2013-04-09 16:41:00 -04:00
<?xml version='1.0'?> <!-- * - nxml - * -->
2016-04-06 16:55:34 +02:00
<?xml-stylesheet type="text/xsl" href="http://docbook.sourceforge.net/release/xsl/current/xhtml/docbook.xsl"?>
2013-04-09 16:41:00 -04:00
< !DOCTYPE refentry PUBLIC "-//OASIS//DTD DocBook XML V4.2//EN"
2016-04-06 16:55:34 +02:00
"http://www.oasis-open.org/docbook/xml/4.2/docbookx.dtd" [
<!ENTITY % entities SYSTEM "common.ent" >
%entities;
]>
2013-04-09 16:41:00 -04:00
<!--
2016-04-06 16:42:19 +02:00
NetworkManager.conf(5) manual page
Copyright 2010 - 2016 Red Hat, Inc.
Permission is granted to copy, distribute and/or modify this document
under the terms of the GNU Free Documentation License, Version 1.1
or any later version published by the Free Software Foundation;
with no Invariant Sections, no Front-Cover Texts, and no Back-Cover
Texts. You may obtain a copy of the GNU Free Documentation License
from the Free Software Foundation by visiting their Web site or by
writing to:
Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
2013-04-09 16:41:00 -04:00
-->
<refentry id= "NetworkManager.conf" >
<refentryinfo >
<title > NetworkManager.conf</title>
2013-05-07 09:54:28 +02:00
<author > NetworkManager developers</author>
2013-04-09 16:41:00 -04:00
</refentryinfo>
<refmeta >
<refentrytitle > NetworkManager.conf</refentrytitle>
<manvolnum > 5</manvolnum>
2013-05-07 09:54:28 +02:00
<refmiscinfo class= "source" > NetworkManager</refmiscinfo>
<refmiscinfo class= "manual" > Configuration</refmiscinfo>
2016-04-06 16:55:34 +02:00
<refmiscinfo class= "version" > &NM_VERSION; </refmiscinfo>
2013-04-09 16:41:00 -04:00
</refmeta>
<refnamediv >
<refname > NetworkManager.conf</refname>
<refpurpose > NetworkManager configuration file</refpurpose>
</refnamediv>
<refsynopsisdiv >
<para > <filename > /etc/NetworkManager/NetworkManager.conf</filename> ,
2015-06-08 17:51:04 +02:00
<filename > /etc/NetworkManager/conf.d/<replaceable > name</replaceable> .conf</filename> ,
2016-11-03 16:58:42 +01:00
<filename > /run/NetworkManager/conf.d/<replaceable > name</replaceable> .conf</filename> ,
2015-06-24 20:11:42 +02:00
<filename > /usr/lib/NetworkManager/conf.d/<replaceable > name</replaceable> .conf</filename> ,
<filename > /var/lib/NetworkManager/NetworkManager-intern.conf</filename>
2013-04-09 16:41:00 -04:00
</para>
</refsynopsisdiv>
<refsect1 >
<title > Description</title>
2015-06-24 20:11:42 +02:00
<para > <literal > NetworkManager.conf</literal> is the configuration file for NetworkManager. It is used
2013-04-09 16:41:00 -04:00
to set up various aspects of NetworkManager's behavior. The
2015-06-24 20:11:42 +02:00
location of the main file and configuration directories may be changed
through use of the <option > --config</option> , <option > --config-dir</option> ,
<option > --system-config-dir</option> , and <option > --intern-config</option>
argument for NetworkManager, respectively.
2013-04-09 16:41:00 -04:00
</para>
<para > If a default <literal > NetworkManager.conf</literal> is
provided by your distribution's packages, you should not modify
it, since your changes may get overwritten by package
updates. Instead, you can add additional <literal > .conf</literal>
2015-06-08 17:51:04 +02:00
files to the <literal > /etc/NetworkManager/conf.d</literal> directory.
These will be read in order, with later files overriding earlier ones.
Packages might install further configuration snippets to <literal > /usr/lib/NetworkManager/conf.d</literal> .
This directory is parsed first, even before <literal > NetworkManager.conf</literal> .
2016-11-03 16:58:42 +01:00
Scripts can also put per-boot configuration into <literal > /run/NetworkManager/conf.d</literal> .
2016-10-23 13:31:42 +02:00
This directory is parsed second, also before <literal > NetworkManager.conf</literal> .
2016-11-03 16:58:42 +01:00
The loading of a file <literal > /run/NetworkManager/conf.d/<replaceable > name</replaceable> .conf</literal>
2015-06-08 17:51:04 +02:00
can be prevented by adding a file <literal > /etc/NetworkManager/conf.d/<replaceable > name</replaceable> .conf</literal> .
2016-10-23 13:31:42 +02:00
Likewise, a file <literal > /usr/lib/NetworkManager/conf.d/<replaceable > name</replaceable> .conf</literal>
can be shadowed by putting a file of the same name to either <literal > /etc/NetworkManager/conf.d</literal>
2016-11-03 16:58:42 +01:00
or <literal > /run/NetworkManager/conf.d</literal> .
2013-04-09 16:41:00 -04:00
</para>
2015-06-24 20:11:42 +02:00
<para >
NetworkManager can overwrite certain user configuration options via D-Bus or other internal
operations. In this case it writes those changes to <literal > /var/lib/NetworkManager/NetworkManager-intern.conf</literal> .
This file is not intended to be modified by the user, but it is read last and can shadow
user configuration from <literal > NetworkManager.conf</literal> .
</para>
2016-05-30 16:43:39 +02:00
<para >
Certain settings from the configuration can be reloaded at runtime either by sending SIGHUP signal or via
D-Bus' Reload call.
</para>
2013-04-09 16:41:00 -04:00
</refsect1>
<refsect1 >
<title > File Format</title>
<para >
The configuration file format is so-called key file (sort of
ini-style format). It consists of sections (groups) of
key-value pairs. Lines beginning with a '#' and blank lines are
considered comments. Sections are started by a header line
containing the section enclosed in '[' and ']', and ended
implicitly by the start of the next section or the end of the
file. Each key-value pair must be contained in a section.
</para>
<para >
For keys that take a list of devices as their value, you can
specify devices by their MAC addresses or interface names, or
2015-06-07 12:03:19 +02:00
"*" to specify all devices. See <xref linkend= "device-spec" />
below.
2013-04-09 16:41:00 -04:00
</para>
<para >
Minimal system settings configuration file looks like this:
<programlisting >
2016-01-27 12:54:18 +01:00
[main]
plugins=keyfile
</programlisting>
2013-04-09 16:41:00 -04:00
</para>
<para >
As an extension to the normal keyfile format, you can also
append a value to a previously-set list-valued key by doing:
<programlisting >
2016-01-27 12:54:18 +01:00
plugins+=another-plugin
plugins-=remove-me
</programlisting>
2013-04-09 16:41:00 -04:00
</para>
</refsect1>
<refsect1 >
2013-12-23 13:37:10 -05:00
<title > <literal > main</literal> section</title>
2013-04-09 16:41:00 -04:00
<variablelist >
<varlistentry >
2016-01-27 12:54:18 +01:00
<term > <varname > plugins</varname> </term>
<listitem >
<para >
Lists system settings plugin names separated by ','. These
plugins are used to read and write system-wide
2018-06-06 13:10:09 +02:00
connection profiles. When multiple plugins are specified, the
2016-01-27 12:54:18 +01:00
connections are read from all listed plugins. When writing
connections, the plugins will be asked to save the
connection in the order listed here; if the first plugin
cannot write out that connection type (or can't write out
any connections) the next plugin is tried, etc. If none of
the plugins can save the connection, an error is returned
to the user.
</para>
<para >
2018-06-06 13:10:09 +02:00
The default value and the number of available plugins is
distro-specific. See <xref linkend= "settings-plugins" />
below for the available plugins.
Note that NetworkManager's native <literal > keyfile</literal>
plugin is always appended to the end of this list (if it doesn't
already appear earlier in the list).
2016-01-27 12:54:18 +01:00
</para>
</listitem>
2013-04-09 16:41:00 -04:00
</varlistentry>
2013-05-23 19:05:40 -03:00
<varlistentry >
2016-01-27 12:54:18 +01:00
<term > <varname > monitor-connection-files</varname> </term>
2020-04-10 14:49:44 +02:00
<listitem > <para > This setting is deprecated and has no effect. Profiles
from disk are never automatically reloaded. Use for example <literal > nmcli connection (re)load</literal>
for that.</para> </listitem>
2013-05-23 19:05:40 -03:00
</varlistentry>
2014-08-14 13:34:57 +02:00
<varlistentry >
2016-01-27 12:54:18 +01:00
<term > <varname > auth-polkit</varname> </term>
<listitem > <para > Whether the system uses PolicyKit for authorization.
2019-12-10 08:51:03 +01:00
If <literal > true</literal> , non-root requests are authorized using PolicyKit.
Requests from root (user ID zero) are always granted without asking PolicyKit.
If <literal > false</literal> , all requests will be allowed and PolicyKit is
not used. If set to <literal > root-only</literal> PolicyKit is not used and
all requests except root are denied.
2016-11-25 14:17:30 +01:00
The default value is <literal > &NM_CONFIG_DEFAULT_MAIN_AUTH_POLKIT_TEXT; </literal> .
2016-01-27 12:54:18 +01:00
</para> </listitem>
2014-08-14 13:34:57 +02:00
</varlistentry>
2013-04-09 16:41:00 -04:00
<varlistentry >
2016-01-27 12:54:18 +01:00
<term > <varname > dhcp</varname> </term>
<listitem > <para > This key sets up what DHCP client
NetworkManager will use. Allowed values are
<literal > dhclient</literal> , <literal > dhcpcd</literal> , and
<literal > internal</literal> . The <literal > dhclient</literal>
and <literal > dhcpcd</literal> options require the indicated
clients to be installed. The <literal > internal</literal>
option uses a built-in DHCP client which is not currently as
2016-02-21 11:43:34 +01:00
featureful as the external clients.</para>
2016-11-25 14:17:30 +01:00
<para > If this key is missing, it defaults to <literal > &NM_CONFIG_DEFAULT_MAIN_DHCP; </literal> .
2020-07-30 14:01:44 +00:00
If the chosen plugin is not available, clients are looked for
2016-10-14 11:00:16 +02:00
in this order: <literal > dhclient</literal> , <literal > dhcpcd</literal> ,
2016-01-27 12:54:18 +01:00
<literal > internal</literal> .</para> </listitem>
2013-04-09 16:41:00 -04:00
</varlistentry>
<varlistentry >
2016-01-27 12:54:18 +01:00
<term > <varname > no-auto-default</varname> </term>
<listitem > <para > Specify devices for which
NetworkManager shouldn't create default wired connection
(Auto eth0). By default, NetworkManager creates a temporary
wired connection for any Ethernet device that is managed and
doesn't have a connection configured. List a device in this
option to inhibit creating the default connection for the
device. May have the special value <literal > *</literal> to
apply to all devices.</para>
<para > When the default wired connection is deleted or saved
to a new persistent connection by a plugin, the device is
added to a list in the file
2018-10-24 18:48:41 +02:00
<filename > &nmstatedir; /no-auto-default.state</filename>
2016-01-27 12:54:18 +01:00
to prevent creating the default connection for that device
again.</para>
<para > See <xref linkend= "device-spec" /> for the syntax how to
specify a device.
</para>
<para >
Example:
<programlisting >
2015-02-09 15:33:28 +01:00
no-auto-default=00:22:68:5c:5d:c4,00:1e:65:ff:aa:ee
no-auto-default=eth0,eth1
no-auto-default=*
2016-01-27 12:54:18 +01:00
</programlisting>
2016-01-27 12:54:18 +01:00
</para>
</listitem>
2013-04-09 16:41:00 -04:00
</varlistentry>
<varlistentry >
2016-01-27 12:54:18 +01:00
<term > <varname > ignore-carrier</varname> </term>
<listitem >
<para >
config: make "ignore-carrier" a per-device configuration option
NetworkManager.conf already contains several per-device settings,
that is, settings that have a device-spec as argument.
main.ignore-carrier
main.no-auto-default
main.assume-ipv6ll-only
keyfile.unmanged-devices
Optimally, these settings should be moved to the new [device*]
section.
For now, only move main.ignore-carrier there. For the others
it may not make sense to do so:
- main.no-auto-default: is already merged with internal state
from /var/lib/NetworkManager/no-auto-default.state. While
NMConfig's write API would be fine to also persist and merge
the no-auto-default setting, we'd still have to read the old
file too. Thus, deprecating this setting gets quite cumbersome
to still handle the old state file.
Also, it seems a less useful setting to configure in the
global configuration aside setting main.no-auto-default=*.
- main.assume-ipv6ll-only: one day, I hope that we no longer
assume connections at all, and this setting becomes entirely
obsolete.
- keyfile.unmanged-devices: this sets NM_UNMANAGED_USER_SETTINGS,
which cannot be overruled via D-Bus. For a future device.managed
setting we want it it to be overwritable via D-Bus by an explicit
user action. Thus, a device.managed property should have a different
semantic, this should be more like a device.unmanaged-force setting,
which could be done.
2016-06-22 13:43:08 +02:00
This setting is deprecated for the per-device setting
<literal > ignore-carrier</literal> which overwrites this setting
2016-08-26 20:52:28 +02:00
if specified (See <xref linkend= "ignore-carrier" /> ).
config: make "ignore-carrier" a per-device configuration option
NetworkManager.conf already contains several per-device settings,
that is, settings that have a device-spec as argument.
main.ignore-carrier
main.no-auto-default
main.assume-ipv6ll-only
keyfile.unmanged-devices
Optimally, these settings should be moved to the new [device*]
section.
For now, only move main.ignore-carrier there. For the others
it may not make sense to do so:
- main.no-auto-default: is already merged with internal state
from /var/lib/NetworkManager/no-auto-default.state. While
NMConfig's write API would be fine to also persist and merge
the no-auto-default setting, we'd still have to read the old
file too. Thus, deprecating this setting gets quite cumbersome
to still handle the old state file.
Also, it seems a less useful setting to configure in the
global configuration aside setting main.no-auto-default=*.
- main.assume-ipv6ll-only: one day, I hope that we no longer
assume connections at all, and this setting becomes entirely
obsolete.
- keyfile.unmanged-devices: this sets NM_UNMANAGED_USER_SETTINGS,
which cannot be overruled via D-Bus. For a future device.managed
setting we want it it to be overwritable via D-Bus by an explicit
user action. Thus, a device.managed property should have a different
semantic, this should be more like a device.unmanaged-force setting,
which could be done.
2016-06-22 13:43:08 +02:00
Otherwise, it is a list of matches to specify for which device
carrier should be ignored. See <xref linkend= "device-spec" /> for the
2017-06-01 23:57:03 +02:00
syntax how to specify a device. Note that master types like
bond, bridge, and team ignore carrier by default. You can however
revert that default using the "except:" specifier (or better,
use the per-device setting instead of the deprecated setting).
