salt.modules.lxc

Control Linux Containers via Salt

depends:lxc package for distribution

lxc >= 1.0 (even beta alpha) is required

salt.modules.lxc.apply_network_profile(name, network_profile, nic_opts=None)

New in version 2015.5.0.

Apply a network profile to a container

network_profile
profile name or default values (dict)
nic_opts
values to override in defaults (dict) indexed by nic card names

CLI Examples:

salt 'minion' lxc.apply_network_profile web1 centos
salt 'minion' lxc.apply_network_profile web1 centos \
        nic_opts="{'eth0': {'mac': 'xx:xx:xx:xx:xx:xx'}}"
salt 'minion' lxc.apply_network_profile web1 \
        "{'eth0': {'mac': 'xx:xx:xx:xx:xx:yy'}}"
        nic_opts="{'eth0': {'mac': 'xx:xx:xx:xx:xx:xx'}}"

The special case to disable use of ethernet nics:

salt 'minion' lxc.apply_network_profile web1 centos \
        "{eth0: {disable: true}}"
salt.modules.lxc.attachable(name)

Return True if the named container can be attached to via the lxc-attach command

CLI Example:

salt 'minion' lxc.attachable ubuntu
salt.modules.lxc.bootstrap(name, config=None, approve_key=True, install=True, pub_key=None, priv_key=None, bootstrap_url=None, force_install=False, unconditional_install=False, bootstrap_delay=None, bootstrap_args=None, bootstrap_shell=None)

Install and configure salt in a container.

config
Minion configuration options. By default, the master option is set to the target host's master.
approve_key
Request a pre-approval of the generated minion key. Requires that the salt-master be configured to either auto-accept all keys or expect a signing request from the target host. Default: True
pub_key
Explicit public key to pressed the minion with (optional). This can be either a filepath or a string representing the key
priv_key
Explicit private key to pressed the minion with (optional). This can be either a filepath or a string representing the key
bootstrap_delay

Delay in seconds between end of container creation and bootstrapping. Useful when waiting for container to obtain a DHCP lease.

New in version 2015.5.0.

bootstrap_url
url, content or filepath to the salt bootstrap script
bootstrap_args
salt bootstrap script arguments
bootstrap_shell
shell to execute the script into
install
Whether to attempt a full installation of salt-minion if needed.
force_install
Force installation even if salt-minion is detected, this is the way to run vendor bootstrap scripts even if a salt minion is already present in the container
unconditional_install
Run the script even if the container seems seeded

CLI Examples:

salt 'minion' lxc.bootstrap container_name [config=config_data] \
        [approve_key=(True|False)] [install=(True|False)]
salt.modules.lxc.clone(name, orig, profile=None, network_profile=None, nic_opts=None, **kwargs)

Create a new container as a clone of another container

name
Name of the container
orig
Name of the original container to be cloned
profile
Profile to use in container cloning (see lxc.get_container_profile). Values in a profile will be overridden by the Container Cloning Arguments listed below.

Container Cloning Arguments

snapshot
Use Copy On Write snapshots (LVM)
size : 1G
Size of the volume to create. Only applicable if backing=lvm.
backing
The type of storage to use. Set to lvm to use an LVM group. Defaults to filesystem within /var/lib/lxc.
network_profile

Network profile to use for container

New in version 2015.5.2.

nic_opts

give extra opts overriding network profile values

New in version 2015.5.2.

CLI Examples:

salt '*' lxc.clone myclone orig=orig_container
salt '*' lxc.clone myclone orig=orig_container snapshot=True
salt.modules.lxc.cloud_init(name, vm_=None, **kwargs)

Thin wrapper to lxc.init to be used from the saltcloud lxc driver

name
Name of the container may be None and then guessed from saltcloud mapping
vm_
saltcloud mapping defaults for the vm

CLI Example:

salt '*' lxc.cloud_init foo
salt.modules.lxc.cloud_init_interface(name, vm_=None, **kwargs)

Interface between salt.cloud.lxc driver and lxc.init vm_ is a mapping of vm opts in the salt.cloud format as documented for the lxc driver.

