LXD
install
apt remove lxd lxd-client snap install lxd # create zfs dataset on pool rpool zfs create rpool/lxd # create lxd storage called zfspool using previous defined dataset lxc storage create zfspool zfs source=rpool/lxd # define default storage pool lxc profile device add default root disk path=/ pool=zfspool # initialize network sudo lxd init
Because group membership is only applied at login, you then either need to close and re-open your user session or use the “newgrp lxd” command in the shell you're going to interact with lxd from
newgrp lxd
- lxc-prepare (chmod +x)
#!/bin/bash NAME=$1 ALIAS=$2 ALIAS=${ALIAS:=xenial} lxc image show $ALIAS >/dev/null 2>&1 if [ ! $? = 0 ]; then echo lxc image copy images:ubuntu/xenial/amd64 local: --alias xenial exit 0 fi if [ ! -f /etc/apt/apt.conf.d/proxy.conf ]; then sudo apt install apt-cacher-ng PROXY=$( lxc network show lxdbr0 | sed -n 's/\s\+ipv4.address: \([0-9\.]\+\).*/\1/p' ) echo "Acquire::http::Proxy \"http://$PROXY:3142\";" | sudo tee /etc/apt/apt.conf.d/proxy.conf echo "PfilePattern = .*" | sudo tee -a /etc/apt-cacher-ng/acng.conf echo "PassThroughPattern: .*" | sudo tee -a /etc/apt-cacher-ng/acng.conf systemctl restart apt-cacher-ng fi lxc info $NAME >/dev/null 2>&1 if [ ! $? = 0 ]; then lxc launch $ALIAS $NAME fi if [ -f /etc/apt/apt.conf.d/proxy.conf ]; then lxc file push /etc/apt/apt.conf.d/proxy.conf $NAME/etc/apt/apt.conf.d/ fi lxc file push /etc/inputrc $NAME/etc/
basic
list remote images
lxc image list images:
auto update remote images
lxc config set images.auto_update_cached true
import image
lxc image copy images:ubuntu/xenial/amd64 local: --alias xenial
create profile
lxc profile create juju-default cat profile.yaml | lxc profile edit juju-default
profile.yaml
name: juju-default config: boot.autostart: "true" security.nesting: "true" security.privileged: "true" linux.kernel_modules: openvswitch,nbd,ip_tables,ip6_tables devices: eth0: mtu: "9000" name: eth0 nictype: bridged parent: br-mng type: nic kvm: path: /dev/kvm type: unix-char mem: path: /dev/mem type: unix-char root: path: / type: disk tun: path: /dev/net/tun type: unix-char
create container from local image
lxc image list lxc launch xenial test1 --profile juju-default
create container from remote image
lxc launch images:ubuntu/xenial/amd64 xenial1 lxc config set xenial1 boot.autostart false lxc list
create custom image from local container
lxc publish local-container --alias mycustomimage
create container from previous image
lxc launch mycustomimage newcontainer
bash inside
lxc exec trusty1 -- /bin/bash
stop and delete
lxc stop trusty1 lxc delete trusty1
autostart on host boot
lxc config set <name> boot.autostart true
show container configuration
lxc config show <name>
proxy
apt install apt-cacher-ng NAME=x11test lxc file push /etc/apt/apt.conf.d/proxy.conf $NAME/etc/apt/apt.conf.d/
- /etc/apt/apt.conf.d/proxy
Acquire::http::Proxy "http://10.106.191.1:3142";
network
lxc network create br0 lxc network show br0 lxc network edit br0
static IP container
istance=c1 lxc stop $instance lxc network attach lxdbr0 $istance eth0 eth0 lxc config device set $istance eth0 ipv4.address 10.99.10.42 lxc start $istance
servers
prepare lxd server
# bind to port 8443 lxc config set core.https_address "[::]" # password lxc config set core.trust_password some-password
from client add remote server
lxc remote add myserver <ip address or DNS>
run command
lxc exec myserver:trusty1 -- bash
xorg integration
container
create container
NAME=x11test lxc launch images:ubuntu/bionic/amd64 $NAME
install simpler X program
lxc exec $NAME -- apt install xterm lxc exec $NAME bash apt install mesa-utils x11-apps
NAME=nvidia-sdk-manager
# lxc config set $NAME environment.DISPLAY <ip-of-host-lxdbr0-bridge>:0
lxc config set $NAME environment.DISPLAY :0
lxc config device add $NAME X0 disk path=/tmp/.X11-unix/X0 source=/tmp/.X11-unix/X0
lxc config device add $NAME Xauthority disk path=/root/.Xauthority source=${XAUTHORITY}
on host
for gmd (ubuntu >= 17.10) or …
- /etc/gdm3/custom.conf
[security] DisallowTCP=false [xdmcp] Enable=true
… or for lightdm
- /etc/lightdm/lightdm.conf
xserver-allow-tcp=true xserver-command=X -listen tcp
add ip of container on /etc/X0.hosts
NAME=x11test lxc info $NAME | sed -n "s/\s*eth0:\s*inet\s\([0-9\.]*\).*/\1/p" >> /etc/X0.hosts