2016-01-27 12:54:18 +01:00
</para>
</listitem>
2013-04-09 16:41:00 -04:00
</varlistentry>
2015-05-20 11:36:37 +02:00
<varlistentry >
2016-01-27 12:54:18 +01:00
<term > <varname > assume-ipv6ll-only</varname> </term>
<listitem >
<para >
Specify devices for which NetworkManager will try to
generate a connection based on initial configuration when
the device only has an IPv6 link-local address.
</para>
<para > See <xref linkend= "device-spec" /> for the syntax how to
specify a device.
</para>
</listitem>
2015-05-20 11:36:37 +02:00
</varlistentry>
2014-05-23 13:24:15 +02:00
<varlistentry >
2016-01-27 12:54:18 +01:00
<term > <varname > configure-and-quit</varname> </term>
<listitem >
<para >
When set to '<literal > true</literal> ', NetworkManager quits after
performing initial network configuration but spawns small helpers
to preserve DHCP leases and IPv6 addresses. This is useful in
environments where network setup is more or less static or it is
desirable to save process time but still handle some dynamic
configurations. When this option is <literal > true</literal> ,
2018-11-21 11:32:38 +01:00
network configuration for Wi-Fi, WWAN, Bluetooth, ADSL, and PPPoE
2016-01-27 12:54:18 +01:00
interfaces cannot be preserved due to their use of external
services, and these devices will be deconfigured when NetworkManager
quits even though other interface's configuration may be preserved.
Also, to preserve DHCP addresses the '<literal > dhcp</literal> ' option
must be set to '<literal > internal</literal> '. The default value of
the '<literal > configure-and-quit</literal> ' option is
'<literal > false</literal> ', meaning that NetworkManager will continue
running after initial network configuration and continue responding
to system and hardware events, D-Bus requests, and user commands.
</para>
</listitem>
2014-05-23 13:24:15 +02:00
</varlistentry>
2017-03-07 10:27:58 +01:00
<varlistentry >
<term > <varname > hostname-mode</varname> </term>
<listitem >
<para >
Set the management mode of the hostname. This parameter will
affect only the transient hostname. If a valid static hostname is set,
NetworkManager will skip the update of the hostname despite the value of
this option. An hostname empty or equal to 'localhost', 'localhost6',
'localhost.localdomain' or 'localhost6.localdomain' is considered invalid.
</para>
<para > <literal > default</literal> : NetworkManager will update the hostname
with the one provided via DHCP on the main connection (the one with a default
route). If not present, the hostname will be updated to the last one set
outside NetworkManager. If it is not valid, NetworkManager will try to recover
the hostname from the reverse lookup of the IP address of the main connection.
If this fails too, the hostname will be set to 'localhost.localdomain'.
</para>
<para > <literal > dhcp</literal> : NetworkManager will update the transient hostname
only with information coming from DHCP. No fallback nor reverse lookup will be
performed, but when the dhcp connection providing the hostname is deactivated,
the hostname is reset to the last hostname set outside NetworkManager or
'localhost' if none valid is there.
</para>
<para > <literal > none</literal> : NetworkManager will not manage the transient
hostname and will never set it.
</para>
</listitem>
</varlistentry>
2013-04-09 16:41:00 -04:00
<varlistentry >
2016-01-27 12:54:18 +01:00
<term > <varname > dns</varname> </term>
2018-09-21 10:54:08 +02:00
<listitem > <para > Set the DNS processing mode.</para>
<para > If the key is unspecified, <literal > default</literal> is used,
2016-11-14 12:32:11 +01:00
unless <filename > /etc/resolv.conf</filename> is a symlink to
2018-02-18 09:49:26 +01:00
<filename > /run/systemd/resolve/stub-resolv.conf</filename> ,
2016-11-14 12:32:11 +01:00
<filename > /run/systemd/resolve/resolv.conf</filename> ,
<filename > /lib/systemd/resolv.conf</filename> or
<filename > /usr/lib/systemd/resolv.conf</filename> .
In that case, <literal > systemd-resolved</literal> is chosen automatically.
</para>
2016-09-08 11:53:55 +02:00
<para > <literal > default</literal> : NetworkManager will update
2016-11-14 12:32:11 +01:00
<filename > /etc/resolv.conf</filename> to reflect the nameservers
provided by currently active connections.</para>
2016-01-27 12:54:18 +01:00
<para > <literal > dnsmasq</literal> : NetworkManager will run
2019-05-12 17:18:44 +02:00
dnsmasq as a local caching nameserver, using "Conditional Forwarding"
if you are connected to a VPN, and then update
2016-01-27 12:54:18 +01:00
<filename > resolv.conf</filename> to point to the local
2016-09-20 15:04:39 +02:00
nameserver. It is possible to pass custom options to the
dnsmasq instance by adding them to files in the
"<filename > &sysconfdir; /NetworkManager/dnsmasq.d/</filename> "
directory. Note that when multiple upstream servers are
available, dnsmasq will initially contact them in parallel and
then use the fastest to respond, probing again other servers
after some time. This behavior can be modified passing the
'all-servers' or 'strict-order' options to dnsmasq (see the
manual page for more details).</para>
dns: write original DNS servers to /var/run/NetworkManager/no-stub-resolv.conf
When a DNS plugin is enabled (like "main.dns=dnsmasq" or "main.dns=systemd-resolved"),
the name servers announced to the rc-manager are coerced to be 127.0.0.1
or 127.0.0.53.
Depending on the "main.rc-manager" setting, also "/etc/resolv.conf"
contains only this coerced name server to the local caching service.
The same is true for "/var/run/NetworkManager/resolv.conf" file, which
contains what we would write to "/etc/resolv.conf" (depending on
the "main.rc-manager" configuration).
Write a new file "/var/run/NetworkManager/no-stub-resolv.conf", which contains
the original name servers, uncoerced. Like "/var/run/NetworkManager/resolv.conf",
this file is always written.
The effect is, when one enables "main.dns=systemd-resolved", then there
is still a file "no-stub-resolv.conf" with the same content as with
"main.dns=default".
The no-stub-resolv.conf may be a possible solution, when a user wants
NetworkManager to update systemd-resolved, but still have a regular
/etc/resolv.conf [1]. For that, the user could configure
[main]
dns=systemd-resolved
rc-manager=unmanaged
and symlink "/etc/resolv.conf" to "/var/run/NetworkManager/no-stub-resolv.conf".
This is not necessarily the only solution for the problem and does not preclude
options for updating systemd-resolved in combination with other DNS plugins.
[1] https://gitlab.freedesktop.org/NetworkManager/NetworkManager/issues/20
2018-09-12 21:16:34 +02:00
<para > <literal > systemd-resolved</literal> : NetworkManager will
push the DNS configuration to systemd-resolved</para>
2016-01-27 12:54:18 +01:00
<para > <literal > unbound</literal> : NetworkManager will talk
2019-05-12 17:18:44 +02:00
to unbound and dnssec-triggerd, using "Conditional Forwarding"
with DNSSEC support. <filename > /etc/resolv.conf</filename>
2016-01-27 12:54:18 +01:00
will be managed by dnssec-trigger daemon.</para>
<para > <literal > none</literal> : NetworkManager will not
2016-05-23 08:46:27 +02:00
modify resolv.conf. This implies
<literal > rc-manager</literal> <literal > unmanaged</literal> </para>
dns: write original DNS servers to /var/run/NetworkManager/no-stub-resolv.conf
When a DNS plugin is enabled (like "main.dns=dnsmasq" or "main.dns=systemd-resolved"),
the name servers announced to the rc-manager are coerced to be 127.0.0.1
or 127.0.0.53.
Depending on the "main.rc-manager" setting, also "/etc/resolv.conf"
contains only this coerced name server to the local caching service.
The same is true for "/var/run/NetworkManager/resolv.conf" file, which
contains what we would write to "/etc/resolv.conf" (depending on
the "main.rc-manager" configuration).
Write a new file "/var/run/NetworkManager/no-stub-resolv.conf", which contains
the original name servers, uncoerced. Like "/var/run/NetworkManager/resolv.conf",
this file is always written.
The effect is, when one enables "main.dns=systemd-resolved", then there
is still a file "no-stub-resolv.conf" with the same content as with
"main.dns=default".
The no-stub-resolv.conf may be a possible solution, when a user wants
NetworkManager to update systemd-resolved, but still have a regular
/etc/resolv.conf [1]. For that, the user could configure
[main]
dns=systemd-resolved
rc-manager=unmanaged
and symlink "/etc/resolv.conf" to "/var/run/NetworkManager/no-stub-resolv.conf".
This is not necessarily the only solution for the problem and does not preclude
options for updating systemd-resolved in combination with other DNS plugins.
[1] https://gitlab.freedesktop.org/NetworkManager/NetworkManager/issues/20
2018-09-12 21:16:34 +02:00
<para > Note that the plugins <literal > dnsmasq</literal> , <literal > systemd-resolved</literal>
and <literal > unbound</literal> are caching local nameservers.
Hence, when NetworkManager writes <filename > &nmrundir; /resolv.conf</filename>
and <filename > /etc/resolv.conf</filename> (according to <literal > rc-manager</literal>
setting below), the name server there will be localhost only.
NetworkManager also writes a file <filename > &nmrundir; /no-stub-resolv.conf</filename>
that contains the original name servers pushed to the DNS plugin.</para>
2020-01-29 16:47:37 +01:00
<para > When using <literal > dnsmasq</literal> and <literal > systemd-resolved</literal>
per-connection added dns servers will always be queried using
2020-01-29 17:32:47 +01:00
the device the connection has been activated on.</para>
2016-01-27 12:54:18 +01:00
</listitem>
2013-04-09 16:41:00 -04:00
</varlistentry>
2014-03-06 22:04:44 +01:00
2015-04-20 14:45:09 +02:00
<varlistentry >
2016-01-27 12:54:18 +01:00
<term > <varname > rc-manager</varname> </term>
<listitem > <para > Set the <filename > resolv.conf</filename>
2016-09-02 17:04:37 +02:00
management mode. The default value depends on NetworkManager build
options, and this version of NetworkManager was build with a default of
2016-11-25 14:17:30 +01:00
"<literal > &NM_CONFIG_DEFAULT_MAIN_RC_MANAGER; </literal> ".
2016-05-19 16:43:48 +02:00
Regardless of this setting, NetworkManager will
dns: change behavior for "rc-manager=symlink" to preserve "/etc/resolv.conf" as file
The purpose of "rc-manager=symlink" is so that the administrator can point
the "/etc/resolv.conf" as a symlink to a certain file, and thus indicating
that a certain component is responsible to manage resolv.conf, while others
should stay away from it.
For example, systemd-resolved never touches "/etc/resolv.conf", but
expects the admin to setup the symlink appropriately. It also recognizes
whether the symlink points to it's own resolv.conf in /run or to another
component.
Previously, "rc-manager=symlink" would always replace a regular file
with a symlink to "/var/run/NetworkManager/resolv.conf". Only if
"/etc/resolv.conf" is already a symlink somewhere else, NM would not
touch it. This with the exception that if "/etc/resolv.conf" points to
"/var/run/NetworkManager/resolv.conf", it would replace the symlink
with the same link to raise inotify events.
Change behavior so if "/etc/resolv.conf" is already a regular file, keep
it as file.
This means, if you have multiple components that don't care, everybody
can write the "/etc/resolv.conf" (as file) and there is no clear
expressed responsibility.
It was wrong that NetworkManager would convert the file to a symlink,
this should be reserved to the admin. Instead, NetworkManager should
accept that the intent is unspecified and preserve the regular file.
It's up to the admin to replace the symlink to somewhere else (to keep
NM off), or to point it to "/var/run/NetworkManager/resolv.conf", to show
the explicit intent.
The wrong behavior causes dangling symlinks when somebody disables
NetworkManager for good.
https://bugzilla.redhat.com/show_bug.cgi?id=1367551
2017-02-10 16:30:59 +01:00
always write resolv.conf to its runtime state directory
<filename > &nmrundir; /resolv.conf</filename> .</para>
<para > <literal > symlink</literal> : If <filename > /etc/resolv.conf</filename> is
a regular file, NetworkManager will replace the file on update. If
<filename > /etc/resolv.conf</filename> is instead a symlink, NetworkManager
will leave it alone. Unless the symlink points to the internal file
<filename > &nmrundir; /resolv.conf</filename> ,
in which case the symlink will be updated to emit an inotify notification.
This allows the user to conveniently instruct NetworkManager not
to manage <filename > /etc/resolv.conf</filename> by replacing it with
a symlink.</para>
2016-03-21 21:01:25 +01:00
<para > <literal > file</literal> : NetworkManager will write
2016-06-03 10:33:54 +02:00
<filename > /etc/resolv.conf</filename> as file. If it finds
2018-06-01 08:29:15 +02:00
a symlink to an existing target, it will follow the symlink and
dns: change main.rc-manager=file behavior to always follow symlink
With "main.rc-manager=file", if /etc/resolv.conf is a symlink, NetworkManager
would follow the symlink and update the file instead.
However, note that realpath() only returns a target, if the file actually
exists. That means, if /etc/resolv.conf is a dangling symlink, NetworkManager
would replace the symlink with a file.
This was the only case in which NetworkManager would every change a symlink
resolv.conf to a file. I think this is undesired behavior.
This is a change in long established behavior. Although note that there were several
changes regarding rc-manager settings in the past. See for example commit [1] and [2].
Now, first still try using realpath() as before. Only if that fails, try
to resolve /etc/resolv.conf as a symlink with readlink().
Following the dangling symlink is likely not a problem for the user, it
probably is even desired. The part that most likely can cause problems
is if the destination file is not writable. That happens for example, if
the destination's parent directories are missing. In this case, NetworkManager
will now fail to write resolv.conf and log a warning. This has the potential of
breaking existing setups, but it really is a mis-configuration from the user's
side.
This fixes for example the problem, if the user configures
/etc/resolv.conf as symlink to /tmp/my-resolv.conf. At boot, the file
would not exist, and NetworkManager would previously always replace the
link with a plain file. Instead, it should follow the symlink and create
the file.
[1] 718fd2243690b8c72dd1cb32f67114f304542082
[2] 15177a34be297654086005f2d796e6a4c6a1b918
https://github.com/NetworkManager/NetworkManager/pull/127
2018-06-01 10:08:07 +02:00
update the target instead. In no case will an existing symlink
be replaced by a file. Note that older versions of NetworkManager
behaved differently and would replace dangling symlinks with a
plain file.</para>
2016-01-27 12:54:18 +01:00
<para > <literal > resolvconf</literal> : NetworkManager will run
resolvconf to update the DNS configuration.</para>
<para > <literal > netconfig</literal> : NetworkManager will run
netconfig to update the DNS configuration.</para>
2016-04-21 18:44:48 +02:00
<para > <literal > unmanaged</literal> : don't touch
2016-04-29 21:54:45 +02:00
<filename > /etc/resolv.conf</filename> .</para>
2016-04-21 16:41:40 +02:00
<para > <literal > none</literal> : deprecated alias for
<literal > symlink</literal> .</para>
2016-01-27 12:54:18 +01:00
</listitem>
2015-04-20 14:45:09 +02:00
</varlistentry>
2018-07-03 14:35:32 +02:00
<varlistentry >
<term > <varname > systemd-resolved</varname> </term>
<listitem > <para > Send the connection DNS configuration to
<literal > systemd-resolved</literal> . Defaults to "<literal > true</literal> ".
