viewed from inside a container

Install lxcfs and start lxcfs.service.

lxd will need to be restarted. Enable lxcfs.service for the service to be started at boot time.

Starting a virtual machine fails

If you see the error:

Error: Couldn't find one of the required UEFI firmware files: [{code:OVMF_CODE.4MB.fd vars:OVMF_VARS.4MB.ms.fd} {code:OVMF_CODE.2MB.fd vars:OVMF_VARS.2MB.ms.fd} {code:OVMF_CODE.fd vars:OVMF_VARS.ms.fd} {code:OVMF_CODE.fd vars:qemu.nvram}]

It's because Arch Linux does not distribute secure boot signed ovmf firmware. To boot virtual machines you need to disable secure boot for the time being:

$ lxc launch ubuntu:18.04 test-vm --vm -c security.secureboot=false

This can also be added to the default profile by doing:

$ lxc profile set default security.secureboot=false

No IPv4 with systemd-networkd

Starting with version version 244.1, systemd detects if /sys is writable by containers. If it is, udev is automatically started and breaks IPv4 in unprivileged containers. See commit bf331d8 and discussion on linuxcontainers.

On containers created past 2020, there should already be a systemd.networkd.service override to work around this issue, create it if it is not:

/etc/systemd/system/systemd-networkd.service.d/lxc.conf
[Service]
BindReadOnlyPaths=/sys

You could also work around this issue by setting raw.lxc: lxc.mount.auto = proc:rw sys:ro in the profile of the container to ensure /sys is read-only for the entire container, although this may be problematic, as per the linked discussion above.

No networking with ufw

When running LXD on a system with ufw, the output of lxc ls will contain an empty IPv4 field, outbound requests will not be forwarded out of the container, and inbound requests will not be forwarded into the container. As seen in a thread on LXC's Discourse instance, ufw will block traffic from LXD bridges by default. The solution is to configure two new ufw rules for each bridge:

# ufw route allow in on lxdbr0
# ufw allow in on lxdbr0

For more information on these two commands, check out this thread which describes these commands and their limitations in more detail.

No networking with Docker installed

You have Docker installed on the host, and you're not able to access LAN or internet from within a lxc container.

# iptables -I DOCKER-USER -i lxdbr0 -o interface -j ACCEPT
# iptables -I DOCKER-USER -o lxdbr0 -m conntrack --ctstate RELATED,ESTABLISHED -j ACCEPT

On the first line, replace interface with the external network interface (what connects the host to LAN/internet, e.g. enp6so, wlp5s0, ...). Also replace lxdbr0 if needed.

For more details, see this note in the LXD documentation.

Building a snap with snapcraft: craft-providers error: LXD requires additional permissions.

If you get the following error:

craft-providers error: LXD requires additional permissions.
Ensure that the user is in the 'lxd' group.
Visit https://documentation.ubuntu.com/lxd/en/latest/getting_started/ for instructions on installing and configuring LXD for your operating system.
Full execution log: '/home/your_user/.local/state/snapcraft/log/snapcraft-20231129-221308.209638.log'

This might occur if you try to build a snap with the snapcraft command. If the error is still there even after you added yourself to the lxd group and rebooted, try installing lxd via snap instead of as an Arch package (don't install both), that might solve this problem.

Uninstall

Stop and disable lxd.service and lxd.socket. Then uninstall the lxd package.

If you uninstalled the package without disabling the service, you might have a lingering broken symlink at /etc/systemd/system/multi-user.wants/lxd.service.

If you want to remove all data:

# rm -r /var/lib/lxd

If you used any of the example networking configuration, you should remove those as well.

See also