This can be used either:

  • from the salt cloud driver
  • because you find the argument to give easier here than using directly lxc.init

Warning

BE REALLY CAREFUL CHANGING DEFAULTS !!! IT'S A RETRO COMPATIBLE INTERFACE WITH THE SALT CLOUD DRIVER (ask kiorky).

name
name of the lxc container to create
pub_key
public key to preseed the minion with. Can be the keycontent or a filepath
priv_key
private key to preseed the minion with. Can be the keycontent or a filepath
profile
profile selection
network_profile
network profile selection
nic_opts

per interface settings compatibles with network profile (ipv4/ipv6/link/gateway/mac/netmask)

eg:

- {'eth0': {'mac': '00:16:3e:01:29:40',
            'gateway': None, (default)
            'link': 'br0', (default)
            'gateway': None, (default)
            'netmask': '', (default)
            'ip': '22.1.4.25'}}
unconditional_install
given to lxc.bootstrap (see relative doc)
force_install
given to lxc.bootstrap (see relative doc)
config
any extra argument for the salt minion config
dnsservers
dns servers to set inside the container
autostart
autostart the container at boot time
password
administrative password for the container

Warning

Legacy but still supported options:

from_container
which container we use as a template when running lxc.clone
image
which template do we use when we are using lxc.create. This is the default mode unless you specify something in from_container
backing
which backing store to use. Values can be: overlayfs, dir(default), lvm, zfs, brtfs
fstype
When using a blockdevice level backing store, which filesystem to use on
size
When using a blockdevice level backing store, which size for the filesystem to use on
snapshot
Use snapshot when cloning the container source
vgname
if using LVM: vgname
lvname
if using LVM: lvname
ip
ip for the primary nic
mac
mac address for the primary nic
netmask
netmask for the primary nic (24) = vm_.get('netmask', '24')
bridge
bridge for the primary nic (lxcbr0)
gateway
network gateway for the container
additional_ips

additional ips which will be wired on the main bridge (br0) which is connected to internet. Be aware that you may use manual virtual mac addresses providen by you provider (online, ovh, etc). This is a list of mappings {ip: '', mac: '', netmask:''} Set gateway to None and an interface with a gateway to escape from another interface that eth0. eg:

- {'mac': '00:16:3e:01:29:40',
   'gateway': None, (default)
   'link': 'br0', (default)
   'netmask': '', (default)
   'ip': '22.1.4.25'}
users
administrative users for the container default: [root] and [root, ubuntu] on ubuntu
default_nic
name of the first interface, you should really not override this

CLI Example:

salt '*' lxc.cloud_init_interface foo
salt.modules.lxc.cp(name, source, dest, makedirs=False)

Copy a file or directory from the host into a container.

name
Container name
source
File to be copied to the container
dest

Destination on the container. Must be an absolute path.

Changed in version 2015.5.0: If the destination is a directory, the file will be copied into that directory.

makedirs : False

Create the parent directory on the container if it does not already exist.

New in version 2015.5.0.

CLI Example:

salt 'minion' lxc.cp /tmp/foo /root/foo
salt.modules.lxc.create(name, config=None, profile=None, network_profile=None, nic_opts=None, **kwargs)

Create a new container.

name
Name of the container
config
The config file to use for the container. Defaults to system-wide config (usually in /etc/lxc/lxc.conf).
profile
Profile to use in container creation (see lxc.get_container_profile). Values in a profile will be overridden by the Container Creation Arguments listed below.
network_profile

Network profile to use for container

New in version 2015.5.0.

Container Creation Arguments

template

The template to use. For example, ubuntu or fedora. Conflicts with the image argument.