</para>
<para > Note that this setting is complementary to the
2018-12-01 15:48:19 +01:00
<varname > dns</varname> setting. You can keep this enabled while using
2018-07-03 14:35:32 +02:00
<varname > dns</varname> set to another DNS plugin alongside
<literal > systemd-resolved</literal> , or <varname > dns</varname> set to
<literal > systemd-resolved</literal> to configure the system resolver to use
<literal > systemd-resolved</literal> .</para>
2018-12-01 15:48:19 +01:00
<para > If systemd-resolved is enabled, the connectivity check resolves the
hostname per-device.</para>
2018-07-03 14:35:32 +02:00
</listitem>
</varlistentry>
2014-03-06 22:04:44 +01:00
<varlistentry >
<term > <varname > debug</varname> </term>
<listitem > <para > Comma separated list of options to aid
debugging. This value will be combined with the environment
2020-07-01 17:20:40 -04:00
variable <literal > NM_DEBUG</literal> . Currently, the following
2014-03-06 22:04:44 +01:00
values are supported:</para>
<para >
<literal > RLIMIT_CORE</literal> : set ulimit -c unlimited
2015-03-31 07:41:56 +02:00
to write out core dumps. Beware, that a core dump can contain
sensitive information such as passwords or configuration settings.
</para>
<para >
<literal > fatal-warnings</literal> : set g_log_set_always_fatal()
to core dump on warning messages from glib. This is equivalent
to the --g-fatal-warnings command line option.
2014-03-06 22:04:44 +01:00
</para>
</listitem>
</varlistentry>
2016-10-08 10:27:18 +02:00
<varlistentry >
<term > <varname > autoconnect-retries-default</varname> </term>
<listitem >
<para >
The number of times a connection activation should be
automatically tried before switching to another one. This
value applies only to connections that can auto-connect
and have a
<literal > connection.autoconnect-retries</literal> property
2017-10-04 13:57:15 +02:00
set to -1. If not specified, connections will be tried 4
times. Setting this value to 1 means to try activation once,
without retry.
2016-10-08 10:27:18 +02:00
</para>
</listitem>
</varlistentry>
2017-05-15 17:17:26 +02:00
<varlistentry >
<term > <varname > slaves-order</varname> </term>
<listitem >
<para >
This key specifies in which order slave connections are
auto-activated on boot or when the master activates
them. Allowed values are <literal > name</literal> (order
connection by interface name, the default), or
<literal > index</literal> (order slaves by their kernel
index).
</para>
</listitem>
</varlistentry>
2013-04-09 16:41:00 -04:00
</variablelist>
</refsect1>
<refsect1 >
<title > <literal > keyfile</literal> section</title>
2013-12-23 13:37:10 -05:00
<para > This section contains keyfile-plugin-specific options, and
is normally only used when you are not using any other
distro-specific plugin.</para>
2013-04-09 16:41:00 -04:00
<para >
<variablelist >
2016-01-27 12:54:18 +01:00
<varlistentry >
<term > <varname > hostname</varname> </term>
<listitem > <para > This key is deprecated and has no effect
2016-11-11 14:26:25 +01:00
since the hostname is now stored in <filename > /etc/hostname</filename>
or other system configuration files according to build options.
2016-01-27 12:54:18 +01:00
</para> </listitem>
</varlistentry>
2015-10-09 18:16:42 +02:00
<varlistentry >
<term > <varname > path</varname> </term>
<listitem >
<para > The location where keyfiles are read and stored.
2017-03-24 10:36:51 +01:00
This defaults to "<filename > &sysconfdir; /NetworkManager/system-connections</filename> ".
2015-10-09 18:16:42 +02:00
</para>
</listitem>
</varlistentry>
2016-01-27 12:54:18 +01:00
<varlistentry >
<term > <varname > unmanaged-devices</varname> </term>
<listitem > <para > Set devices that should be ignored by
NetworkManager.
</para>
2020-07-01 17:20:40 -04:00
<para > See <xref linkend= "device-spec" /> for the syntax on how to
2016-01-27 12:54:18 +01:00
specify a device.
</para>
<para >
Example:
<programlisting >
2013-04-09 16:41:00 -04:00
unmanaged-devices=interface-name:em4
unmanaged-devices=mac:00:22:68:1c:59:b1;mac:00:1E:65:30:D1:C4;interface-name:eth2
2016-01-27 12:54:18 +01:00
</programlisting>
2016-01-27 12:54:18 +01:00
</para>
</listitem>
</varlistentry>
2013-04-09 16:41:00 -04:00
</variablelist>
</para>
</refsect1>
<refsect1 >
<title > <literal > ifupdown</literal> section</title>
2013-12-23 13:37:10 -05:00
<para > This section contains ifupdown-specific options and thus only
has effect when using the <literal > ifupdown</literal> plugin.</para>
2013-04-09 16:41:00 -04:00
<para >
<variablelist >
2016-01-27 12:54:18 +01:00
<varlistentry >
<term > <varname > managed</varname> </term>
<listitem > <para > If set to <literal > true</literal> , then
interfaces listed in
<filename > /etc/network/interfaces</filename> are managed by
NetworkManager. If set to <literal > false</literal> , then
any interface listed in
<filename > /etc/network/interfaces</filename> will be ignored
by NetworkManager. Remember that NetworkManager controls the
default route, so because the interface is ignored,
NetworkManager may assign the default route to some other
interface.</para>
<para >
The default value is <literal > false</literal> .
</para>
</listitem>
</varlistentry>
2013-04-09 16:41:00 -04:00
</variablelist>
</para>
</refsect1>
<refsect1 >
<title > <literal > logging</literal> section</title>
<para > This section controls NetworkManager's logging. Any
settings here are overridden by the <option > --log-level</option>
and <option > --log-domains</option> command-line options.</para>
<para >
<variablelist >
2016-01-27 12:54:18 +01:00
<varlistentry >
<term > <varname > level</varname> </term>
<listitem > <para > The default logging verbosity level.
One of <literal > OFF</literal> , <literal > ERR</literal> ,
<literal > WARN</literal> , <literal > INFO</literal> ,
<literal > DEBUG</literal> , <literal > TRACE</literal> . The ERR
level logs only critical errors. WARN logs warnings that may
reflect operation. INFO logs various informational messages that
are useful for tracking state and operations. DEBUG enables
verbose logging for debugging purposes. TRACE enables even more
verbose logging then DEBUG level. Subsequent levels also log
all messages from earlier levels; thus setting the log level
to INFO also logs error and warning messages.</para> </listitem>
</varlistentry>
<varlistentry >
<term > <varname > domains</varname> </term>
<listitem > <para > The following log domains are available:
PLATFORM, RFKILL, ETHER, WIFI, BT, MB, DHCP4, DHCP6, PPP,
WIFI_SCAN, IP4, IP6, AUTOIP4, DNS, VPN, SHARING, SUPPLICANT,
AGENTS, SETTINGS, SUSPEND, CORE, DEVICE, OLPC, WIMAX,
INFINIBAND, FIREWALL, ADSL, BOND, VLAN, BRIDGE, DBUS_PROPS,
2017-11-06 11:42:25 +01:00
TEAM, CONCHECK, DCB, DISPATCH, AUDIT, SYSTEMD, VPN_PLUGIN,
PROXY.</para>
2016-01-27 12:54:18 +01:00
<para > In addition, these special domains can be used: NONE,
ALL, DEFAULT, DHCP, IP.</para>
<para > You can specify per-domain log level overrides by
adding a colon and a log level to any domain. E.g.,
"<literal > WIFI:DEBUG,WIFI_SCAN:OFF</literal> ".</para> </listitem>
</varlistentry>
<varlistentry >
2014-01-15 17:32:07 +01:00
<para > Domain descriptions:
<simplelist type= "horiz" columns= "1" >
<member > PLATFORM : OS (platform) operations</member>
<member > RFKILL : RFKill subsystem operations</member>
<member > ETHER : Ethernet device operations</member>
<member > WIFI : Wi-Fi device operations</member>
<member > BT : Bluetooth operations</member>
<member > MB : Mobile broadband operations</member>
<member > DHCP4 : DHCP for IPv4</member>
<member > DHCP6 : DHCP for IPv6</member>
<member > PPP : Point-to-point protocol operations</member>
<member > WIFI_SCAN : Wi-Fi scanning operations</member>
<member > IP4 : IPv4-related operations</member>
<member > IP6 : IPv6-related operations</member>
2015-04-20 18:13:41 +02:00
<member > AUTOIP4 : AutoIP operations</member>
2014-01-15 17:32:07 +01:00
<member > DNS : Domain Name System related operations</member>
2014-01-16 09:55:48 +01:00
<member > VPN : Virtual Private Network connections and operations</member>
2017-11-06 11:52:05 +01:00
<member > SHARING : Connection sharing. With TRACE level log queries for dnsmasq instance</member>
2014-01-15 17:32:07 +01:00
<member > SUPPLICANT : WPA supplicant related operations</member>
<member > AGENTS : Secret agents operations and communication</member>
<member > SETTINGS : Settings/config service operations</member>
<member > SUSPEND : Suspend/resume</member>
<member > CORE : Core daemon and policy operations</member>
<member > DEVICE : Activation and general interface operations</member>
<member > OLPC : OLPC Mesh device operations</member>
<member > WIMAX : WiMAX device operations</member>
<member > INFINIBAND : InfiniBand device operations</member>
<member > FIREWALL : FirewallD related operations</member>
<member > ADSL : ADSL device operations</member>
<member > BOND : Bonding operations</member>
<member > VLAN : VLAN operations</member>
<member > BRIDGE : Bridging operations</member>
2014-01-16 09:55:48 +01:00
<member > DBUS_PROPS : D-Bus property changes</member>
2014-01-15 17:32:07 +01:00
<member > TEAM : Teaming operations</member>
<member > CONCHECK : Connectivity check</member>
<member > DCB : Data Center Bridging (DCB) operations</member>
2014-04-24 17:27:47 +02:00
<member > DISPATCH : Dispatcher scripts</member>
2015-07-13 10:19:33 +02:00
<member > AUDIT : Audit records</member>
2016-05-17 19:57:55 +02:00
<member > SYSTEMD : Messages from internal libsystemd</member>
2016-05-17 19:29:38 +02:00
<member > VPN_PLUGIN : logging messages from VPN plugins</member>
2017-11-06 11:42:25 +01:00
<member > PROXY : logging messages for proxy handling</member>
2014-01-15 17:32:07 +01:00
<member > </member>
<member > NONE : when given by itself logging is disabled</member>
<member > ALL : all log domains</member>
<member > DEFAULT : default log domains</member>
<member > DHCP : shortcut for "DHCP4,DHCP6"</member>
<member > IP : shortcut for "IP4,IP6"</member>
2014-05-12 19:12:28 +02:00
<member > </member>
<member > HW : deprecated alias for "PLATFORM"</member>
2014-01-15 17:32:07 +01:00
</simplelist>
</para>
2016-05-17 19:29:38 +02:00
<para >
In general, the logfile should not contain passwords or private data. However,
you are always advised to check the file before posting it online or attaching
2016-06-16 22:30:41 +02:00
to a bug report. <literal > VPN_PLUGIN</literal> is special as it might reveal
private information of the VPN plugins with verbose levels. Therefore this domain
will be excluded when setting <literal > ALL</literal> or <literal > DEFAULT</literal>
to more verbose levels then <literal > INFO</literal> .
</para>
2014-01-15 17:32:07 +01:00
</varlistentry>
2016-01-27 12:54:18 +01:00
<varlistentry >
<term > <varname > backend</varname> </term>
<listitem > <para > The logging backend. Supported values
2018-06-21 09:21:57 +02:00
are "<literal > syslog</literal> " and "<literal > journal</literal> ".
When NetworkManager is started with "<literal > --debug</literal> "
in addition all messages will be printed to stderr.
If unspecified, the default is "<literal > &NM_CONFIG_DEFAULT_LOGGING_BACKEND_TEXT; </literal> ".
2016-01-27 12:54:18 +01:00
</para> </listitem>
</varlistentry>
<varlistentry >
<term > <varname > audit</varname> </term>
<listitem > <para > Whether the audit records are delivered to
auditd, the audit daemon. If <literal > false</literal> , audit
records will be sent only to the NetworkManager logging
system. If set to <literal > true</literal> , they will be also
2016-04-06 16:55:34 +02:00
sent to auditd. The default value is <literal > &NM_CONFIG_DEFAULT_LOGGING_AUDIT_TEXT; </literal> .
2016-01-27 12:54:18 +01:00
</para> </listitem>
</varlistentry>
2013-04-09 16:41:00 -04:00
</variablelist>
</para>
</refsect1>
2015-05-15 11:36:28 +02:00
<refsect1 >
<title > <literal > connection</literal> section</title>
2016-01-27 13:30:25 +01:00
<para > Specify default values for connections.
</para>
<para >
Example:
<programlisting >
[connection]
ipv6.ip6-privacy=0
</programlisting>
</para>
<refsect2 >
<title > Supported Properties</title>
2015-05-15 11:36:28 +02:00
<para >
2016-01-27 13:30:25 +01:00
Not all properties can be overwritten, only the following
properties are supported to have their default values configured
2016-05-30 20:24:32 +02:00
(see <link linkend= 'nm-settings' > <citerefentry > <refentrytitle > nm-settings</refentrytitle> <manvolnum > 5</manvolnum> </citerefentry> </link> for details).
2016-01-27 13:30:25 +01:00
A default value is only consulted if the corresponding per-connection value
explicitly allows for that.
2018-11-14 14:52:21 +01:00
<!-- The following comment is used by check - config - options.sh, don't remove it. -->
<!-- start connection defaults -->
2016-01-27 13:30:25 +01:00
<variablelist >
2018-11-14 16:35:53 +01:00
<varlistentry >
<term > <varname > 802-1x.auth-timeout</varname> </term>
</varlistentry>
<varlistentry >
<term > <varname > cdma.mtu</varname> </term>
</varlistentry>
2017-11-02 09:25:40 +01:00
<varlistentry >
<term > <varname > connection.auth-retries</varname> </term>
<listitem > <para > If left unspecified, the default value is 3 tries before failing the connection.
