207 lines
5.3 KiB
Markdown
207 lines
5.3 KiB
Markdown
interfaces(5)
|
|
|
|
# NAME
|
|
|
|
*/etc/network/interfaces* - interface configuration database
|
|
|
|
# DESCRIPTION
|
|
|
|
The */etc/network/interfaces* file is used to specify how network
|
|
interfaces are configured. The file is processed by *ifquery*(8),
|
|
*ifup*(8) and *ifdown*(8) to introspect and change system state.
|
|
|
|
In most cases, syntax from legacy implementations is supported as
|
|
well, but that syntax is not discussed here.
|
|
|
|
# FILE SYNTAX
|
|
|
|
The interface configuration database is composed of a series of
|
|
stanzas. Hash symbols designate comments, which are ignored by
|
|
the system.
|
|
|
|
A stanza is a collection of triples, where a triple is a key and
|
|
value combination that is related to an *object*. Triples which
|
|
are not associated with an *object* are considered to be part
|
|
of the root of the configuration tree.
|
|
|
|
The following is a simple example of a stanza:
|
|
|
|
```
|
|
auto eth0
|
|
iface eth0
|
|
address 203.0.113.2/24
|
|
gateway 203.0.113.1
|
|
```
|
|
|
|
This stanza defines an interface named *eth0* which is configured
|
|
with an address of *203.0.113.2* and gateway of *203.0.113.1*.
|
|
|
|
# SUPPORTED KEYWORDS FOR UNASSOCIATED TRIPLES
|
|
|
|
*auto* _object_
|
|
Designates that _object_ should be automatically configured
|
|
by the system when appropriate.
|
|
|
|
*iface* _object_
|
|
Begins a new declaration for _object_. Any child keyword
|
|
associated with the declaration will be stored inside
|
|
_object_.
|
|
|
|
*template* _object_
|
|
Begins a new declaration for _object_, like *iface*, except
|
|
that _object_ is defined as a *template*.
|
|
|
|
# SUPPORTED KEYWORDS FOR OBJECT TRIPLES
|
|
|
|
Any keyword may be used inside an interface declaration block, but
|
|
the system will only respond to certain keywords by default:
|
|
|
|
*address* _address_
|
|
Associates an IPv4 or IPv6 address in CIDR notation with
|
|
the parent interface.
|
|
|
|
*gateway* _address_
|
|
Associates an IPv4 or IPv6 address with the parent interface
|
|
for use as a default route (gateway).
|
|
|
|
*netmask* _netmask_
|
|
Associates a fallback netmask with the parent interface for
|
|
addresses which do not have a CIDR length set. This option
|
|
is for backwards compatibility and should not be used in new
|
|
deployments.
|
|
|
|
*link-type* _link-type_
|
|
Denotes the link-type of the interface. When set to _dummy_,
|
|
the interface is created as a virtual dummy interfaces.
|
|
When set to _veth_ the interface is created as virtual veth
|
|
interface (pair).
|
|
|
|
*veth-peer-name* _peer-name_
|
|
Denotes the name of the veth peer interfaces. If not set
|
|
the kernel will name the veth peer interface as _vethN_
|
|
with N being an integer number.
|
|
|
|
*alias* _alias_
|
|
Sets the given alias on the interface.
|
|
|
|
*requires* _interfaces_...
|
|
Designates one or more required interfaces that must be
|
|
brought up before configuration of the parent interface.
|
|
Interfaces associated with the parent are taken down at
|
|
the same time as the parent.
|
|
|
|
*inherit* _object_
|
|
Designates that the configured interface should inherit
|
|
configuration data from _object_. Normally _object_
|
|
must be a *template*.
|
|
|
|
*use* _option_
|
|
Designates that an option should be used. See _OPTIONS_
|
|
section for more information on options.
|
|
|
|
*pre-down* _command_
|
|
Runs _command_ before taking the interface down.
|
|
|
|
*down* _command_
|
|
Runs _command_ when the interface is taken down.
|
|
|
|
*post-down* _command_
|
|
Runs _command_ after taking the interface down.
|
|
|
|
*pre-up* _command_
|
|
Runs _command_ before bringing the interface up.
|
|
|
|
*up* _command_
|
|
Runs _command_ when the interface is brought up.
|
|
|
|
*post-up* _command_
|
|
Runs _command_ after bringing the interface up.
|
|
|
|
Additional packages such as *bonding*, *bridge*, *tunnel*, *vrf* and
|
|
*vxlan* add additional keywords to this vocabulary.
|
|
|
|
# OPTIONS
|
|
|
|
The *use* keyword designates that an _option_ should be used.
|
|
This system is extendable by additional packages, but the
|
|
most common options are:
|
|
|
|
*batman*
|
|
The interface is a B.A.T.M.A.N. adv. mesh interface.
|
|
Configuration of B.A.T.M.A.N. adv. interfaces requires the
|
|
*batctl* untiliy to be installed.
|
|
|
|
*bond*
|
|
The interface is a bonded interface. Configuration
|
|
of bonded interfaces requires the *bonding* package
|
|
to be installed.
|
|
|
|
*bridge*
|
|
The interface is an ethernet bridge. Configuration
|
|
of ethernet bridges requires the *bridge* package
|
|
to be installed.
|
|
|
|
*dhcp*
|
|
Use a DHCP client to learn the IPv4 address of an
|
|
interface.
|
|
|
|
*loopback*
|
|
Designates the interface as a loopback device.
|
|
|
|
*tunnel*
|
|
The interface is a tunnel. Configuration of tunnels
|
|
requires the *tunnel* package to be installed.
|
|
|
|
*vrf*
|
|
The interface is a VRF. Configuration of VRFs requires
|
|
the *vrf* package to be installed.
|
|
|
|
*vxlan*
|
|
The interface is a Virtual Extensible LAN (VXLAN) tunnel
|
|
endpoint.
|
|
|
|
Check *interfaces-<option>(5)* for further informaton about a given
|
|
option and available configuration parameters.
|
|
|
|
# EXAMPLES
|
|
|
|
Configure a bridge interface *br0* with *bond0* attached to it,
|
|
which is a failover between *eth0* and *eth1*. This requires
|
|
the *bonding* and *bridge* packages to be installed:
|
|
|
|
```
|
|
auto br0
|
|
iface br0
|
|
use bridge
|
|
requires bond0
|
|
address 203.0.113.2/24
|
|
gateway 203.0.113.1
|
|
|
|
iface bond0
|
|
use bond
|
|
requires eth0 eth1
|
|
bond-mode 802.3ad
|
|
bond-xmit-hash-policy layer2+3
|
|
```
|
|
|
|
Configure a network interface to use DHCP to learn its IPv4
|
|
address:
|
|
|
|
```
|
|
auto eth0
|
|
iface eth0
|
|
use dhcp
|
|
```
|
|
|
|
# SEE ALSO
|
|
|
|
*interfaces-bond*(5)
|
|
*interfaces-batman*(5)
|
|
*interfaces-bridge*(5)
|
|
*interfaces-vrf*(5)
|
|
*interfaces-vxlan*(5)
|
|
|
|
# AUTHORS
|
|
|
|
Ariadne Conill <ariadne@dereferenced.org>++
|
|
Maximilian Wilhelm <max@sdn.clinic>
|