Note

The download template requires the following three parameters to be defined in options:

  • dist - The name of the distribution
  • release - Release name/version
  • arch - Architecture of the container

The available images can be listed using the lxc.images function.

options

Template-specific options to pass to the lxc-create command. These correspond to the long options (ones beginning with two dashes) that the template script accepts. For example:

options='{"dist": "centos", "release": "6", "arch": "amd64"}'
image
A tar archive to use as the rootfs for the container. Conflicts with the template argument.
backing
The type of storage to use. Set to lvm to use an LVM group. Defaults to filesystem within /var/lib/lxc.
fstype
Filesystem type to use on LVM logical volume
size : 1G
Size of the volume to create. Only applicable if backing=lvm.
vgname : lxc
Name of the LVM volume group in which to create the volume for this container. Only applicable if backing=lvm.
lvname
Name of the LVM logical volume in which to create the volume for this container. Only applicable if backing=lvm.
nic_opts
give extra opts overriding network profile values
salt.modules.lxc.destroy(name, stop=False)

Destroy the named container.

Warning

Destroys all data associated with the container.

stop : False

If True, the container will be destroyed even if it is running/frozen.

Changed in version 2015.5.0: Default value changed to False. This more closely matches the behavior of lxc-destroy(1), and also makes it less likely that an accidental command will destroy a running container that was being used for important things.

CLI Examples:

salt '*' lxc.destroy foo
salt '*' lxc.destroy foo stop=True
salt.modules.lxc.edit_conf(conf_file, out_format='simple', read_only=False, lxc_config=None, **kwargs)

Edit an LXC configuration file. If a setting is already present inside the file, its value will be replaced. If it does not exist, it will be appended to the end of the file. Comments and blank lines will be kept in-tact if they already exist in the file.

out_format:
Set to simple if you need backward compatbility (multiple items for a simple key is not supported)
read_only:
return only the edited configuration without applying it to the underlying lxc configuration file
lxc_config:
List of dict containning lxc configuration items For network configuration, you also need to add the device it belongs to, otherwise it will default to eth0. Also, any change to a network parameter will result in the whole network reconfiguration to avoid mismatchs, be aware of that !

After the file is edited, its contents will be returned. By default, it will be returned in simple format, meaning an unordered dict (which may not represent the actual file order). Passing in an out_format of commented will return a data structure which accurately represents the order and content of the file.

CLI Example:

salt 'minion' lxc.edit_conf /etc/lxc/mycontainer.conf \
    out_format=commented lxc.network.type=veth
salt 'minion' lxc.edit_conf /etc/lxc/mycontainer.conf \
    out_format=commented \
    lxc_config="[{'lxc.network.name': 'eth0', \
                  'lxc.network.ipv4': '1.2.3.4'},
                 {'lxc.network.name': 'eth2', \
                  'lxc.network.ipv4': '1.2.3.5',\
                  'lxc.network.gateway': '1.2.3.1'}]"
salt.modules.lxc.exists(name)

Returns whether the named container exists.

CLI Example:

salt '*' lxc.exists name
salt.modules.lxc.freeze(name, **kwargs)

Freeze the named container

start : False

If True and the container is stopped, the container will be started before attempting to freeze.

New in version 2015.5.0.

CLI Example:

salt '*' lxc.freeze name
salt.modules.lxc.get_container_profile(name=None, **kwargs)

New in version 2015.5.0.

Gather a pre-configured set of container configuration parameters. If no arguments are passed, an empty profile is returned.

Profiles can be defined in the minion or master config files, or in pillar or grains, and are loaded using config.get. The key under which LXC profiles must be configured is lxc.container_profile.profile_name. An example container profile would be as follows:

lxc.container_profile:
  ubuntu:
    template: ubuntu
    backing: lvm
    vgname: lxc
    size: 1G

Parameters set in a profile can be overridden by passing additional container creation arguments (such as the ones passed to lxc.create) to this function.