</para> </listitem>
</varlistentry>
2016-01-27 13:30:25 +01:00
<varlistentry >
<term > <varname > connection.autoconnect-slaves</varname> </term>
</varlistentry>
2020-04-25 07:46:56 +02:00
<varlistentry >
<term > <varname > connection.mud-url</varname> </term>
<listitem > <para > If unspecified, MUD URL defaults to <literal > "none"</literal> .</para> </listitem>
</varlistentry>
2016-01-27 13:30:25 +01:00
<varlistentry >
<term > <varname > connection.lldp</varname> </term>
</varlistentry>
2018-08-31 15:14:39 +02:00
<varlistentry >
<term > <varname > connection.llmnr</varname> </term>
2019-12-10 16:39:44 +01:00
<listitem > <para > If unspecified, the ultimate default values depends on the DNS plugin. With systemd-resolved the default currently is "yes" (2) and for all other plugins "no" (0).</para> </listitem>
2018-08-31 15:14:39 +02:00
</varlistentry>
core: implement setting MDNS setting for systemd
The connection.mdns setting is a per-connection setting,
so one might expect that one activated device can only have
one MDNS setting at a time.
However, with certain VPN plugins (those that don't have their
own IP interface, like libreswan), the VPN configuration is merged
into the configuration of the device. So, in this case, there
might be multiple settings for one device that must be merged.
We already have a mechanism for that. It's NMIP4Config. Let NMIP4Config
track this piece of information. Although, stricitly speaking this
is not tied to IPv4, the alternative would be to introduce a new
object to track such data, which would be a tremendous effort
and more complicated then this.
Luckily, NMDnsManager and NMDnsPlugin are already equipped to
handle multiple NMIPConfig instances per device (IPv4 vs. IPv6,
and Device vs. VPN).
Also make "connection.mdns" configurable via global defaults in
NetworkManager.conf.
2017-12-20 14:49:32 +01:00
<varlistentry >
<term > <varname > connection.mdns</varname> </term>
2019-12-10 16:39:44 +01:00
<listitem > <para > If unspecified, the ultimate default values depends on the DNS plugin. With systemd-resolved the default currently is "no" (0) and for all other plugins also "no" (0).</para> </listitem>
core: implement setting MDNS setting for systemd
The connection.mdns setting is a per-connection setting,
so one might expect that one activated device can only have
one MDNS setting at a time.
However, with certain VPN plugins (those that don't have their
own IP interface, like libreswan), the VPN configuration is merged
into the configuration of the device. So, in this case, there
might be multiple settings for one device that must be merged.
We already have a mechanism for that. It's NMIP4Config. Let NMIP4Config
track this piece of information. Although, stricitly speaking this
is not tied to IPv4, the alternative would be to introduce a new
object to track such data, which would be a tremendous effort
and more complicated then this.
Luckily, NMDnsManager and NMDnsPlugin are already equipped to
handle multiple NMIPConfig instances per device (IPv4 vs. IPv6,
and Device vs. VPN).
Also make "connection.mdns" configurable via global defaults in
NetworkManager.conf.
2017-12-20 14:49:32 +01:00
</varlistentry>
2016-12-18 13:54:26 +01:00
<varlistentry >
<term > <varname > connection.stable-id</varname> </term>
</varlistentry>
device: extend MAC address handling including randomization for ethernet and wifi
Extend the "ethernet.cloned-mac-address" and "wifi.cloned-mac-address"
settings. Instead of specifying an explicit MAC address, the additional
special values "permanent", "preserve", "random", "random-bia", "stable" and
"stable-bia" are supported.
"permanent" means to use the permanent hardware address. Previously that
was the default if no explict cloned-mac-address was set. The default is
thus still "permanent", but it can be overwritten by global
configuration.
"preserve" means not to configure the MAC address when activating the
device. That was actually the default behavior before introducing MAC
address handling with commit 1b49f941a69af910b0e68530be7339e8053068e5.
"random" and "random-bia" use a randomized MAC address for each
connection. "stable" and "stable-bia" use a generated, stable
address based on some token. The "bia" suffix says to generate a
burned-in address. The stable method by default uses as token the
connection UUID, but the token can be explicitly choosen via
"stable:<TOKEN>" and "stable-bia:<TOKEN>".
On a D-Bus level, the "cloned-mac-address" is a bytestring and thus
cannot express the new forms. It is replaced by the new
"assigned-mac-address" field. For the GObject property, libnm's API,
nmcli, keyfile, etc. the old name "cloned-mac-address" is still used.
Deprecating the old field seems more complicated then just extending
the use of the existing "cloned-mac-address" field, although the name
doesn't match well with the extended meaning.
There is some overlap with the "wifi.mac-address-randomization" setting.
https://bugzilla.gnome.org/show_bug.cgi?id=705545
https://bugzilla.gnome.org/show_bug.cgi?id=708820
https://bugzilla.gnome.org/show_bug.cgi?id=758301
2016-05-24 15:57:16 +02:00
<varlistentry >
<term > <varname > ethernet.cloned-mac-address</varname> </term>
2016-09-01 16:18:34 +02:00
<listitem > <para > If left unspecified, it defaults to "preserve".</para> </listitem>
device: extend MAC address handling including randomization for ethernet and wifi
Extend the "ethernet.cloned-mac-address" and "wifi.cloned-mac-address"
settings. Instead of specifying an explicit MAC address, the additional
special values "permanent", "preserve", "random", "random-bia", "stable" and
"stable-bia" are supported.
"permanent" means to use the permanent hardware address. Previously that
was the default if no explict cloned-mac-address was set. The default is
thus still "permanent", but it can be overwritten by global
configuration.
"preserve" means not to configure the MAC address when activating the
device. That was actually the default behavior before introducing MAC
address handling with commit 1b49f941a69af910b0e68530be7339e8053068e5.
"random" and "random-bia" use a randomized MAC address for each
connection. "stable" and "stable-bia" use a generated, stable
address based on some token. The "bia" suffix says to generate a
burned-in address. The stable method by default uses as token the
connection UUID, but the token can be explicitly choosen via
"stable:<TOKEN>" and "stable-bia:<TOKEN>".
On a D-Bus level, the "cloned-mac-address" is a bytestring and thus
cannot express the new forms. It is replaced by the new
"assigned-mac-address" field. For the GObject property, libnm's API,
nmcli, keyfile, etc. the old name "cloned-mac-address" is still used.
Deprecating the old field seems more complicated then just extending
the use of the existing "cloned-mac-address" field, although the name
doesn't match well with the extended meaning.
There is some overlap with the "wifi.mac-address-randomization" setting.
https://bugzilla.gnome.org/show_bug.cgi?id=705545
https://bugzilla.gnome.org/show_bug.cgi?id=708820
https://bugzilla.gnome.org/show_bug.cgi?id=758301
2016-05-24 15:57:16 +02:00
</varlistentry>
all: make MAC address randomization algorithm configurable
For the per-connection settings "ethernet.cloned-mac-address"
and "wifi.cloned-mac-address", and for the per-device setting
"wifi.scan-rand-mac-address", we may generate MAC addresses using
either the "random" or "stable" algorithm.
Add new properties "generate-mac-address-mask" that allow to configure
which bits of the MAC address will be scrambled.
By default, the "random" and "stable" algorithms scamble all bits
of the MAC address, including the OUI part and generate a locally-
administered, unicast address.
By specifying a MAC address mask, we can now configure to perserve
parts of the current MAC address of the device. For example, setting
"FF:FF:FF:00:00:00" will preserve the first 3 octects of the current
MAC address.
One can also explicitly specify a MAC address to use instead of the
current MAC address. For example, "FF:FF:FF:00:00:00 68:F7:28:00:00:00"
sets the OUI part of the MAC address to "68:F7:28" while scrambling
the last 3 octects.
Similarly, "02:00:00:00:00:00 00:00:00:00:00:00" will scamble
all bits of the MAC address, except clearing the second-least
significant bit. Thus, creating a burned-in address, globally
administered.
One can also supply a list of MAC addresses like
"FF:FF:FF:00:00:00 68:F7:28:00:00:00 00:0C:29:00:00:00 ..." in which
case a MAC address is choosen randomly.
To fully scamble the MAC address one can configure
"02:00:00:00:00:00 00:00:00:00:00:00 02:00:00:00:00:00".
which also randomly creates either a locally or globally administered
address.
With this, the following macchanger options can be implemented:
`macchanger --random`
This is the default if no mask is configured.
-> ""
while is the same as:
-> "00:00:00:00:00:00"
-> "02:00:00:00:00:00 02:00:00:00:00:00"
`macchanger --random --bia`
-> "02:00:00:00:00:00 00:00:00:00:00:00"
`macchanger --ending`
This option cannot be fully implemented, because macchanger
uses the current MAC address but also implies --bia.
-> "FF:FF:FF:00:00:00"
This would yields the same result only if the current MAC address
is already a burned-in address too. Otherwise, it has not the same
effect as --ending.
-> "FF:FF:FF:00:00:00 <MAC_ADDR>"
Alternatively, instead of using the current MAC address,
spell the OUI part out. But again, that is not really the
same as macchanger does because you explictly have to name
the OUI part to use.
`machanger --another`
`machanger --another_any`
-> "FF:FF:FF:00:00:00 <MAC_ADDR> <MAC_ADDR> ..."
"$(printf "FF:FF:FF:00:00:00 %s\n" "$(sed -n 's/^\([0-9a-fA-F][0-9a-fA-F]\) \([0-9a-fA-F][0-9a-fA-F]\) \([0-9a-fA-F][0-9a-fA-F]\) .*/\1:\2:\3:00:00:00/p' /usr/share/macchanger/wireless.list | xargs)")"
2016-06-22 20:31:39 +02:00
<varlistentry >
<term > <varname > ethernet.generate-mac-address-mask</varname> </term>
</varlistentry>
2017-01-17 13:11:07 +01:00
<varlistentry >
<term > <varname > ethernet.mtu</varname> </term>
device: don't set MTU of device unless explicitly configured
Since commit 2b51d3967 "device: merge branch 'th/device-mtu-bgo777251'",
we always set the MTU for certain device types during activation. Even
if the MTU is neither specified via the connection nor other means, like
DHCP.
Revert that change. On activation, if nothing explicitly configures the
MTU, leave it unchanged. This is like what we do with ethernet's
cloned-mac-address, which has a default value "preserve".
So, as last resort the default value for MTU is now 0 (don't change),
instead of depending on the device type.
Note that you also can override the default value in global
configuration via NetworkManager.conf.
This behavior makes sense, because whenever NM actively resets the MTU,
it remembers the previous value and restores it when deactivating
the connection. That wasn't implemented before 2b51d3967, and the
MTU would depend on which connection was previously active. That
is no longer an issue as the MTU gets reset when deactivating.
https://bugzilla.redhat.com/show_bug.cgi?id=1460760
2017-06-13 12:26:51 +02:00
<listitem > <para > If configured explicitly to 0, the MTU is not reconfigured during device activation unless it is required due to IPv6 constraints. If left unspecified, a DHCP/IPv6 SLAAC provided value is used or the MTU is not reconfigured during activation.</para> </listitem>
2017-01-17 13:11:07 +01:00
</varlistentry>
2016-01-27 13:30:25 +01:00
<varlistentry >
<term > <varname > ethernet.wake-on-lan</varname> </term>
</varlistentry>
2018-11-14 16:35:53 +01:00
<varlistentry >
<term > <varname > gsm.mtu</varname> </term>
</varlistentry>
2017-01-17 13:11:07 +01:00
<varlistentry >
<term > <varname > infiniband.mtu</varname> </term>
<listitem > <para > If configured explicitly to 0, the MTU is not reconfigured during device activation unless it is required due to IPv6 constraints. If left unspecified, a DHCP/IPv6 SLAAC provided value is used or the MTU is left unspecified on activation.</para> </listitem>
</varlistentry>
<varlistentry >
<term > <varname > ip-tunnel.mtu</varname> </term>
<listitem > <para > If configured explicitly to 0, the MTU is not reconfigured during device activation unless it is required due to IPv6 constraints. If left unspecified, a DHCP/IPv6 SLAAC provided value is used or a default of 1500.</para> </listitem>
</varlistentry>
2016-01-27 13:30:25 +01:00
<varlistentry >
<term > <varname > ipv4.dad-timeout</varname> </term>
</varlistentry>
2018-02-14 15:45:29 +01:00
<varlistentry >
<term > <varname > ipv4.dhcp-client-id</varname> </term>
</varlistentry>
2019-10-15 14:45:47 +02:00
<varlistentry >
<term > <varname > ipv4.dhcp-iaid</varname> </term>
<listitem > <para > If left unspecified, it defaults to "ifname".</para> </listitem>
</varlistentry>
2019-06-25 20:57:15 +02:00
<varlistentry >
<term > <varname > ipv4.dhcp-hostname-flags</varname> </term>
<listitem > <para > If left unspecified, the value 3 (fqdn-encoded,fqdn-serv-update) is used.</para> </listitem>
</varlistentry>
2016-01-27 13:30:25 +01:00
<varlistentry >
<term > <varname > ipv4.dhcp-timeout</varname> </term>
<listitem > <para > If left unspecified, the default value for
the interface type is used.</para> </listitem>
</varlistentry>
2018-11-13 12:49:37 +01:00
<varlistentry >
<term > <varname > ipv4.dns-priority</varname> </term>
<listitem > <para > If unspecified or zero, use 50 for VPN profiles
and 100 for other profiles.</para> </listitem>
</varlistentry>
2016-01-27 13:30:25 +01:00
<varlistentry >
<term > <varname > ipv4.route-metric</varname> </term>
</varlistentry>
2017-09-21 19:03:08 +02:00
<varlistentry >
all: rework configuring route table support by adding "route-table" setting
We added "ipv4.route-table-sync" and "ipv6.route-table-sync" to not change
behavior for users that configured policy routing outside of NetworkManager,
for example, via a dispatcher script. Users had to explicitly opt-in
for NetworkManager to fully manage all routing tables.
These settings were awkward. Replace them with new settings "ipv4.route-table"
and "ipv6.route-table". Note that this commit breaks API/ABI on the unstable
development branch by removing recently added API.
As before, a connection will have no route-table set by default. This
has the meaning that policy-routing is not enabled and only the main table
will be fully synced. Once the user sets a table, we recognize that and
NetworkManager manages all routing tables.
The new route-table setting has other important uses: analog to
"ipv4.route-metric", it is the default that applies to all routes.
Currently it only works for static routes, not DHCP, SLAAC,
default-route, etc. That will be implemented later.
For static routes, each route still can explicitly set a table, and
overwrite the per-connection setting in "ipv4.route-table" and
"ipv6.route-table".
2017-09-28 08:40:41 +02:00
<term > <varname > ipv4.route-table</varname> </term>
<listitem > <para > If left unspecified, routes are only added to the main table. Note that this
is different from explicitly selecting the main table 254, because of how NetworkManager
removes extraneous routes from the tables.