launch X application in container
xhost + lxc exec $NAME -- xterm
audio integration
misc devices
lxc config device add <name> rfxcom unix-char path=/dev/ttyACM0 lxc config device set <name> rfxcom mode 666
share folder
# only first time
echo "root:$UID:1" | sudo tee -a /etc/subuid
echo "root:${id -d}:1" | sudo tee -a /etc/subgid
lxc profile set default security.privileged true
# for every share
# lxc init stretch giano
lxc config set gianocop security.privileged true
lxc config set giano raw.idmap "both $UID $UID"
# source is on host, path is inside container
lxc config device add giano develop disk source=/mnt/giano path=/mnt/giano
migration
on host-destination
lxc config set core.https_address 0.0.0.0:8443 lxc config set core.trust_password PASSWORDhere
on host-origin
# add destination lxd lxc remote add other-server <ip-address> # take snap0 on gianocop container lxc snapshot gianocop snap0 lxc copy gianocop/snap0 other-server:gianocop --verbose lxc delete gianocop/snap0
on host-destination delete volatile in “lxc config”
volatile.base_image: 6adc9ca1a1124ebd954ba787e83dd9318866fd0b9ddce1cffc612559cfe3bc88 volatile.eth0.hwaddr: 00:16:3e:50:f6:e8 volatile.eth0.name: eth0 volatile.idmap.base: "0" volatile.idmap.next: '[{"Isuid":true,"Isgid":false,"Hostid":165536,"Nsid":0,"Maprange":1000},{"Isuid":true,"Isgid":true,"Hostid":1000,"Nsid":1000,"Maprange":1},{"Isuid":true,"Isgid":false,"Hostid":166537,"Nsid":1001,"Maprange":64535},{"Isuid":false,"Isgid":true,"Hostid":165536,"Nsid":0,"Maprange":1000},{"Isuid":true,"Isgid":true,"Hostid":1000,"Nsid":1000,"Maprange":1},{"Isuid":false,"Isgid":true,"Hostid":166537,"Nsid":1001,"Maprange":64535}]' volatile.last_state.idmap: '[{"Isuid":true,"Isgid":false,"Hostid":165536,"Nsid":0,"Maprange":1000},{"Isuid":true,"Isgid":true,"Hostid":1000,"Nsid":1000,"Maprange":1},{"Isuid":true,"Isgid":false,"Hostid":166537,"Nsid":1001,"Maprange":64535},{"Isuid":false,"Isgid":true,"Hostid":165536,"Nsid":0,"Maprange":1000},{"Isuid":true,"Isgid":true,"Hostid":1000,"Nsid":1000,"Maprange":1},{"Isuid":false,"Isgid":true,"Hostid":166537,"Nsid":1001,"Maprange":64535}]' volatile.last_state.power: STOPPED
export image from container
Vlan attach
apt-get install vlan
sudo modprobe 8021q
sudo vconfig add eth1 10
sudo ip addr add 10.0.0.1/24 dev eth1.10
ip addr del 10.22.30.44/16 dev eth0
sudo ip link set up eth1.10
sudo su -c 'echo "8021q" >> /etc/modules'
auto eth1.10 iface eth1.10 inet static address 10.0.0.1 netmask 255.255.255.0 vlan-raw-device eth1
Send file to your new host
On image hosts
lxc publish --force 'name of container" --alias 'new name'
example
lxc publish --force 'lxc-limesurvey' --alias 'lxc-docuwiki'
Export image
lxc image export 'new name'
Output is in efaa243331f0a7c175376edaf796545a01ad09bb47f25a297b798e09fe66ee66.tar.gz Show size of export
du -h efaa243331f0a7c175376edaf796545a01ad09bb47f25a297b798e09fe66ee66.tar.gz
check sum of image
md5sum efaa243331f0a7c175376edaf796545a01ad09bb47f25a297b798e09fe66ee66.tar.gz > exportmd5.txt cat exportmd5.txt | nc 10.18.49.73 1234 cat efaa243331f0a7c175376edaf796545a01ad09bb47f25a297b798e09fe66ee66.tar.gz | nc 10.18.49.73 1234
NB: 10.18.49.73 is your new lxd host
1234 is a free port
Transfer image and checksum to new LXD host
nc -l 1234 > efaa243331f0a7c175376edaf796545a01ad09bb47f25a297b798e09fe66ee66.tar.gz nc -l 1234 > exportmd5.txt
check file
md5sum efaa243331f0a7c175376edaf796545a01ad09bb47f25a297b798e09fe66ee66.tar.gz md5sum -c exportmd5.txt
Import image to new LXD host
lxc image import efaa243331f0a7c175376edaf796545a01ad09bb47f25a297b798e09fe66ee66.tar.gz --alias lxc-docuwiki
Transferring image: 100%
lxc launch image_name container_name
Creating container_name Starting container_name
In some instances the publish command may lead to a split xz tar-ball — but both formats are supported. Simply import the meta-data and rootfs components with
lxc image import <metadata tarball> <rootfs tarball> --alias image_name
Edit LXD default profile: networking
Put lxc network interface to host network
lxc stop lxc-docuwiki lxc profile device set default eth0 parent ens3 lxc profile device set default eth0 nictype macvlan service lxd restart service lxd-containers restart
launch your container
lxc start lxc-docuwiki lxc exec lxc-docuwiki /bin/bash