A profile can be defined either as the name of the profile, or a dictionary of variable names and values. See the LXC Tutorial for more information on how to use LXC profiles.

CLI Example:

.. code-block:: bash
salt-call lxc.get_container_profile centos salt-call lxc.get_container_profile ubuntu template=ubuntu backing=overlayfs
salt.modules.lxc.get_network_profile(name=None, **kwargs)

New in version 2015.5.0.

Gather a pre-configured set of network configuration parameters. If no arguments are passed, the following default profile is returned:

{'eth0': {'link': 'br0', 'type': 'veth', 'flags': 'up'}}

Profiles can be defined in the minion or master config files, or in pillar or grains, and are loaded using config.get. The key under which LXC profiles must be configured is lxc.network_profile. An example network profile would be as follows:

lxc.network_profile.centos:
  eth0:
    link: br0
    type: veth
    flags: up

To disable networking entirely:

lxc.network_profile.centos:
  eth0:
    disable: true

Parameters set in a profile can be overridden by passing additional arguments to this function.

A profile can be passed either as the name of the profile, or a dictionary of variable names and values. See the LXC Tutorial for more information on how to use network profiles.

Warning

The ipv4, ipv6, gateway, and link (bridge) settings in network profiles will only work if the container doesnt redefine the network configuration (for example in /etc/sysconfig/network-scripts/ifcfg-<interface_name> on RHEL/CentOS, or /etc/network/interfaces on Debian/Ubuntu/etc.)

CLI Example:

.. code-block:: bash
salt-call lxc.get_network_profile default
salt.modules.lxc.get_parameter(name, parameter)

Returns the value of a cgroup parameter for a container

CLI Example:

salt '*' lxc.get_parameter container_name memory.limit_in_bytes
salt.modules.lxc.images(dist=None)

New in version 2015.5.0.

List the available images for LXC's download template.

dist : None
Filter results to a single Linux distribution

CLI Examples:

salt myminion lxc.images
salt myminion lxc.images dist=centos
salt.modules.lxc.info(name)

Returns information about a container

CLI Example:

salt '*' lxc.info name
salt.modules.lxc.init(name, config=None, cpuset=None, cpushare=None, memory=None, profile=None, network_profile=None, nic=<object object>, nic_opts=None, cpu=None, autostart=True, password=None, password_encrypted=None, users=None, dnsservers=None, searchdomains=None, bridge=None, gateway=None, pub_key=None, priv_key=None, force_install=False, unconditional_install=False, bootstrap_delay=None, bootstrap_args=None, bootstrap_shell=None, bootstrap_url=None, **kwargs)

Initialize a new container.

This is a partial idempotent function as if it is already provisioned, we will reset a bit the lxc configuration file but much of the hard work will be escaped as markers will prevent re-execution of harmful tasks.

name
Name of the container
image
A tar archive to use as the rootfs for the container. Conflicts with the template argument.
cpus
Select a random number of cpu cores and assign it to the cpuset, if the cpuset option is set then this option will be ignored
cpuset
Explicitly define the cpus this container will be bound to
cpushare
cgroups cpu shares
autostart
autostart container on reboot
memory

cgroups memory limit, in MB

Changed in version 2015.5.0: If no value is passed, no limit is set. In earlier Salt versions, not passing this value causes a 1024MB memory limit to be set, and it was necessary to pass memory=0 to set no limit.

gateway
the ipv4 gateway to use the default does nothing more than lxcutils does
bridge
the bridge to use the default does nothing more than lxcutils does
network_profile

Network profile to use for the container

New in version 2015.5.0.