</para> </listitem>
2017-09-21 19:03:08 +02:00
</varlistentry>
2020-02-14 16:58:55 +01:00
<varlistentry >
<term > <varname > ipv6.ra-timeout</varname> </term>
<listitem > <para > If left unspecified, the default value depends on the sysctl solicitation settings.</para> </listitem>
</varlistentry>
2017-09-11 10:43:48 +02:00
<varlistentry >
2018-03-06 16:10:01 +01:00
<term > <varname > ipv6.dhcp-duid</varname> </term>
<listitem > <para > If left unspecified, it defaults to "lease".</para> </listitem>
</varlistentry>
2019-10-15 14:45:47 +02:00
<varlistentry >
<term > <varname > ipv6.dhcp-iaid</varname> </term>
<listitem > <para > If left unspecified, it defaults to "ifname".</para> </listitem>
</varlistentry>
2019-06-25 20:57:15 +02:00
<varlistentry >
<term > <varname > ipv6.dhcp-hostname-flags</varname> </term>
<listitem > <para > If left unspecified, the value 1 (fqdn-serv-update) is used.</para> </listitem>
</varlistentry>
2018-03-06 16:10:01 +01:00
<varlistentry >
2017-09-11 10:43:48 +02:00
<term > <varname > ipv6.dhcp-timeout</varname> </term>
<listitem > <para > If left unspecified, the default value for
the interface type is used.</para> </listitem>
</varlistentry>
2018-11-13 12:49:37 +01:00
<varlistentry >
<term > <varname > ipv6.dns-priority</varname> </term>
<listitem > <para > If unspecified or zero, use 50 for VPN profiles
and 100 for other profiles.</para> </listitem>
</varlistentry>
2016-01-27 13:30:25 +01:00
<varlistentry >
<term > <varname > ipv6.ip6-privacy</varname> </term>
<listitem > <para > If <literal > ipv6.ip6-privacy</literal> is unset, use the content of
"/proc/sys/net/ipv6/conf/default/use_tempaddr" as last fallback.
</para> </listitem>
</varlistentry>
<varlistentry >
<term > <varname > ipv6.route-metric</varname> </term>
</varlistentry>
2017-09-21 19:03:08 +02:00
<varlistentry >
all: rework configuring route table support by adding "route-table" setting
We added "ipv4.route-table-sync" and "ipv6.route-table-sync" to not change
behavior for users that configured policy routing outside of NetworkManager,
for example, via a dispatcher script. Users had to explicitly opt-in
for NetworkManager to fully manage all routing tables.
These settings were awkward. Replace them with new settings "ipv4.route-table"
and "ipv6.route-table". Note that this commit breaks API/ABI on the unstable
development branch by removing recently added API.
As before, a connection will have no route-table set by default. This
has the meaning that policy-routing is not enabled and only the main table
will be fully synced. Once the user sets a table, we recognize that and
NetworkManager manages all routing tables.
The new route-table setting has other important uses: analog to
"ipv4.route-metric", it is the default that applies to all routes.
Currently it only works for static routes, not DHCP, SLAAC,
default-route, etc. That will be implemented later.
For static routes, each route still can explicitly set a table, and
overwrite the per-connection setting in "ipv4.route-table" and
"ipv6.route-table".
2017-09-28 08:40:41 +02:00
<term > <varname > ipv6.route-table</varname> </term>
<listitem > <para > If left unspecified, routes are only added to the main table. Note that this
is different from explicitly selecting the main table 254, because of how NetworkManager
removes extraneous routes from the tables.
</para> </listitem>
2017-09-21 19:03:08 +02:00
</varlistentry>
2018-05-22 10:48:11 +02:00
<varlistentry >
<term > <varname > sriov.autoprobe-drivers</varname> </term>
<listitem > <para > If left unspecified, drivers are autoprobed when the SR-IOV VF gets created.</para> </listitem>
</varlistentry>
2016-02-11 17:19:28 +01:00
<varlistentry >
<term > <varname > vpn.timeout</varname> </term>
<listitem > <para > If left unspecified, default value of 60 seconds is used.</para> </listitem>
</varlistentry>
2020-06-29 17:33:12 +02:00
<varlistentry >
<term > <varname > wifi.ap-isolation</varname> </term>
<listitem > <para > If left unspecified, AP isolation is disabled.</para> </listitem>
</varlistentry>
device: extend MAC address handling including randomization for ethernet and wifi
Extend the "ethernet.cloned-mac-address" and "wifi.cloned-mac-address"
settings. Instead of specifying an explicit MAC address, the additional
special values "permanent", "preserve", "random", "random-bia", "stable" and
"stable-bia" are supported.
"permanent" means to use the permanent hardware address. Previously that
was the default if no explict cloned-mac-address was set. The default is
thus still "permanent", but it can be overwritten by global
configuration.
"preserve" means not to configure the MAC address when activating the
device. That was actually the default behavior before introducing MAC
address handling with commit 1b49f941a69af910b0e68530be7339e8053068e5.
"random" and "random-bia" use a randomized MAC address for each
connection. "stable" and "stable-bia" use a generated, stable
address based on some token. The "bia" suffix says to generate a
burned-in address. The stable method by default uses as token the
connection UUID, but the token can be explicitly choosen via
"stable:<TOKEN>" and "stable-bia:<TOKEN>".
On a D-Bus level, the "cloned-mac-address" is a bytestring and thus
cannot express the new forms. It is replaced by the new
"assigned-mac-address" field. For the GObject property, libnm's API,
nmcli, keyfile, etc. the old name "cloned-mac-address" is still used.
Deprecating the old field seems more complicated then just extending
the use of the existing "cloned-mac-address" field, although the name
doesn't match well with the extended meaning.
There is some overlap with the "wifi.mac-address-randomization" setting.
https://bugzilla.gnome.org/show_bug.cgi?id=705545
https://bugzilla.gnome.org/show_bug.cgi?id=708820
https://bugzilla.gnome.org/show_bug.cgi?id=758301
2016-05-24 15:57:16 +02:00
<varlistentry >
<term > <varname > wifi.cloned-mac-address</varname> </term>
2016-09-01 16:18:34 +02:00
<listitem > <para > If left unspecified, it defaults to "preserve".</para> </listitem>
device: extend MAC address handling including randomization for ethernet and wifi
Extend the "ethernet.cloned-mac-address" and "wifi.cloned-mac-address"
settings. Instead of specifying an explicit MAC address, the additional
special values "permanent", "preserve", "random", "random-bia", "stable" and
"stable-bia" are supported.
"permanent" means to use the permanent hardware address. Previously that
was the default if no explict cloned-mac-address was set. The default is
thus still "permanent", but it can be overwritten by global
configuration.
"preserve" means not to configure the MAC address when activating the
device. That was actually the default behavior before introducing MAC
address handling with commit 1b49f941a69af910b0e68530be7339e8053068e5.
"random" and "random-bia" use a randomized MAC address for each
connection. "stable" and "stable-bia" use a generated, stable
address based on some token. The "bia" suffix says to generate a
burned-in address. The stable method by default uses as token the
connection UUID, but the token can be explicitly choosen via
"stable:<TOKEN>" and "stable-bia:<TOKEN>".
On a D-Bus level, the "cloned-mac-address" is a bytestring and thus
cannot express the new forms. It is replaced by the new
"assigned-mac-address" field. For the GObject property, libnm's API,
nmcli, keyfile, etc. the old name "cloned-mac-address" is still used.
Deprecating the old field seems more complicated then just extending
the use of the existing "cloned-mac-address" field, although the name
doesn't match well with the extended meaning.
There is some overlap with the "wifi.mac-address-randomization" setting.
https://bugzilla.gnome.org/show_bug.cgi?id=705545
https://bugzilla.gnome.org/show_bug.cgi?id=708820
https://bugzilla.gnome.org/show_bug.cgi?id=758301
2016-05-24 15:57:16 +02:00
</varlistentry>
all: make MAC address randomization algorithm configurable
For the per-connection settings "ethernet.cloned-mac-address"
and "wifi.cloned-mac-address", and for the per-device setting
"wifi.scan-rand-mac-address", we may generate MAC addresses using
either the "random" or "stable" algorithm.
Add new properties "generate-mac-address-mask" that allow to configure
which bits of the MAC address will be scrambled.
By default, the "random" and "stable" algorithms scamble all bits
of the MAC address, including the OUI part and generate a locally-
administered, unicast address.
By specifying a MAC address mask, we can now configure to perserve
parts of the current MAC address of the device. For example, setting
"FF:FF:FF:00:00:00" will preserve the first 3 octects of the current
MAC address.
One can also explicitly specify a MAC address to use instead of the
current MAC address. For example, "FF:FF:FF:00:00:00 68:F7:28:00:00:00"
sets the OUI part of the MAC address to "68:F7:28" while scrambling
the last 3 octects.
Similarly, "02:00:00:00:00:00 00:00:00:00:00:00" will scamble
all bits of the MAC address, except clearing the second-least
significant bit. Thus, creating a burned-in address, globally
administered.
One can also supply a list of MAC addresses like
"FF:FF:FF:00:00:00 68:F7:28:00:00:00 00:0C:29:00:00:00 ..." in which
case a MAC address is choosen randomly.
To fully scamble the MAC address one can configure
"02:00:00:00:00:00 00:00:00:00:00:00 02:00:00:00:00:00".
which also randomly creates either a locally or globally administered
address.
With this, the following macchanger options can be implemented:
`macchanger --random`
This is the default if no mask is configured.
-> ""
while is the same as:
-> "00:00:00:00:00:00"
-> "02:00:00:00:00:00 02:00:00:00:00:00"
`macchanger --random --bia`
-> "02:00:00:00:00:00 00:00:00:00:00:00"
`macchanger --ending`
This option cannot be fully implemented, because macchanger
uses the current MAC address but also implies --bia.
-> "FF:FF:FF:00:00:00"
This would yields the same result only if the current MAC address
is already a burned-in address too. Otherwise, it has not the same
effect as --ending.
-> "FF:FF:FF:00:00:00 <MAC_ADDR>"
Alternatively, instead of using the current MAC address,
spell the OUI part out. But again, that is not really the
same as macchanger does because you explictly have to name
the OUI part to use.
`machanger --another`
`machanger --another_any`
-> "FF:FF:FF:00:00:00 <MAC_ADDR> <MAC_ADDR> ..."
"$(printf "FF:FF:FF:00:00:00 %s\n" "$(sed -n 's/^\([0-9a-fA-F][0-9a-fA-F]\) \([0-9a-fA-F][0-9a-fA-F]\) \([0-9a-fA-F][0-9a-fA-F]\) .*/\1:\2:\3:00:00:00/p' /usr/share/macchanger/wireless.list | xargs)")"
2016-06-22 20:31:39 +02:00
<varlistentry >
<term > <varname > wifi.generate-mac-address-mask</varname> </term>
</varlistentry>
2016-02-11 17:19:28 +01:00
<varlistentry >
<term > <varname > wifi.mac-address-randomization</varname> </term>
2016-06-20 12:28:04 +02:00
<listitem > <para > If left unspecified, MAC address randomization is disabled.
This setting is deprecated for <literal > wifi.cloned-mac-address</literal> .
</para> </listitem>
2016-02-11 17:19:28 +01:00
</varlistentry>
2017-01-17 13:11:07 +01:00
<varlistentry >
<term > <varname > wifi.mtu</varname> </term>
<listitem > <para > If configured explicitly to 0, the MTU is not reconfigured during device activation unless it is required due to IPv6 constraints. If left unspecified, a DHCP/IPv6 SLAAC provided value is used or a default of 1500.</para> </listitem>
</varlistentry>
2016-02-10 16:46:34 +01:00
<varlistentry >
<term > <varname > wifi.powersave</varname> </term>
<listitem > <para > If left unspecified, the default value
"<literal > ignore</literal> " will be used.</para> </listitem>
</varlistentry>
2017-04-18 13:39:43 +02:00
<varlistentry >
<term > <varname > wifi-sec.pmf</varname> </term>
<listitem > <para > If left unspecified, the default value
"<literal > optional</literal> " will be used.</para> </listitem>
</varlistentry>
2018-01-16 06:28:33 +09:00
<varlistentry >
<term > <varname > wifi-sec.fils</varname> </term>
<listitem > <para > If left unspecified, the default value
"<literal > optional</literal> " will be used.</para> </listitem>
</varlistentry>
2018-11-14 16:35:53 +01:00
<varlistentry >
<term > <varname > wifi.wake-on-wlan</varname> </term>
</varlistentry>
2019-03-02 23:33:15 +01:00
<varlistentry >
<term > <varname > wireguard.mtu</varname> </term>
</varlistentry>
2016-01-27 13:30:25 +01:00
</variablelist>
2018-11-14 14:52:21 +01:00
<!-- The following comment is used by check - config - options.sh, don't remove it. -->
<!-- end connection defaults -->
2016-01-27 13:30:25 +01:00
</para>
</refsect2>
2016-08-26 14:58:19 +02:00
<refsect2 id= "connection-sections" xreflabel= "“Sections” under the section called “CONNECTION SECTION”" >
2016-01-27 13:30:25 +01:00
<title > Sections</title>
<para >
You can configure multiple <literal > connection</literal>
sections, by having different sections with a name that all start
with "connection".
2015-05-15 11:36:28 +02:00
Example:
<programlisting >
[connection]
ipv6.ip6-privacy=0
2015-06-09 16:05:10 +02:00
connection.autoconnect-slaves=1
2015-09-17 12:39:56 +02:00
vpn.timeout=120
2015-05-15 11:36:28 +02:00
[connection-wifi-wlan0]
match-device=interface-name:wlan0
ipv4.route-metric=50
[connection-wifi-other]
match-device=type:wifi
ipv4.route-metric=55
ipv6.ip6-privacy=1
</programlisting>
</para>
<para >
2015-06-25 13:02:10 +02:00
The sections within one file are considered in order of appearance, with the
2015-05-15 11:36:28 +02:00
exception that the <literal > [connection]</literal> section is always
considered last. In the example above, this order is <literal > [connection-wifi-wlan0]</literal> ,
<literal > [connection-wlan-other]</literal> , and <literal > [connection]</literal> .
2016-01-27 12:32:07 +01:00
When checking for a default configuration value, the sections are searched until
2015-05-15 11:36:28 +02:00
the requested value is found.
In the example above, "ipv4.route-metric" for wlan0 interface is set to 50,
and for all other Wi-Fi typed interfaces to 55. Also, Wi-Fi devices would have
IPv6 private addresses enabled by default, but other devices would have it disabled.
Note that also "wlan0" gets "ipv6.ip6-privacy=1", because although the section
"[connection-wifi-wlan0]" matches the device, it does not contain that property
and the search continues.
</para>
2015-06-25 13:02:10 +02:00
<para >
When having different sections in multiple files, sections from files that are read
later have higher priority. So within one file the priority of the sections is
top-to-bottom. Across multiple files later definitions take precedence.
</para>
2015-05-15 11:36:28 +02:00
<para >
2016-01-27 13:30:25 +01:00
The following properties further control how a connection section applies.
2015-05-15 11:36:28 +02:00
<variablelist >
<varlistentry >
<term > <varname > match-device</varname> </term>
<listitem > <para > An optional device spec that restricts
when the section applies. See <xref linkend= "device-spec" />
for the possible values.
