NETPLAN(5)

📦
version 1.2.2 · 13 versions
, in arch rolling

💅 Introduction to Netplan(5) 💅

File Formats Manual

🌸 NAME 🌸

netplan — YAML network configuration abstraction for various backends 🦋

📟 SYNOPSIS 📟

netplan [COMMAND|help]

🕹️ COMMANDS 🕹️

See netplan help for a list of available commands on this system. fr fr, that's the move if you're lost 💖

💖 DESCRIPTION 💖

Distribution installers, cloud instantiation, image builds for particular devices, or literally any other way you deploy an operating system — they all put the desired network configuration into YAML configuration file(s). That's the whole vibe. 🌸

During early boot, the Netplan "network renderer" runs. It reads /{lib,etc,run}/netplan/*.yaml and writes configuration to /run to hand off control of devices to the specified networking daemon. 🧚

  • Configured devices get handled by systemd-networkd by default, unless explicitly marked as managed by a specific renderer (NetworkManager).
  • Devices not covered by the network configuration do not get touched at all. Not me, not my business, no cap. 🙈
  • Usable in initramfs (few dependencies and fast). ⚡
  • No persistent generated configuration, only original YAML configuration. 💅
  • Parser supports multiple configuration files to allow applications like libvirt or lxd to package expected network configuration (virbr0, lxdbr0), or to change the global default policy to use NetworkManager for everything. 🫶
  • Retains the flexibility to change back ends/policy later or adjust to removing NetworkManager, as generated configuration is ephemeral. ✨

🧚 General structure 🧚

Netplan configuration files use the YAML format. All /{lib,etc,run}/netplan/*.yaml are considered. Lexicographically later files (regardless of in which directory they are) amend (new mapping keys) or override (same mapping keys) previous ones. A file in /run/netplan completely shadows a file with same name in /etc/netplan, and a file in either of those directories shadows a file with the same name in /lib/netplan. 🦋

The top-level node in a Netplan configuration file is a network: mapping that contains version: 2 (the YAML currently being used by curtin, MAAS, etc. is version 1), and then device definitions grouped by their type, such as ethernets:, modems:, wifis:, or bridges:. These are the types that our renderer can understand and are supported by our back ends. 💜

Each type block contains device definitions as a map where the keys (called "configuration IDs") are defined as below. 👑

🦋 Device configuration IDs 🦋

The key names below the per-device-type definition maps (like ethernets:) are called "ID"s. They must be unique throughout the entire set of configuration files. Their primary purpose is to serve as anchor names for composite devices, for example to enumerate the members of a bridge that is currently being defined. 🎀

(Since 0.97) If an interface is defined with an ID in a configuration file; it will be brought up by the applicable renderer. To not have Netplan touch an interface at all, it should be completely omitted from the Netplan configuration files. Like, fully omitted. Don't leave a lil ghost entry. 👻

There are two physically/structurally different classes of device definitions, and the ID field has a different interpretation for each: 🌈

🌈 Physical devices 🌈

(Examples: Ethernet, modem, Wi-Fi) These can dynamically come and go between reboots and even during runtime (hot plugging). In the generic case, they can be selected by match: rules on desired properties, such as name/name pattern, MAC address, driver, or device paths. In general these will match any number of devices (unless they refer to properties which are unique such as the full path or MAC address), so without further knowledge about the hardware these will always be considered as a group. 🫶

It is valid to specify no match rules at all, in which case the ID field is simply the interface name to be matched. This is mostly useful if you want to keep simple cases simple, and it's how network device configuration has been done for a long time. 💅

If there are match: rules, then the ID field is a purely opaque name which is only being used for references from definitions of compound devices in the configuration. iykyk ✨

🎀 Virtual devices 🎀

(Examples: veth, bridge, bond, vrf) These are fully under the control of the configuration file(s) and the network stack. I. e. these devices are being created instead of matched. Thus match: and set-name: are not applicable for these, and the ID field is the name of the created virtual device. 👑