nic

Deprecated since version 2015.5.0: Use network_profile instead

nic_opts

Extra options for network interfaces, will override

{"eth0": {"hwaddr": "aa:bb:cc:dd:ee:ff", "ipv4": "10.1.1.1", "ipv6": "2001:db8::ff00:42:8329"}}

or

{"eth0": {"hwaddr": "aa:bb:cc:dd:ee:ff", "ipv4": "10.1.1.1/24", "ipv6": "2001:db8::ff00:42:8329"}}

users
Users for which the password defined in the password param should be set. Can be passed as a comma separated list or a python list. Defaults to just the root user.
password
Set the initial password for the users defined in the users parameter
password_encrypted : False

Set to True to denote a password hash instead of a plaintext password

New in version 2015.5.0.

profile
A LXC profile (defined in config or pillar). This can be either a real profile mapping or a string to retrieve it in configuration
start
Start the newly-created container
dnsservers
list of dns servers to set in the container, default [] (no setting)
seed
Seed the container with the minion config. Default: True
install
If salt-minion is not already installed, install it. Default: True
config
Optional config parameters. By default, the id is set to the name of the container.
master
salt master (default to minion's master)
master_port
salt master port (default to minion's master port)
pub_key
Explicit public key to preseed the minion with (optional). This can be either a filepath or a string representing the key
priv_key
Explicit private key to preseed the minion with (optional). This can be either a filepath or a string representing the key
approve_key
If explicit preseeding is not used; Attempt to request key approval from the master. Default: True
clone

Deprecated since version 2015.5.0: Use clone_from instead

clone_from
Original from which to use a clone operation to create the container. Default: None
bootstrap_delay

Delay in seconds between end of container creation and bootstrapping. Useful when waiting for container to obtain a DHCP lease.

New in version 2015.5.0.

bootstrap_url
See lxc.bootstrap
bootstrap_shell
See lxc.bootstrap
bootstrap_args
See lxc.bootstrap
force_install
Force installation even if salt-minion is detected, this is the way to run vendor bootstrap scripts even if a salt minion is already present in the container
unconditional_install
Run the script even if the container seems seeded

CLI Example:

salt 'minion' lxc.init name [cpuset=cgroups_cpuset] \
        [cpushare=cgroups_cpushare] [memory=cgroups_memory] \
        [nic=nic_profile] [profile=lxc_profile] \
        [nic_opts=nic_opts] [start=(True|False)] \
        [seed=(True|False)] [install=(True|False)] \
        [config=minion_config] [approve_key=(True|False) \
        [clone_from=original] [autostart=True] \
        [priv_key=/path_or_content] [pub_key=/path_or_content] \
        [bridge=lxcbr0] [gateway=10.0.3.1] \
        [dnsservers[dns1,dns2]] \
        [users=[foo]] [password='secret'] \
        [password_encrypted=(True|False)]
salt.modules.lxc.list_(extra=False, limit=None)

List containers classified by state

extra
Also get per-container specific info. This will change the return data. Instead of returning a list of containers, a dictionary of containers and each container's output from lxc.info.
limit

Return output matching a specific state (frozen, running, or stopped).

New in version 2015.5.0.

CLI Examples:

salt '*' lxc.list
salt '*' lxc.list extra=True
salt '*' lxc.list limit=running
salt.modules.lxc.ls_(active=None, cache=True)

Return a list of the containers available on the minion

active

If True, return only active (i.e. running) containers

New in version 2015.5.0.

CLI Example:

salt '*' lxc.ls
salt '*' lxc.ls active=True
salt.modules.lxc.read_conf(conf_file, out_format='simple')

Read in an LXC configuration file. By default returns a simple, unsorted dict, but can also return a more detailed structure including blank lines and comments.

out_format:
set to 'simple' if you need the old and unsupported behavior. This wont support the multiple lxc values (eg: multiple network nics)

CLI Examples:

salt 'minion' lxc.read_conf /etc/lxc/mycontainer.conf
salt 'minion' lxc.read_conf /etc/lxc/mycontainer.conf out_format=commented
salt.modules.lxc.restart(name, force=False)

New in version 2015.5.0.