</para> </listitem>
</varlistentry>
<varlistentry >
<term > <varname > stop-match</varname> </term>
<listitem > <para > An optional boolean value which defaults to
<literal > no</literal> . If the section matches (based on
<literal > match-device</literal> ), further sections will not be
considered even if the property in question is not present. In
the example above, if <literal > [connection-wifi-wlan0]</literal> would
have <literal > stop-match</literal> set to <literal > yes</literal> ,
2016-05-13 15:02:55 +02:00
the device <literal > wlan0</literal> would have <literal > ipv6.ip6-privacy</literal>
property unspecified. That is, the search for the property would not continue
in the connection sections <literal > [connection-wifi-other]</literal>
or <literal > [connection]</literal> .
2015-05-15 11:36:28 +02:00
</para> </listitem>
</varlistentry>
</variablelist>
</para>
2016-01-27 13:30:25 +01:00
</refsect2>
2015-05-15 11:36:28 +02:00
</refsect1>
2016-06-22 13:14:10 +02:00
<refsect1 >
<title > <literal > device</literal> section</title>
<para > Contains per-device persistent configuration.
</para>
<para >
Example:
<programlisting >
[device]
match-device=interface-name:eth3
2017-09-26 16:28:49 +02:00
managed=1
2016-06-22 13:14:10 +02:00
</programlisting>
</para>
<refsect2 >
<title > Supported Properties</title>
<para >
The following properties can be configured per-device.
<variablelist >
2017-09-26 16:28:49 +02:00
<varlistentry id= "managed" >
<term > <varname > managed</varname> </term>
<listitem >
<para >
Whether the device is managed or not. A device can be
marked as managed via udev rules (ENV{NM_UNMANAGED}),
or via setting plugins (keyfile.unmanaged-devices).
This is yet another way. Note that this configuration
can be overruled at runtime via D-Bus. Also, it has
higher priority then udev rules.
</para>
</listitem>
</varlistentry>
2017-11-21 15:48:57 +01:00
<varlistentry id= "carrier-wait-timeout" >
<term > <varname > carrier-wait-timeout</varname> </term>
<listitem >
<para >
Specify the timeout for waiting for carrier in milliseconds.
When the device loses carrier, NetworkManager does not react
immediately. Instead, it waits for this timeout before considering
the link lost. Also, on startup, NetworkManager considers the
device as busy for this time, as long as the device has no carrier.
This delays startup-complete signal and NetworkManager-wait-online.
Configuring this too high means to block NetworkManager-wait-online
longer then necessary. Configuring it too low, means that NetworkManager
will declare startup-complete, although carrier is about to come
and auto-activation to kick in.
The default is 5000 milliseconds.
</para>
</listitem>
</varlistentry>
config: make "ignore-carrier" a per-device configuration option
NetworkManager.conf already contains several per-device settings,
that is, settings that have a device-spec as argument.
main.ignore-carrier
main.no-auto-default
main.assume-ipv6ll-only
keyfile.unmanged-devices
Optimally, these settings should be moved to the new [device*]
section.
For now, only move main.ignore-carrier there. For the others
it may not make sense to do so:
- main.no-auto-default: is already merged with internal state
from /var/lib/NetworkManager/no-auto-default.state. While
NMConfig's write API would be fine to also persist and merge
the no-auto-default setting, we'd still have to read the old
file too. Thus, deprecating this setting gets quite cumbersome
to still handle the old state file.
Also, it seems a less useful setting to configure in the
global configuration aside setting main.no-auto-default=*.
- main.assume-ipv6ll-only: one day, I hope that we no longer
assume connections at all, and this setting becomes entirely
obsolete.
- keyfile.unmanged-devices: this sets NM_UNMANAGED_USER_SETTINGS,
which cannot be overruled via D-Bus. For a future device.managed
setting we want it it to be overwritable via D-Bus by an explicit
user action. Thus, a device.managed property should have a different
semantic, this should be more like a device.unmanaged-force setting,
which could be done.
2016-06-22 13:43:08 +02:00
<varlistentry id= "ignore-carrier" >
<term > <varname > ignore-carrier</varname> </term>
<listitem >
<para >
Specify devices for which NetworkManager will (partially)
ignore the carrier state. Normally, for
device types that support carrier-detect, such as Ethernet
and InfiniBand, NetworkManager will only allow a
connection to be activated on the device if carrier is
present (ie, a cable is plugged in), and it will
deactivate the device if carrier drops for more than a few
seconds.
</para>
<para >
A device with carrier ignored will allow activating connections on
that device even when it does not have carrier, provided
that the connection uses only statically-configured IP
addresses. Additionally, it will allow any active
connection (whether static or dynamic) to remain active on
the device when carrier is lost.
</para>
<para >
Note that the "carrier" property of NMDevices and device D-Bus
interfaces will still reflect the actual device state; it's just
that NetworkManager will not make use of that information.
</para>
2017-06-01 23:57:03 +02:00
<para >
Master types like bond, bridge and team ignore carrier by default,
while other device types react on carrier changes by default.
</para>
config: make "ignore-carrier" a per-device configuration option
NetworkManager.conf already contains several per-device settings,
that is, settings that have a device-spec as argument.
main.ignore-carrier
main.no-auto-default
main.assume-ipv6ll-only
keyfile.unmanged-devices
Optimally, these settings should be moved to the new [device*]
section.
For now, only move main.ignore-carrier there. For the others
it may not make sense to do so:
- main.no-auto-default: is already merged with internal state
from /var/lib/NetworkManager/no-auto-default.state. While
NMConfig's write API would be fine to also persist and merge
the no-auto-default setting, we'd still have to read the old
file too. Thus, deprecating this setting gets quite cumbersome
to still handle the old state file.
Also, it seems a less useful setting to configure in the
global configuration aside setting main.no-auto-default=*.
- main.assume-ipv6ll-only: one day, I hope that we no longer
assume connections at all, and this setting becomes entirely
obsolete.
- keyfile.unmanged-devices: this sets NM_UNMANAGED_USER_SETTINGS,
which cannot be overruled via D-Bus. For a future device.managed
setting we want it it to be overwritable via D-Bus by an explicit
user action. Thus, a device.managed property should have a different
semantic, this should be more like a device.unmanaged-force setting,
which could be done.
2016-06-22 13:43:08 +02:00
<para >
This setting overwrites the deprecated <literal > main.ignore-carrier</literal>
setting above.
</para>
</listitem>
</varlistentry>
2016-06-22 14:42:20 +02:00
<varlistentry >
<term > <varname > wifi.scan-rand-mac-address</varname> </term>
<listitem >
<para >
2016-06-30 09:22:12 +02:00
Configures MAC address randomization of a Wi-Fi device during
scanning. This defaults to <literal > yes</literal> in which case
a random, locally-administered MAC address will be used.
The setting <literal > wifi.scan-generate-mac-address-mask</literal>
allows to influence the generated MAC address to use certain vendor
OUIs.
If disabled, the MAC address during scanning is left unchanged to
whatever is configured.
For the configured MAC address while the device is associated, see instead
2016-06-22 14:42:20 +02:00
the per-connection setting <literal > wifi.cloned-mac-address</literal> .
</para>
</listitem>
</varlistentry>
2017-12-13 15:45:47 +01:00
<varlistentry id= "wifi.backend" >
<term > <varname > wifi.backend</varname> </term>
<listitem >
<para >
2020-07-01 17:20:40 -04:00
Specify the Wi-Fi backend used for the device. Currently, supported
2017-12-13 15:45:47 +01:00
are <literal > wpa_supplicant</literal> and <literal > iwd</literal> (experimental).
</para>
</listitem>
</varlistentry>
all: make MAC address randomization algorithm configurable
For the per-connection settings "ethernet.cloned-mac-address"
and "wifi.cloned-mac-address", and for the per-device setting
"wifi.scan-rand-mac-address", we may generate MAC addresses using
either the "random" or "stable" algorithm.
Add new properties "generate-mac-address-mask" that allow to configure
which bits of the MAC address will be scrambled.
By default, the "random" and "stable" algorithms scamble all bits
of the MAC address, including the OUI part and generate a locally-
administered, unicast address.
By specifying a MAC address mask, we can now configure to perserve
parts of the current MAC address of the device. For example, setting
"FF:FF:FF:00:00:00" will preserve the first 3 octects of the current
MAC address.
One can also explicitly specify a MAC address to use instead of the
current MAC address. For example, "FF:FF:FF:00:00:00 68:F7:28:00:00:00"
sets the OUI part of the MAC address to "68:F7:28" while scrambling
the last 3 octects.
Similarly, "02:00:00:00:00:00 00:00:00:00:00:00" will scamble
all bits of the MAC address, except clearing the second-least
significant bit. Thus, creating a burned-in address, globally
administered.
One can also supply a list of MAC addresses like
"FF:FF:FF:00:00:00 68:F7:28:00:00:00 00:0C:29:00:00:00 ..." in which
case a MAC address is choosen randomly.
To fully scamble the MAC address one can configure
"02:00:00:00:00:00 00:00:00:00:00:00 02:00:00:00:00:00".
which also randomly creates either a locally or globally administered
address.
With this, the following macchanger options can be implemented:
`macchanger --random`
This is the default if no mask is configured.
-> ""
while is the same as:
-> "00:00:00:00:00:00"
-> "02:00:00:00:00:00 02:00:00:00:00:00"
`macchanger --random --bia`
-> "02:00:00:00:00:00 00:00:00:00:00:00"
`macchanger --ending`
This option cannot be fully implemented, because macchanger
uses the current MAC address but also implies --bia.
-> "FF:FF:FF:00:00:00"
This would yields the same result only if the current MAC address
is already a burned-in address too. Otherwise, it has not the same
effect as --ending.
-> "FF:FF:FF:00:00:00 <MAC_ADDR>"
Alternatively, instead of using the current MAC address,
spell the OUI part out. But again, that is not really the
same as macchanger does because you explictly have to name
the OUI part to use.
`machanger --another`
`machanger --another_any`
-> "FF:FF:FF:00:00:00 <MAC_ADDR> <MAC_ADDR> ..."
"$(printf "FF:FF:FF:00:00:00 %s\n" "$(sed -n 's/^\([0-9a-fA-F][0-9a-fA-F]\) \([0-9a-fA-F][0-9a-fA-F]\) \([0-9a-fA-F][0-9a-fA-F]\) .*/\1:\2:\3:00:00:00/p' /usr/share/macchanger/wireless.list | xargs)")"
2016-06-22 20:31:39 +02:00
<varlistentry >
<term > <varname > wifi.scan-generate-mac-address-mask</varname> </term>
<listitem >
<para >
2016-06-30 09:22:12 +02:00
Like the per-connection settings <literal > ethernet.generate-mac-address-mask</literal>
and <literal > wifi.generate-mac-address-mask</literal> , this allows to configure the
generated MAC addresses during scanning. See <citerefentry > <refentrytitle > nm-settings</refentrytitle> <manvolnum > 5</manvolnum> </citerefentry>
all: make MAC address randomization algorithm configurable
For the per-connection settings "ethernet.cloned-mac-address"
and "wifi.cloned-mac-address", and for the per-device setting
"wifi.scan-rand-mac-address", we may generate MAC addresses using
either the "random" or "stable" algorithm.
Add new properties "generate-mac-address-mask" that allow to configure
which bits of the MAC address will be scrambled.
By default, the "random" and "stable" algorithms scamble all bits
of the MAC address, including the OUI part and generate a locally-
administered, unicast address.
By specifying a MAC address mask, we can now configure to perserve
parts of the current MAC address of the device. For example, setting
"FF:FF:FF:00:00:00" will preserve the first 3 octects of the current
MAC address.
One can also explicitly specify a MAC address to use instead of the
current MAC address. For example, "FF:FF:FF:00:00:00 68:F7:28:00:00:00"
sets the OUI part of the MAC address to "68:F7:28" while scrambling
the last 3 octects.
Similarly, "02:00:00:00:00:00 00:00:00:00:00:00" will scamble
all bits of the MAC address, except clearing the second-least
significant bit. Thus, creating a burned-in address, globally
administered.
One can also supply a list of MAC addresses like
"FF:FF:FF:00:00:00 68:F7:28:00:00:00 00:0C:29:00:00:00 ..." in which
case a MAC address is choosen randomly.
To fully scamble the MAC address one can configure
"02:00:00:00:00:00 00:00:00:00:00:00 02:00:00:00:00:00".
which also randomly creates either a locally or globally administered
address.
With this, the following macchanger options can be implemented:
`macchanger --random`
This is the default if no mask is configured.
-> ""
while is the same as:
-> "00:00:00:00:00:00"
-> "02:00:00:00:00:00 02:00:00:00:00:00"
`macchanger --random --bia`
-> "02:00:00:00:00:00 00:00:00:00:00:00"
`macchanger --ending`
This option cannot be fully implemented, because macchanger
uses the current MAC address but also implies --bia.
-> "FF:FF:FF:00:00:00"
This would yields the same result only if the current MAC address
is already a burned-in address too. Otherwise, it has not the same
effect as --ending.
-> "FF:FF:FF:00:00:00 <MAC_ADDR>"
Alternatively, instead of using the current MAC address,
spell the OUI part out. But again, that is not really the
same as macchanger does because you explictly have to name
the OUI part to use.
`machanger --another`
`machanger --another_any`
-> "FF:FF:FF:00:00:00 <MAC_ADDR> <MAC_ADDR> ..."
"$(printf "FF:FF:FF:00:00:00 %s\n" "$(sed -n 's/^\([0-9a-fA-F][0-9a-fA-F]\) \([0-9a-fA-F][0-9a-fA-F]\) \([0-9a-fA-F][0-9a-fA-F]\) .*/\1:\2:\3:00:00:00/p' /usr/share/macchanger/wireless.list | xargs)")"
2016-06-22 20:31:39 +02:00
for details.
</para>
</listitem>
</varlistentry>
2017-04-14 23:33:14 +02:00
<varlistentry id= "sriov-num-vfs" >
2020-07-01 17:20:40 -04:00
, <term > <varname > sriov-num-vfs</varname> </term>
2017-04-14 23:33:14 +02:00
<listitem >
<para >
Specify the number of virtual functions (VF) to enable
for a PCI physical device that supports single-root I/O
virtualization (SR-IOV).
</para>
</listitem>
</varlistentry>
2016-06-22 13:14:10 +02:00
</variablelist>
</para>
</refsect2>
<refsect2 >
<title > Sections</title>
<para >
The <literal > [device]</literal> section works the same as the <literal > [connection]</literal> section.
That is, multiple sections that all start with the prefix "device" can be specified.
The settings "match-device" and "stop-match" are available to match a device section
on a device. The order of multiple sections is also top-down within the file and
later files overwrite previous settings. See <xref linkend= "connection-sections" />
for details.