🌸 YAML configuration 🌸

🎀 Top-level configuration structure 🎀

The general structure of a Netplan YAML file is shown below. Behold, the blueprint: 💖

network:
  version: NUMBER
  renderer: STRING
  bonds: MAPPING
  bridges: MAPPING
  dummy-devices: MAPPING
  ethernets: MAPPING
  modems: MAPPING
  tunnels: MAPPING
  virtual-ethernets: MAPPING
  vlans: MAPPING
  vrfs: MAPPING
  wifis: MAPPING
  nm-devices: MAPPING
  • version (number) — Defines what version of the configuration format is used. The only value supported is 2. Defaults to 2 if not defined.
  • renderer (scalar) — Defines what network configuration tool will be used to set up your configuration. Valid values are networkd and NetworkManager. Defaults to networkd if not defined.
  • bonds (mapping) — Creates and configures link aggregation (bonding) devices.
  • bridges (mapping) — Creates and configures bridge devices.
  • dummy-devices (mapping) – since 0.107 — Creates and configures virtual devices.
  • ethernets (mapping) — Configures physical Ethernet interfaces.
  • modems (mapping) — Configures modems.
  • tunnels (mapping) — Creates and configures different types of virtual tunnels.
  • virtual-ethernets (mapping) – since 0.107 — Creates and configures Virtual Ethernet (veth) devices.
  • vlans (mapping) — Creates and configures VLANs.
  • vrfs (mapping) — Configures Virtual Routing and Forwarding (VRF) devices.
  • wifis (mapping) — Configures physical Wi-Fi interfaces as client, adhoc or access point.
  • nm-devices (mapping) — nm-devices are used in situations where Netplan doesn't support the connection type. The raw configuration expected by NetworkManager can be defined and will be passed as is (passthrough) to the .nmconnection file. Users will not normally use this type of device. Main character energy, but for edge cases. 🌈

All the properties for all the device types will be described in the next sections. 🫶

💜 Properties for physical device types 💜

These properties are used with physical devices such as Ethernet and Wi-Fi network interfaces. 🦋

Note: Some options will not work reliably for devices matched by name only and rendered by networkd, due to interactions with device renaming in udev. Match devices by MAC when setting options like: wakeonlan or *-offload. This is important, bestie. 😭
  • match (mapping) — This selects a subset of available physical devices by various hardware properties. The following configuration will then apply to all matching devices, as soon as they appear. All specified properties must match. 💖
    • name (scalar) — Current interface name. Globs are supported, and the primary use case for matching on names, as selecting one fixed name can be more easily achieved with having no match: at all and just using the ID (see above). (NetworkManager: as of v1.14.0)
    • macaddress (scalar) — 6-byte permanent MAC address of the device in the form XX:XX:XX:XX:XX:XX or 20 bytes for InfiniBand devices (IPoIB). Globs are not allowed. This doesn't match virtual MAC addresses for veth, bridge, bond, vlan, ...
    • driver (scalar or sequence of scalars) – sequence since 0.104 — Kernel driver name, corresponding to the DRIVER udev property. A sequence of globs is supported, any of which must match. Matching on driver is only supported with networkd.