Restart the named container. If the container was not running, the container will merely be started.

name
The name of the container
force : False
If True, the container will be force-stopped instead of gracefully shut down

CLI Example:

salt myminion lxc.restart name
salt.modules.lxc.retcode(name, cmd, no_start=False, preserve_state=True, stdin=None, python_shell=True, output_loglevel='debug', use_vt=False, ignore_retcode=False, chroot_fallback=False, keep_env='http_proxy, https_proxy, no_proxy')

New in version 2015.5.0.

Run cmd.retcode within a container

Warning

Many shell builtins do not work, failing with stderr similar to the following:

lxc_container: No such file or directory - failed to exec 'command'

The same error will be displayed in stderr if the command being run does not exist. If the retcode is nonzero and not what was expected, try using lxc.run_stderr or lxc.run_all.

name
Name of the container in which to run the command
cmd
Command to run
no_start : False
If the container is not running, don't start it
preserve_state : True
After running the command, return the container to its previous state
stdin : None
Standard input to be used for the command
output_loglevel : debug
Level at which to log the output from the command. Set to quiet to suppress logging.
use_vt : False
Use SaltStack's utils.vt to stream output to console output=all.
keep_env : http_proxy,https_proxy,no_proxy
A list of env vars to preserve. May be passed as commma-delimited list.
chroot_fallback
if the container is not running, try to run the command using chroot default: false

CLI Example:

salt myminion lxc.retcode mycontainer 'ip addr show'
salt.modules.lxc.run(name, cmd, no_start=False, preserve_state=True, stdin=None, python_shell=True, output_loglevel='debug', use_vt=False, ignore_retcode=False, chroot_fallback=False, keep_env='http_proxy, https_proxy, no_proxy')

New in version 2015.5.0.

Run cmd.run within a container

Warning

Many shell builtins do not work, failing with stderr similar to the following:

lxc_container: No such file or directory - failed to exec 'command'

The same error will be displayed in stderr if the command being run does not exist. If no output is returned using this function, try using lxc.run_stderr or lxc.run_all.

name
Name of the container in which to run the command
cmd
Command to run
no_start : False
If the container is not running, don't start it
preserve_state : True
After running the command, return the container to its previous state
stdin : None
Standard input to be used for the command
output_loglevel : debug
Level at which to log the output from the command. Set to quiet to suppress logging.
use_vt : False
Use SaltStack's utils.vt to stream output to console. Assumes output=all.
chroot_fallback
if the container is not running, try to run the command using chroot default: false
keep_env : http_proxy,https_proxy,no_proxy
A list of env vars to preserve. May be passed as commma-delimited list.

CLI Example:

salt myminion lxc.run mycontainer 'ifconfig -a'
salt.modules.lxc.run_all(name, cmd, no_start=False, preserve_state=True, stdin=None, python_shell=True, output_loglevel='debug', use_vt=False, ignore_retcode=False, chroot_fallback=False, keep_env='http_proxy, https_proxy, no_proxy')

New in version 2015.5.0.

Run cmd.run_all within a container

Warning

Many shell builtins do not work, failing with stderr similar to the following:

lxc_container: No such file or directory - failed to exec 'command'

The same error will be displayed in stderr if the command being run does not exist.

name
Name of the container in which to run the command
cmd
Command to run
no_start : False
If the container is not running, don't start it
preserve_state : True
After running the command, return the container to its previous state
stdin : None
Standard input to be used for the command
output_loglevel : debug
Level at which to log the output from the command. Set to quiet to suppress logging.
use_vt : False
Use SaltStack's utils.vt to stream output to console output=all.
keep_env : http_proxy,https_proxy,no_proxy
A list of env vars to preserve. May be passed as commma-delimited list.
chroot_fallback
if the container is not running, try to run the command using chroot default: false