</para>
</refsect2>
</refsect1>
2013-04-09 16:41:00 -04:00
<refsect1 >
<title > <literal > connectivity</literal> section</title>
2019-02-08 16:05:36 +01:00
2013-04-09 16:41:00 -04:00
<para > This section controls NetworkManager's optional connectivity
checking functionality. This allows NetworkManager to detect
whether or not the system can actually access the internet or
whether it is behind a captive portal.</para>
2019-02-08 16:05:36 +01:00
2019-02-05 10:52:51 +01:00
<para > Connectivity checking serves two purposes. For one, it exposes
a connectivity state on D-Bus, which other applications may use. For example,
Gnome's portal helper uses this as signal to show a captive portal login
page.
The other use is that default-route of devices without global connectivity
get a penalty of +20000 to the route-metric. This has the purpose to give a
better default-route to devices that have global connectivity. For example,
when being connected to WWAN and to a Wi-Fi network which is behind a captive
portal, WWAN still gets preferred until login.</para>
2013-04-09 16:41:00 -04:00
2019-02-08 16:05:36 +01:00
<para > Note that your distribution might set <literal > /proc/sys/net/ipv4/conf/*/rp_filter</literal> to
strict filtering. That works badly with per-device connectivity checking,
which uses SO_BINDDEVICE to send requests on all devices. A strict rp_filter
setting will reject any response and the connectivity check on all but the
best route will fail.</para>
2013-04-09 16:41:00 -04:00
<para >
<variablelist >
2019-12-28 15:20:06 +01:00
<varlistentry >
<term > <varname > enabled</varname> </term>
<listitem > <para > Whether connectivity check is enabled.
Note that to enable connectivity check, a valid uri must
also be configured. The value defaults to true, but since
the uri is unset by default, connectivity check may be disabled.
The main purpose of this option is to have a single flag
to disable connectivity check. Note that this setting can
also be set via D-Bus API at runtime. In that case, the value gets
stored in <filename > /var/lib/NetworkManager/NetworkManager-intern.conf</filename>
file.
</para> </listitem>
</varlistentry>
2016-01-27 12:54:18 +01:00
<varlistentry >
<term > <varname > uri</varname> </term>
<listitem > <para > The URI of a web page to periodically
request when connectivity is being checked. This page
should return the header "X-NetworkManager-Status" with a
2018-04-18 14:13:28 +02:00
value of "online". Alternatively, its body content should
2016-01-27 12:54:18 +01:00
be set to "NetworkManager is online". The body content
check can be controlled by the <literal > response</literal>
option. If this option is blank or missing, connectivity
checking is disabled.
</para> </listitem>
</varlistentry>
<varlistentry >
<term > <varname > interval</varname> </term>
<listitem > <para > Specified in seconds; controls how often
connectivity is checked when a network connection exists. If
set to 0 connectivity checking is disabled. If missing, the
default is 300 seconds.</para> </listitem>
</varlistentry>
<varlistentry >
<term > <varname > response</varname> </term>
2019-01-14 18:58:45 +01:00
<listitem > <para > If set, controls what body content
2016-01-27 12:54:18 +01:00
NetworkManager checks for when requesting the URI for
2019-01-14 18:58:45 +01:00
connectivity checking. Note that this only compares
that the HTTP response starts with the specifid text,
it does not compare the exact string. This behavior
might change in the future, so avoid relying on it.
If missing, the response defaults to "NetworkManager is online".
If set to empty, the HTTP server is expected to answer with
status code 204 or send no data.</para> </listitem>
2016-01-27 12:54:18 +01:00
</varlistentry>
2013-04-09 16:41:00 -04:00
</variablelist>
</para>
</refsect1>
2015-06-29 15:59:44 +02:00
<refsect1 >
<title > <literal > global-dns</literal> section</title>
<para > This section specifies global DNS settings that override
connection-specific configuration.</para>
<para >
<variablelist >
2016-01-27 12:54:18 +01:00
<varlistentry >
<term > <varname > searches</varname> </term>
<listitem >
<para >
A list of search domains to be used during hostname lookup.
</para>
</listitem>
</varlistentry>
<varlistentry >
<term > <varname > options</varname> </term>
<listitem >
<para >
2019-01-24 17:19:15 +01:00
A list of options to be passed to the hostname resolver.
2016-01-27 12:54:18 +01:00
</para>
</listitem>
</varlistentry>
2015-06-29 15:59:44 +02:00
</variablelist>
</para>
</refsect1>
<refsect1 >
<title > <literal > global-dns-domain</literal> sections</title>
<para > Sections with a name starting with the "global-dns-domain-"
prefix allow to define global DNS configuration for specific
domains. The part of section name after "global-dns-domain-"
specifies the domain name a section applies to. More specific
domains have the precedence over less specific ones and the
default domain is represented by the wildcard "*". A default
domain section is mandatory.
</para>
<para >
<variablelist >
2016-01-27 12:54:18 +01:00
<varlistentry >
<term > <varname > servers</varname> </term>
<listitem >
<para >
A list of addresses of DNS servers to be used for the given domain.
</para>
</listitem>
</varlistentry>
<varlistentry >
<term > <varname > options</varname> </term>
<listitem >
<para >
A list of domain-specific DNS options. Not used at the moment.
</para>
</listitem>
</varlistentry>
2015-06-29 15:59:44 +02:00
</variablelist>
</para>
</refsect1>
config: allow to enable/disable configuration snippets
Support a new configuration option
[.config]
enable=<ENABLED>
for configuration snippets.
This new [.config] section is only relevant within the snippet itself
and it is not merged into the combined configuration.
Currently only the "enable" key is supported. If the "enable" key is
missing, it obviously defaults to being enabled. It allows snippets
to be skipped from loading. The main configuration "NetworkManager.conf"
cannot be skipped.
<ENABLED> can be a boolean value (false), to skip a configuration
snippet from loading.
It can also be a string to match against the NetworkManager version,
like "enable=nm-version-min:1.1,nm-version-min:1.0.6"
There are several motivations for this:
- the user can disable an entire configuration snippet by toggeling
one entry.
This generalizes the functionality of the global-dns.enable
setting, but in a way that applies to configuration on a per-file
basis.
- for developing, we often switch between different versions of
NetworkManager. Thus, we might want to use different configuration.
E.g. before global-dns options, I want to use "dns=none" and manage
resolv.conf myself. Now, I can use global-dns setting to do that.
That can be achieved with something like the following (not exactly,
it's an example only):
[.config]
enable=nm-version-min:1.1
[main]
dns=default
[global-dns-domain-*]
nameserver=127.0.0.1
Arguably, this would be more awesome, if we would bump our micro devel
version (1.1.0) more often while developing 1.2.0 (*hint*).
- in principle, packages could drop configuration snippets and enable
them based on the NetworkManager version.
- with the "env:" spec, you can enable/disable snippets by configuring
an environment variable. Again, useful for testing and developing.
2015-10-01 10:43:33 +02:00
<refsect1 >
<title > <literal > .config</literal> sections</title>
<para > This is a special section that contains options which apply
to the configuration file that contains the option.
</para>
<para >
<variablelist >
<varlistentry >
<term > <varname > enable</varname> </term>
<listitem >
<para >
Defaults to "<literal > true</literal> ". If "<literal > false</literal> ",
the configuration file will be skipped during loading.
2015-12-24 09:48:20 +01:00
Note that the main configuration file <literal > NetworkManager.conf</literal>
config: allow to enable/disable configuration snippets
Support a new configuration option
[.config]
enable=<ENABLED>
for configuration snippets.
This new [.config] section is only relevant within the snippet itself
and it is not merged into the combined configuration.
Currently only the "enable" key is supported. If the "enable" key is
missing, it obviously defaults to being enabled. It allows snippets
to be skipped from loading. The main configuration "NetworkManager.conf"
cannot be skipped.
<ENABLED> can be a boolean value (false), to skip a configuration
snippet from loading.
It can also be a string to match against the NetworkManager version,
like "enable=nm-version-min:1.1,nm-version-min:1.0.6"
There are several motivations for this:
- the user can disable an entire configuration snippet by toggeling
one entry.
This generalizes the functionality of the global-dns.enable
setting, but in a way that applies to configuration on a per-file
basis.
- for developing, we often switch between different versions of
NetworkManager. Thus, we might want to use different configuration.
E.g. before global-dns options, I want to use "dns=none" and manage
resolv.conf myself. Now, I can use global-dns setting to do that.
That can be achieved with something like the following (not exactly,
it's an example only):
[.config]
enable=nm-version-min:1.1
[main]
dns=default
[global-dns-domain-*]
nameserver=127.0.0.1
Arguably, this would be more awesome, if we would bump our micro devel
version (1.1.0) more often while developing 1.2.0 (*hint*).
- in principle, packages could drop configuration snippets and enable
them based on the NetworkManager version.
- with the "env:" spec, you can enable/disable snippets by configuring
an environment variable. Again, useful for testing and developing.
2015-10-01 10:43:33 +02:00
cannot be disabled.
<programlisting >
# always skip loading the config file
[.config]
enable=false
</programlisting>
</para>
<para >
You can also match against the version of NetworkManager. For example
the following are valid configurations:
<programlisting >
# only load on version 1.0.6
[.config]
enable=nm-version:1.0.6
# load on all versions 1.0.x, but not 1.2.x
[.config]
enable=nm-version:1.0
# only load on versions > = 1.1.6. This does not match
# with version 1.2.0 or 1.4.4. Only the last digit is considered.
[.config]
enable=nm-version-min:1.1.6
# only load on versions > = 1.2. Contrary to the previous
# example, this also matches with 1.2.0, 1.2.10, 1.4.4, etc.
[.config]
enable=nm-version-min:1.2
# Match against the maximum allowed version. The example matches
# versions 1.2.0, 1.2.2, 1.2.4. Again, only the last version digit
2020-07-04 11:37:01 +03:00
# is allowed to be smaller. So this would not match on 1.1.10.
config: allow to enable/disable configuration snippets
Support a new configuration option
[.config]
enable=<ENABLED>
for configuration snippets.
This new [.config] section is only relevant within the snippet itself
and it is not merged into the combined configuration.
Currently only the "enable" key is supported. If the "enable" key is
missing, it obviously defaults to being enabled. It allows snippets
to be skipped from loading. The main configuration "NetworkManager.conf"
cannot be skipped.
<ENABLED> can be a boolean value (false), to skip a configuration
snippet from loading.
It can also be a string to match against the NetworkManager version,
like "enable=nm-version-min:1.1,nm-version-min:1.0.6"
There are several motivations for this:
- the user can disable an entire configuration snippet by toggeling
one entry.
This generalizes the functionality of the global-dns.enable
setting, but in a way that applies to configuration on a per-file
basis.
- for developing, we often switch between different versions of
NetworkManager. Thus, we might want to use different configuration.
E.g. before global-dns options, I want to use "dns=none" and manage
resolv.conf myself. Now, I can use global-dns setting to do that.
That can be achieved with something like the following (not exactly,
it's an example only):
[.config]
enable=nm-version-min:1.1
[main]
dns=default
[global-dns-domain-*]
nameserver=127.0.0.1
Arguably, this would be more awesome, if we would bump our micro devel
version (1.1.0) more often while developing 1.2.0 (*hint*).
- in principle, packages could drop configuration snippets and enable
them based on the NetworkManager version.
- with the "env:" spec, you can enable/disable snippets by configuring
an environment variable. Again, useful for testing and developing.
2015-10-01 10:43:33 +02:00
[.config]
enable=nm-version-max:1.2.6
</programlisting>
</para>
<para >
You can also match against the value of the environment variable
<literal > NM_CONFIG_ENABLE_TAG</literal> , like:
<programlisting >
# always skip loading the file when running NetworkManager with
# environment variable "NM_CONFIG_ENABLE_TAG=TAG1"
[.config]
enable=env:TAG1
</programlisting>
</para>
<para >
More then one match can be specified. The configuration will be
enabled if one of the predicates matches ("or"). The special prefix "except:" can
be used to negate the match. Note that if one except-predicate
matches, the entire configuration will be disabled.
In other words, a except predicate always wins over other predicates.
core: fix match spec behavior for a list of all "except:"
If the spec specifies only negative matches (and none of them matches),
then the result shall be positive.
Meaning:
[connection*] match-device=except:dhcp-plugin:dhclient
[connection*] match-device=except:interface-name:eth0
[.config] enabled=except:nm-version:1.14
should be the same as:
[connection*] match-device=*,except:dhcp-plugin:dhclient
[connection*] match-device=*,except:interface-name:eth0
[.config] enabled=*,except:nm-version:1.14
and match by default. Previously, such specs would never yield a
positive match, which seems wrong.
Note that "except:" already has a special meaning. It is not merely
"not:". That is because we don't support "and:" nor grouping, but all
matches are combined by an implicit "or:". With such a meaning, having
a "not:" would be unclear to define. Instead it is defined that any
"except:" match always wins and makes the entire condition to explicitly
not match. As such, it makes sense to treat a match that only consists
of "except:" matches special.
This is a change in behavior, but the alternative meaning makes
little sense.
2018-12-11 12:35:40 +01:00
If the setting only consists of "except:" matches and none of the
negative conditions are satisfied, the configuration is still enabled.
config: allow to enable/disable configuration snippets
Support a new configuration option
[.config]
enable=<ENABLED>
for configuration snippets.
This new [.config] section is only relevant within the snippet itself
and it is not merged into the combined configuration.
Currently only the "enable" key is supported. If the "enable" key is
missing, it obviously defaults to being enabled. It allows snippets
to be skipped from loading. The main configuration "NetworkManager.conf"
cannot be skipped.
<ENABLED> can be a boolean value (false), to skip a configuration
snippet from loading.
It can also be a string to match against the NetworkManager version,
like "enable=nm-version-min:1.1,nm-version-min:1.0.6"
There are several motivations for this:
- the user can disable an entire configuration snippet by toggeling
one entry.
This generalizes the functionality of the global-dns.enable
setting, but in a way that applies to configuration on a per-file
basis.
- for developing, we often switch between different versions of
NetworkManager. Thus, we might want to use different configuration.
E.g. before global-dns options, I want to use "dns=none" and manage
resolv.conf myself. Now, I can use global-dns setting to do that.
That can be achieved with something like the following (not exactly,
it's an example only):
[.config]
enable=nm-version-min:1.1
[main]
dns=default
[global-dns-domain-*]
nameserver=127.0.0.1
Arguably, this would be more awesome, if we would bump our micro devel
version (1.1.0) more often while developing 1.2.0 (*hint*).
- in principle, packages could drop configuration snippets and enable
them based on the NetworkManager version.
- with the "env:" spec, you can enable/disable snippets by configuring
an environment variable. Again, useful for testing and developing.