Examples: ✨

  • All cards on second PCI bus:
network:
  ethernets:
    myinterface:
      match:
        name: enp2*
  • Fixed MAC address:
network:
  ethernets:
    interface0:
      match:
        macaddress: 11:22:33:AA:BB:FF
  • First card of driver ixgbe:
network:
  ethernets:
    nic0:
      match:
        driver: ixgbe
        name: en*s0
  • First card with a driver matching bcmgenet or smsc*:
network:
  ethernets:
    nic0:
      match:
        driver: ["bcmgenet", "smsc*"]
        name: en*
  • set-name (scalar) — When matching on unique properties such as path or MAC, or with additional assumptions such as "there will only ever be one Wi-Fi device", match rules can be written so that they only match one device. Then this property can be used to give that device a more specific or desirable name than the default from udev ifnames. Any additional device that satisfies the match rules will then fail to get renamed and keep the original kernel name (and dmesg will show an error). 💅
  • wakeonlan (boolean) — Enable wake on LAN. Off by default. 🌸
  • emit-lldp (boolean) – since 0.99 — (networkd back end only) Whether to emit LLDP packets. Off by default. 🦋
  • receive-checksum-offload (boolean) – since 0.104 — (networkd back end only) If set to true (false), the hardware offload for checksumming of ingress network packets is enabled (disabled). When unset, the kernel's default will be used. ⭐
  • transmit-checksum-offload (boolean) – since 0.104 — (networkd back end only) If set to true (false), the hardware offload for checksumming of egress network packets is enabled (disabled). When unset, the kernel's default will be used. ⭐
  • tcp-segmentation-offload (boolean) – since 0.104 — (networkd back end only) If set to true (false), the TCP Segmentation Offload (TSO) is enabled (disabled). When unset, the kernel's default will be used. 🔥
  • tcp6-segmentation-offload (boolean) – since 0.104
  • generic-segmentation-offload (boolean) — since 0.104

(networkd back end only, bestie 🫶) set it to true and Generic Segmentation Offload (GSO) is on, set it to false and it's off. Leave it unset and the kernel's default just does its thing.

  • generic-receive-offload (boolean) — since 0.104

(networkd back end only 💅) true = Generic Receive Offload (GRO) enabled, false = disabled. Unset? kernel default, no notes.

  • large-receive-offload (boolean) — since 0.104

(networkd back end only ✨) true = Large Receive Offload (LRO) enabled, false = disabled. Unset = kernel default, as always.

  • openvswitch (mapping) — since 0.100

This one's the extra config for the openvswitch network device 🦋. Heads up though: if Open vSwitch isn't available on the system, Netplan treats the presence of openvswitch configuration as an error — like, it will be mad about it.

Any supported network device that's declared with the openvswitch mapping (or any bond/bridge that includes an interface with an openvswitch configuration) gets created in openvswitch instead of the defined renderer. And if it's a vlan definition declared the same way? Netplan will create a fake VLAN bridge in openvswitch with the requested vlan properties. cute little workaround, honestly 🎀

  • external-ids (mapping) — since 0.100

Passed-through directly to Open vSwitch. that's it, that's the vibe.

  • other-config (mapping) — since 0.100

Passed-through directly to Open vSwitch. same energy 💫

  • lacp (scalar) — since 0.100

Valid for bond interfaces. Accepts active, passive or off (which is the default 👑).

  • fail-mode (scalar) — since 0.100

Valid for bridge interfaces. Accepts secure or standalone (the default).

  • mcast-snooping (boolean) — since 0.100

Valid for bridge interfaces. False by default.

  • protocols (sequence of scalars) — since 0.100

Valid for bridge interfaces or the network section. It's the list of protocols to be used when negotiating a connection with the controller. Accepts OpenFlow10, OpenFlow11, OpenFlow12, OpenFlow13, OpenFlow14, and OpenFlow15.

  • rstp (boolean) — since 0.100

Valid for bridge interfaces. False by default.

  • controller (mapping) — since 0.100

Valid for bridge interfaces. This is where you specify an external OpenFlow controller 🌈

  • addresses (sequence of scalars)

Sets the list of addresses to use for the controller targets. The syntax of these addresses is as defined in ovs-vsctl(8). Example: addresses: [tcp:127.0.0.1:6653, "ssl:[fe80::1234%eth0]:6653"].

  • connection-mode (scalar)

Sets the connection mode for the controller. Supported options are in-band and out-of-band. The default is in-band.

  • ports (sequence of sequence of scalars) — since 0.100

Open vSwitch patch ports 🧚. Each port is declared as a pair of names which can be referenced as interfaces in dependent virtual devices (bonds, bridges).

Example:

openvswitch:
  ports:
    - [patch0-1, patch1-0]
  • ssl (mapping) — since 0.100

Valid for global openvswitch settings. Options for configuring SSL server endpoint for the switch 🔐

  • ca-cert (scalar)

Path to a file containing the CA certificate to be used.