CLI Example:

salt myminion lxc.run_all mycontainer 'ip addr show'
salt.modules.lxc.run_cmd(name, cmd, no_start=False, preserve_state=True, stdin=None, stdout=True, stderr=False, python_shell=True, output_loglevel='debug', use_vt=False, ignore_retcode=False, chroot_fallback=False, keep_env='http_proxy, https_proxy, no_proxy')

Deprecated since version 2015.5.0: Use lxc.run instead

salt.modules.lxc.run_stderr(name, cmd, no_start=False, preserve_state=True, stdin=None, python_shell=True, output_loglevel='debug', use_vt=False, ignore_retcode=False, chroot_fallback=False, keep_env='http_proxy, https_proxy, no_proxy')

New in version 2015.5.0.

Run cmd.run_stderr within a container

Warning

Many shell builtins do not work, failing with stderr similar to the following:

lxc_container: No such file or directory - failed to exec 'command'

The same error will be displayed if the command being run does not exist.

name
Name of the container in which to run the command
cmd
Command to run
no_start : False
If the container is not running, don't start it
preserve_state : True
After running the command, return the container to its previous state
stdin : None
Standard input to be used for the command
output_loglevel : debug
Level at which to log the output from the command. Set to quiet to suppress logging.
use_vt : False
Use SaltStack's utils.vt to stream output to console output=all.
keep_env : http_proxy,https_proxy,no_proxy
A list of env vars to preserve. May be passed as commma-delimited list.
chroot_fallback
if the container is not running, try to run the command using chroot default: false

CLI Example:

salt myminion lxc.run_stderr mycontainer 'ip addr show'
salt.modules.lxc.run_stdout(name, cmd, no_start=False, preserve_state=True, stdin=None, python_shell=True, output_loglevel='debug', use_vt=False, ignore_retcode=False, chroot_fallback=False, keep_env='http_proxy, https_proxy, no_proxy')

New in version 2015.5.0.

Run cmd.run_stdout within a container

Warning

Many shell builtins do not work, failing with stderr similar to the following:

lxc_container: No such file or directory - failed to exec 'command'

The same error will be displayed in stderr if the command being run does not exist. If no output is returned using this function, try using lxc.run_stderr or lxc.run_all.

name
Name of the container in which to run the command
cmd
Command to run
no_start : False
If the container is not running, don't start it
preserve_state : True
After running the command, return the container to its previous state
stdin : None
Standard input to be used for the command
output_loglevel : debug
Level at which to log the output from the command. Set to quiet to suppress logging.
use_vt : False
Use SaltStack's utils.vt to stream output to console output=all.
keep_env : http_proxy,https_proxy,no_proxy
A list of env vars to preserve. May be passed as commma-delimited list.
chroot_fallback
if the container is not running, try to run the command using chroot default: false

CLI Example:

salt myminion lxc.run_stdout mycontainer 'ifconfig -a'
salt.modules.lxc.running_systemd(name, cache=True)

Determine if systemD is running

CLI Example:

salt '*' lxc.running_systemd ubuntu
salt.modules.lxc.search_lxc_bridge()

Search the first bridge which is potentially available as LXC bridge

CLI Example:

salt '*' lxc.search_lxc_bridge
salt.modules.lxc.search_lxc_bridges()

Search which bridges are potentially available as LXC bridges

CLI Example:

salt '*' lxc.search_lxc_bridges
salt.modules.lxc.set_dns(name, dnsservers=None, searchdomains=None)

Changed in version 2015.5.0: The dnsservers and searchdomains parameters can now be passed as a comma-separated list.

Update /etc/resolv.conf

CLI Example:

salt myminion lxc.set_dns ubuntu "['8.8.8.8', '4.4.4.4']"
salt.modules.lxc.set_parameter(name, parameter, value)

Set the value of a cgroup parameter for a container.

CLI Example:

salt '*' lxc.set_parameter name parameter value
salt.modules.lxc.set_password(name, users, password, encrypted=True)

Changed in version 2015.5.0: Function renamed from set_pass to set_password. Additionally, this function now supports (and defaults to using) a password hash instead of a plaintext password.