2015-10-01 10:43:33 +02:00
<programlisting >
# enable the configuration either when the environment variable
# is present or the version is at least 1.2.0.
[.config]
enable=env:TAG2,nm-version-min:1.2
# enable the configuration for version > = 1.2.0, but disable
# it when the environment variable is set to "TAG3"
[.config]
enable=except:env:TAG3,nm-version-min:1.2
# enable the configuration on > = 1.3, > = 1.2.6, and > = 1.0.16.
# Useful if a certain feature is only present since those releases.
[.config]
enable=nm-version-min:1.3,nm-version-min:1.2.6,nm-version-min:1.0.16
</programlisting>
</para>
</listitem>
</varlistentry>
</variablelist>
</para>
</refsect1>
2018-06-06 13:10:09 +02:00
<refsect1 id= "settings-plugins" >
2013-04-09 16:41:00 -04:00
<title > Plugins</title>
2018-06-06 13:10:09 +02:00
<para >
Settings plugins for reading and writing connection profiles. The number of
available plugins is distribution specific.
</para>
2013-04-09 16:41:00 -04:00
<variablelist >
<varlistentry >
2016-01-27 12:54:18 +01:00
<term > <varname > keyfile</varname> </term>
<listitem >
<para >
The <literal > keyfile</literal> plugin is the generic
plugin that supports all the connection types and
capabilities that NetworkManager has. It writes files out
in an .ini-style format in
2016-11-11 14:26:25 +01:00
<filename > /etc/NetworkManager/system-connections</filename> .
2018-06-06 13:10:09 +02:00
See <link linkend= 'nm-settings-keyfile' > <citerefentry > <refentrytitle > nm-settings-keyfile</refentrytitle> <manvolnum > 5</manvolnum> </citerefentry> </link>
for details about the file format.
2016-01-27 12:54:18 +01:00
</para>
<para >
2017-09-28 17:29:45 +02:00
The stored connection file may contain passwords, secrets and
private keys in plain text, so it will be made readable only to
root, and the plugin will ignore files that are readable or
writable by any user or group other than root. See "Secret flag types"
in <link linkend= 'nm-settings' > <citerefentry > <refentrytitle > nm-settings</refentrytitle> <manvolnum > 5</manvolnum> </citerefentry> </link>
for how to avoid storing passwords in plain text.
2016-01-27 12:54:18 +01:00
</para>
<para >
This plugin is always active, and will automatically be
used to store any connections that aren't supported by any
other active plugin.
</para>
</listitem>
2013-04-09 16:41:00 -04:00
</varlistentry>
2017-12-20 10:18:14 +01:00
2013-04-09 16:41:00 -04:00
<varlistentry >
2016-01-27 12:54:18 +01:00
<term > <varname > ifcfg-rh</varname> </term>
<listitem >
<para >
This plugin is used on the Fedora and Red Hat Enterprise
Linux distributions to read and write configuration from
the standard
<filename > /etc/sysconfig/network-scripts/ifcfg-*</filename>
files. It currently supports reading Ethernet, Wi-Fi,
InfiniBand, VLAN, Bond, Bridge, and Team connections.
Enabling <literal > ifcfg-rh</literal> implicitly enables
<literal > ibft</literal> plugin, if it is available.
This can be disabled by adding <literal > no-ibft</literal> .
2018-06-06 13:10:09 +02:00
See <filename > /usr/share/doc/initscripts/sysconfig.txt</filename>
and <link linkend= 'nm-settings-ifcfg-rh' > <citerefentry > <refentrytitle > nm-settings-ifcfg-rh</refentrytitle> <manvolnum > 5</manvolnum> </citerefentry> </link>
for more information about the ifcfg file format.
2016-01-27 12:54:18 +01:00
</para>
</listitem>
2013-04-09 16:41:00 -04:00
</varlistentry>
<varlistentry >
2016-01-27 12:54:18 +01:00
<term > <varname > ifupdown</varname> </term>
<listitem >
<para >
This plugin is used on the Debian and Ubuntu
distributions, and reads Ethernet and Wi-Fi connections
from <filename > /etc/network/interfaces</filename> .
</para>
<para >
This plugin is read-only; any connections (of any type)
added from within NetworkManager when you are using this
plugin will be saved using the <literal > keyfile</literal>
plugin instead.
</para>
</listitem>
2013-04-09 16:41:00 -04:00
</varlistentry>
2014-10-21 09:24:06 +02:00
<varlistentry >
2016-01-27 12:54:18 +01:00
<term > <varname > ibft</varname> , <varname > no-ibft</varname> </term>
<listitem >
<para >
2019-06-19 07:47:20 +02:00
These plugins are deprecated and their selection has no effect.
This is now handled by nm-initrd-generator.
2016-09-02 13:35:00 +02:00
</para>
2016-01-27 12:54:18 +01:00
</listitem>
2014-10-21 09:24:06 +02:00
</varlistentry>
2017-12-20 10:18:14 +01:00
<varlistentry >
2017-12-20 10:42:41 +01:00
<term > <varname > ifcfg-suse</varname> , <varname > ifnet</varname> </term>
2017-12-20 10:18:14 +01:00
<listitem >
<para >
2017-12-20 10:42:41 +01:00
These plugins are deprecated and their selection has no effect.
2017-12-20 10:18:14 +01:00
The <literal > keyfile</literal> plugin should be used
instead.
</para>
</listitem>
</varlistentry>
2013-04-09 16:41:00 -04:00
</variablelist>
</refsect1>
2015-02-09 15:33:28 +01:00
<refsect1 >
<title > Appendix</title>
<refsect2 id= "device-spec" >
<title > Device List Format</title>
<para >
The configuration options <literal > main.no-auto-default</literal> , <literal > main.ignore-carrier</literal> ,
config: make "ignore-carrier" a per-device configuration option
NetworkManager.conf already contains several per-device settings,
that is, settings that have a device-spec as argument.
main.ignore-carrier
main.no-auto-default
main.assume-ipv6ll-only
keyfile.unmanged-devices
Optimally, these settings should be moved to the new [device*]
section.
For now, only move main.ignore-carrier there. For the others
it may not make sense to do so:
- main.no-auto-default: is already merged with internal state
from /var/lib/NetworkManager/no-auto-default.state. While
NMConfig's write API would be fine to also persist and merge
the no-auto-default setting, we'd still have to read the old
file too. Thus, deprecating this setting gets quite cumbersome
to still handle the old state file.
Also, it seems a less useful setting to configure in the
global configuration aside setting main.no-auto-default=*.
- main.assume-ipv6ll-only: one day, I hope that we no longer
assume connections at all, and this setting becomes entirely
obsolete.
- keyfile.unmanged-devices: this sets NM_UNMANAGED_USER_SETTINGS,
which cannot be overruled via D-Bus. For a future device.managed
setting we want it it to be overwritable via D-Bus by an explicit
user action. Thus, a device.managed property should have a different
semantic, this should be more like a device.unmanaged-force setting,
which could be done.
2016-06-22 13:43:08 +02:00
<literal > keyfile.unmanaged-devices</literal> , <literal > connection*.match-device</literal> and
<literal > device*.match-device</literal> select devices based on a list of matchings.
2015-02-09 15:33:28 +01:00
Devices can be specified using the following format:
</para>
<para >
<variablelist >
<varlistentry >
<term > *</term>
<listitem > <para > Matches every device.</para> </listitem>
</varlistentry>
<varlistentry >
<term > IFNAME</term>
<listitem > <para > Case sensitive match of interface name of the device. Globbing is not supported.</para> </listitem>
</varlistentry>
<varlistentry >
<term > HWADDR</term>
2016-06-15 11:25:29 +02:00
<listitem > <para > Match the permanent MAC address of the device. Globbing is not supported</para> </listitem>
2015-02-09 15:33:28 +01:00
</varlistentry>
<varlistentry >
<term > interface-name:IFNAME</term>
<term > interface-name:~IFNAME</term>
<listitem > <para > Case sensitive match of interface name of the device. Simple globbing is supported with
<literal > *</literal> and <literal > ?</literal> . Ranges and escaping is not supported.</para> </listitem>
</varlistentry>
<varlistentry >
<term > interface-name:=IFNAME</term>
<listitem > <para > Case sensitive match of interface name of the device. Globbing is disabled and <literal > IFNAME</literal>
is taken literally.</para> </listitem>
</varlistentry>
<varlistentry >
<term > mac:HWADDR</term>
2016-06-15 11:25:29 +02:00
<listitem > <para > Match the permanent MAC address of the device. Globbing is not supported</para> </listitem>
2015-02-09 15:33:28 +01:00
</varlistentry>
<varlistentry >
<term > s390-subchannels:HWADDR</term>
<listitem > <para > Match the device based on the subchannel address. Globbing is not supported</para> </listitem>
</varlistentry>
2015-05-15 14:30:40 +02:00
<varlistentry >
<term > type:TYPE</term>
<listitem > <para > Match the device type. Valid type names are as reported by "<literal > nmcli -f GENERAL.TYPE device show</literal> ".
Globbing is not supported.</para> </listitem>
</varlistentry>
2017-03-17 16:18:48 +01:00
<varlistentry >
<term > driver:DRIVER</term>
<listitem > <para > Match the device driver as reported by "<literal > nmcli -f GENERAL.DRIVER,GENERAL.DRIVER-VERSION device show</literal> ".
"<literal > DRIVER</literal> " must match the driver name exactly and does not support globbing.
Optionally, a driver version may be specified separated by '/'. Globbing is supported for the version.
</para> </listitem>
</varlistentry>
device: add "dhcp-plugin" match spec for device
The need for this is the following:
"ipv4.dhcp-client-id" can be specified via global connection defaults.
In absence of any configuration in NetworkManager, the default depends
on the DHCP client plugin. In case of "dhclient", the default further
depends on /etc/dhcp.
For "internal" plugin, we may very well want to change the default
client-id to "mac" by universally installing a configuration
snippet
[connection-use-mac-client-id]
ipv4.dhcp-client-id=mac
However, if we the user happens to enable "dhclient" plugin, this also
forces the client-id and overrules configuration from /etc/dhcp. The real
problem is, that dhclient can be configured via means outside of NetworkManager,
so our defaults shall not overwrite defaults from /etc/dhcp.
With the new device spec, we can avoid this issue:
[connection-dhcp-client-id]
match-device=except:dhcp-plugin:dhclient
ipv4.dhcp-client-id=mac
This will be part of the solution for rh#1640494. Note that merely
dropping a configuration snippet is not yet enough. More fixes for
DHCP will follow. Also, bug rh#1640494 may have alternative solutions
as well. The nice part of this new feature is that it is generally
useful for configuring connection defaults and not specifically for
the client-id issue.
Note that this match spec is per-device, although the plugin is selected
globally. That makes some sense, because in the future we may or may not
configure the DHCP plugin per-device or per address family.
https://bugzilla.redhat.com/show_bug.cgi?id=1640494
2018-10-24 08:43:45 +02:00
<varlistentry >
<term > dhcp-plugin:DHCP</term>
<listitem > <para > Match the configured DHCP plugin "<literal > main.dhcp</literal> ".
</para> </listitem>
</varlistentry>
2015-02-09 15:33:28 +01:00
<varlistentry >
<term > except:SPEC</term>
<listitem > <para > Negative match of a device. <literal > SPEC</literal> must be explicitly qualified with
a prefix such as <literal > interface-name:</literal> . A negative match has higher priority then the positive
core: fix match spec behavior for a list of all "except:"
If the spec specifies only negative matches (and none of them matches),
then the result shall be positive.
Meaning:
[connection*] match-device=except:dhcp-plugin:dhclient
[connection*] match-device=except:interface-name:eth0
[.config] enabled=except:nm-version:1.14
should be the same as:
[connection*] match-device=*,except:dhcp-plugin:dhclient
[connection*] match-device=*,except:interface-name:eth0
[.config] enabled=*,except:nm-version:1.14
and match by default. Previously, such specs would never yield a
positive match, which seems wrong.
Note that "except:" already has a special meaning. It is not merely
"not:". That is because we don't support "and:" nor grouping, but all
matches are combined by an implicit "or:". With such a meaning, having
a "not:" would be unclear to define. Instead it is defined that any
"except:" match always wins and makes the entire condition to explicitly
not match. As such, it makes sense to treat a match that only consists
of "except:" matches special.
This is a change in behavior, but the alternative meaning makes
little sense.
2018-12-11 12:35:40 +01:00
matches above.</para>
<para > If there is a list consisting only of negative matches, the behavior is the same as if there
is also match-all. That means, if none of all the negative matches is satisfied, the overall result is
still a positive match. That means, <literal > "except:interface-name:eth0"</literal> is the same as
<literal > "*,except:interface-name:eth0"</literal> .</para> </listitem>
2015-02-09 15:33:28 +01:00
</varlistentry>
<varlistentry >
<term > SPEC[,;]SPEC</term>
2015-12-24 09:48:20 +01:00
<listitem > <para > Multiple specs can be concatenated with commas or semicolons. The order does not matter as
2015-12-09 13:28:11 +01:00
matches are either inclusive or negative (<literal > except:</literal> ), with negative matches having higher
priority.
</para>
2015-02-09 15:33:28 +01:00
<para > Backslash is supported to escape the separators ';' and ',', and to express special
characters such as newline ('\n'), tabulator ('\t'), whitespace ('\s') and backslash ('\\'). The globbing of
2015-12-09 13:28:11 +01:00
interface names cannot be escaped. Whitespace is not a separator but will be trimmed between
two specs (unless escaped as '\s').
</para>
</listitem>
2015-02-09 15:33:28 +01:00
</varlistentry>
</variablelist>
</para>
<para >
Example:
<programlisting >
interface-name:em4
mac:00:22:68:1c:59:b1;mac:00:1E:65:30:D1:C4;interface-name:eth2
interface-name:vboxnet*,except:interface-name:vboxnet2
*,except:mac:00:22:68:1c:59:b1
2016-01-27 12:54:18 +01:00
</programlisting>
2015-02-09 15:33:28 +01:00
</para>
</refsect2>
</refsect1>
2013-04-09 16:41:00 -04:00
<refsect1 >
<title > See Also</title>
<para >
2016-05-30 20:24:32 +02:00
<link linkend= 'NetworkManager' > <citerefentry > <refentrytitle > NetworkManager</refentrytitle> <manvolnum > 8</manvolnum> </citerefentry> </link> ,
<link linkend= 'nmcli' > <citerefentry > <refentrytitle > nmcli</refentrytitle> <manvolnum > 1</manvolnum> </citerefentry> </link> ,
<link linkend= 'nmcli-examples' > <citerefentry > <refentrytitle > nmcli-examples</refentrytitle> <manvolnum > 7</manvolnum> </citerefentry> </link> ,
<link linkend= 'nm-online' > <citerefentry > <refentrytitle > nm-online</refentrytitle> <manvolnum > 1</manvolnum> </citerefentry> </link> ,
<link linkend= 'nm-settings' > <citerefentry > <refentrytitle > nm-settings</refentrytitle> <manvolnum > 5</manvolnum> </citerefentry> </link> ,
2013-04-09 16:41:00 -04:00
<citerefentry > <refentrytitle > nm-applet</refentrytitle> <manvolnum > 1</manvolnum> </citerefentry> ,
<citerefentry > <refentrytitle > nm-connection-editor</refentrytitle> <manvolnum > 1</manvolnum> </citerefentry>
</para>
</refsect1>
</refentry>