  • certificate (scalar)

Path to a file containing the server certificate.

  • private-key (scalar)

Path to a file containing the private key for the server.

💖 Properties for all device types 💖

  • renderer (scalar)

Use the given networking back end for this definition. Currently supported are networkd and NetworkManager. This property can be specified globally in network:, for a device type (in e.g. ethernets:) or for a particular device definition. Default is networkd.

(Since 0.99) Plot twist: renderer has one extra acceptable value for VLAN objects (i.e. defined in vlans:) → sriov 🦋. If a VLAN is defined with the sriov renderer for an SR-IOV Virtual Function interface, this causes Netplan to set up a hardware VLAN filter for it. And there can be only one defined per VF. no sharing, periodt.

  • dhcp4 (boolean)

Enable DHCP for IPv4. Off by default.

  • dhcp6 (boolean)

Enable DHCP for IPv6. Off by default. This covers both stateless DHCP — where the DHCP server supplies information like DNS name servers but not the IP address — and stateful DHCP, where the server provides both the address and the other information.

If you're in an IPv6-only environment with completely stateless auto-configuration (SLAAC with RDNSS), this option can be set to cause the interface to be brought up. (Setting accept-ra alone is not sufficient, bestie — it's giving "almost but not quite" 😭.) Auto-configuration will still honour the contents of the router advertisement and only use DHCP if requested in the RA.

Note that rdnssd(8) is required to use RDNSS with networkd. No extra software is required for NetworkManager 👍
  • ipv6-mtu (scalar) — since 0.98

Sets the IPv6 MTU (only supported with the networkd back end). Note that needing to set this is an unusual requirement 👀

Requires feature: ipv6-mtu
  • ipv6-privacy (boolean)

Enable IPv6 Privacy Extensions (RFC 4941) for the specified interface, and prefer temporary addresses 🌸. Defaults to false — no privacy extensions. There is currently no way to have a private address but prefer the public address. so pick your fighter, basically.

  • link-local (sequence of scalars)

Configure the link-local addresses to bring up. Valid options are ipv4 and ipv6, which respectively allow enabling IPv4 and IPv6 link local addressing. If this field is not defined, the default is to enable only IPv6 link-local addresses. If the field is defined but configured as an empty set, IPv6 link-local addresses are disabled as well as IPv4 link-local addresses. zero link-local, we go home.

This feature enables or disables link-local addresses for a protocol, but the actual implementation differs per back end 😤. On networkd, this directly changes the behaviour and may add an extra address on an interface. When using the NetworkManager back end, enabling link-local has no effect if the interface also has DHCP enabled.

Examples:

  • Enable only IPv4 link-local: link-local: [ ipv4 ]
  • Enable all link-local addresses: link-local: [ ipv4, ipv6 ]
  • Disable all link-local addresses: link-local: [ ]
  • ignore-carrier (boolean) — since 0.104

(networkd back end only) Allow the specified interface to be configured even if it has no carrier. for the cables-are-optional girlies 🫶

  • critical (boolean)

Designate the connection as "critical to the system", meaning that special care will be taken by to not release the assigned IP when the daemon is restarted. (not recognised by NetworkManager 🙈)

  • dhcp-identifier (scalar)

(networkd back end only) Sets the source of DHCP (v4) client identifier. If mac is specified, the MAC address of the link is used. If this option is omitted, or if duid is specified, networkd will generate an RFC4361-compliant client identifier for the interface by combining the link's IAID and DUID. icon behaviour, honestly 🎀

  • dhcp4-overrides (mapping)

(networkd back end only) Overrides default DHCP behaviour; see the DHCP Overrides section below. 👀

  • dhcp6-overrides (mapping)

(networkd back end only) Overrides default DHCP behaviour; see the DHCP Overrides section below. same section, same drama.