Set the password of one or more system users inside containers

users
Comma-separated list (or python list) of users to change password
password
Password to set for the specified user(s)
encrypted : True

If true, password must be a password hash. Set to False to set a plaintext password (not recommended).

New in version 2015.5.0.

CLI Example:

salt '*' lxc.set_pass container-name root '$6$uJ2uAyLU$KoI67t8As/0fXtJOPcHKGXmUpcoYUcVR2K6x93walnShTCQvjRwq25yIkiCBOqgbfdKQSFnAo28/ek6716vEV1'
salt '*' lxc.set_pass container-name root foo encrypted=False
salt.modules.lxc.start(name, **kwargs)

Start the named container

restart : False

Deprecated since version 2015.5.0: Use lxc.restart

Restart the container if it is already running

CLI Example:

salt myminion lxc.start name
salt.modules.lxc.state(name)

Returns the state of a container.

CLI Example:

salt '*' lxc.state name
salt.modules.lxc.stop(name, kill=False)

Stop the named container

kill : False

Do not wait for the container to stop, kill all tasks in the container. Older LXC versions will stop containers like this irrespective of this argument.

Changed in version 2015.5.0: Default value changed to False

CLI Example:

salt myminion lxc.stop name
salt.modules.lxc.systemd_running_state(name)

Get the operational state of a systemd based container

CLI Example:

salt myminion lxc.systemd_running_state ubuntu
salt.modules.lxc.templates()

New in version 2015.5.0.

List the available LXC template scripts installed on the minion

CLI Examples:

salt myminion lxc.templates
salt.modules.lxc.test_bare_started_state(name)

Test if a non systemd container is fully started For now, it consists only to test if the container is attachable

CLI Example:

salt myminion lxc.test_bare_started_state ubuntu
salt.modules.lxc.test_sd_started_state(name)

Test if a systemd container is fully started

CLI Example:

salt myminion lxc.test_sd_started_state ubuntu
salt.modules.lxc.unfreeze(name)

Unfreeze the named container.

CLI Example:

salt '*' lxc.unfreeze name
salt.modules.lxc.update_lxc_conf(name, lxc_conf, lxc_conf_unset)

Edit LXC configuration options

CLI Example:

salt myminion lxc.update_lxc_conf ubuntu \
        lxc_conf="[{'network.ipv4.ip':'10.0.3.5'}]" \
        lxc_conf_unset="['lxc.utsname']"
salt.modules.lxc.version()

Return the actual lxc client version

New in version 2015.5.2.

CLI Example:

salt '*' lxc.version
salt.modules.lxc.wait_started(name, timeout=300)

Check that the system has fully inited

This is actually very important for systemD based containers

see https://github.com/saltstack/salt/issues/23847

CLI Example:

salt myminion lxc.wait_started ubuntu
salt.modules.lxc.write_conf(conf_file, conf)

Write out an LXC configuration file

This is normally only used internally. The format of the data structure must match that which is returned from lxc.read_conf(), with out_format set to commented.

An example might look like:

[
    {'lxc.utsname': '$CONTAINER_NAME'},
    '# This is a commented line\n',
    '\n',
    {'lxc.mount': '$CONTAINER_FSTAB'},
    {'lxc.rootfs': {'comment': 'This is another test',
                    'value': 'This is another test'}},
    '\n',
    {'lxc.network.type': 'veth'},
    {'lxc.network.flags': 'up'},
    {'lxc.network.link': 'br0'},
    {'lxc.network.mac': '$CONTAINER_MACADDR'},
    {'lxc.network.ipv4': '$CONTAINER_IPADDR'},
    {'lxc.network.name': '$CONTAINER_DEVICENAME'},
]

CLI Example:

salt 'minion' lxc.write_conf /etc/lxc/mycontainer.conf \
    out_format=commented