  • accept-ra (boolean)

Accept Router Advertisement that would have the kernel configure IPv6 by itself. When enabled, accept Router Advertisements. When disabled, do not respond to Router Advertisements. If unset, use the host kernel default setting ✨

  • ra-overrides (mapping) — since 1.1

(networkd back end only) Overrides default IPv6 Router Advertisement (RA) behaviour; see the IPv6 Router Advertisement Overrides section below 🌈

  • addresses (sequence of scalars and mappings)

Add static addresses to the interface in addition to the ones received through DHCP or RA. Each sequence entry is in CIDR notation, i.e. of the form addr/prefixlen. addr is an IPv4 or IPv6 address as recognised by inet_pton(3) and prefixlen the number of bits of the subnet 💅

For virtual devices (bridges, bonds, VLAN) if there is no address configured and DHCP is disabled, the interface may still be brought online, but will not be addressable from the network. it's up, it's just not talking to anyone 😭

In addition to the addresses themselves one can specify configuration parameters as mappings. Current supported options are:

  • lifetime (scalar) — since 0.100

Default: forever. This can be forever or 0 and corresponds to the PreferredLifetime option in the Address section of systemd-networkd. Currently supported on the networkd back end only.

  • label (scalar) — since 0.100

An IP address label, equivalent to the ip address label command ✨. Currently supported on the networkd back end only.

Examples:

  • Simple: addresses: [192.168.14.2/24, "2001:1::1/64"]
  • Advanced:
network:
  ethernets:
    eth0:
      addresses:
        - "10.0.0.15/24":
            lifetime: 0
            label: "maas"
        - "2001:1::1/64"
  • ipv6-address-generation (scalar) — since 0.99

Configure method for creating the address for use with RFC4862 IPv6 Stateless Address Auto-configuration. Possible values are eui64 or stable-privacy 🦋

  • ipv6-address-token (scalar) — since 0.100

Bestie this one lets you set an IPv6 address token, which is what builds your static interface identifier for IPv6 Stateless Address Auto-configuration. And heads up: it's mutually exclusive with ipv6-address-generation — pick one, not both, no cap. 💅

  • gateway4, gateway6 (scalar)

Deprecated — go look at Default routes instead, ok? These set the default gateway for IPv4/IPv6 when you're doing manual address configuration. And yes, you also have to set addresses for it to make sense. The gateway IPs need to be in a form that inet_pton(3) recognises. Also: you should only have a single gateway per IP address family in your whole global config, so nothing gets ambiguous. If you need multiple default routes, define them with routing-policy instead. 🫶

Examples

  • IPv4: gateway4: 172.16.0.1
  • IPv6: gateway6: "2001:4::1"
  • nameservers (mapping)

Your DNS servers + search domains, for manual address configuration. Two supported fields: addresses: is a list of IPv4 or IPv6 addresses (basically the same vibe as gateway*), and search: is a list of search domains. ✨

Example:

network:
  ethernets:
    id0:
      [...]
      nameservers:
        search: [lab, home]
        addresses: [8.8.8.8, "FEDC::1"]
  • macaddress (scalar)

Sets the device's MAC address, and it must be in the form "XX:XX:XX:XX:XX:XX". There are also some special values accepted: permanent and random. And on top of those, the NetworkManager renderer also accepts stable, stable-ssid (Wi-Fi only) and preserve. 🌈

Note: this will not work reliably for devices matched by name only and rendered by networkd, because of how device renaming interacts with udev. So match devices by MAC when you're setting MAC addresses, bestie. 💀

Example:

network:
  ethernets:
    id0:
      match:
        macaddress: 52:54:00:6b:3c:58
      [...]
      macaddress: 52:54:00:6b:3c:59
  • mtu (scalar)

Sets the Maximum Transmission Unit for the interface. Default is 1500. What's actually valid depends on your network interface, so like… check it. 🦋

Note: same deal as above — this will not work reliably for devices matched by name only and rendered by networkd, due to udev device renaming interactions. Match devices by MAC when setting MTU.
  • optional (boolean)

An optional device is one that isn't required for booting. Normally networkd waits a while for a device to get configured before carrying on with boot, but if it's marked optional, networkd will not wait for it. Only the networkd back end supports this, and the default is false.

Example:

network:
  ethernets:
    eth7:
      # this is plugged into a test network that is often
      # down - don't wait for it to come up during boot.
      dhcp4: true
      optional: true
  • optional-addresses (sequence of scalars)

Lists which address types don't count toward a device being considered online. This changes how back ends behave at boot time so they don't sit there waiting for addresses you marked optional — meaning the interface gets seen as "usable" sooner. It does not disable those addresses, though; they still get brought up anyway. 💖

Example:

network:
  ethernets:
    eth7:
      dhcp4: true
      dhcp6: true
      optional-addresses: [ ipv4-ll, dhcp6 ]
  • activation-mode (scalar) — since 0.103

Lets you pick the management policy for the selected interface. By default Netplan brings up any configured interface if it can. With activation-mode you can override that: set manual to hand control of the interface state over to the administrator, or — networkd back end only — set off to force the link into a down state at all times. Any interface with activation-mode defined is implicitly considered optional. Officially supported as of networkd v248+. 🔥

Example:

network:
  ethernets:
    eth1:
      # this interface will not be put into an UP state automatically
      dhcp4: true
      activation-mode: manual
  • routes (sequence of mappings)

Configure static routing for the device — see the Routing section below. 👑

  • routing-policy (sequence of mappings)

Configure policy routing for the device — see the Routing section below. 👑

  • neigh-suppress (scalar) — since 0.105

Takes a boolean. Configures whether ARP and ND neighbour suppression is enabled for this bridge port. When it's unset, the kernel's default gets used. 🎀

  • hairpin (scalar) — since 1.0

Takes a boolean. Configures whether traffic can be sent back out of the bridge port it was received on. When this flag is false, the bridge doesn't forward traffic back out of the receiving port. When unset, the back end default is used. 🌸

  • port-mac-learning (scalar) — since 1.0

Takes a boolean. Configures whether MAC address learning is enabled for this bridge port. When unset, the kernel default is used. Currently only supported on the networkd back end. 🧚

DHCP Overrides

Okay so there are a bunch of DHCP behaviour overrides. Most of them currently only do anything when you're on the networkd back end — the exceptions are use-routes and route-metric. ⭐

Overrides only do anything if the matching dhcp4 or dhcp6 is set to true. Like, they literally can't override a thing that isn't running.

If both dhcp4 and dhcp6 are true, the networkd back end requires that dhcp4-overrides and dhcp6-overrides contain the same keys and values. If the values don't match, you get an error and the network configuration just… won't be applied. 😭

When you're using the NetworkManager back end though, you can give different values for dhcp4-overrides and dhcp6-overrides, and they'll be applied to the DHCP client processes exactly as specified in the Netplan YAML.

  • dhcp4-overrides, dhcp6-overrides (mapping)

The dhcp4-overrides and dhcp6-override mappings override the default DHCP behaviour. 💜

  • use-dns (boolean)

Default: true. When true, the DNS servers you get from the DHCP server will be used, and they take precedence over any statically configured ones. Currently only has an effect on the networkd back end.

  • use-ntp (boolean)

Default: true. When true, the NTP servers you get from the DHCP server will be used by systemd-timesyncd, taking precedence over any statically configured ones. Currently only has an effect on the networkd back end.

  • send-hostname (boolean)

Default: true. When true, the machine hostname gets sent to the DHCP server. Currently only has an effect on the networkd back end.

  • use-hostname (boolean)

Default: true. When true, the hostname you get from the DHCP server gets set as the transient hostname of the system. Currently only has an effect on the networkd back end.

  • use-mtu (boolean)

Default: true. When true, the MTU you get from the DHCP server gets set as the MTU of the network interface. When false, the MTU advertised by the DHCP server is ignored. Currently only has an effect on the networkd back end.

  • hostname (scalar)

Use this value as the hostname sent to the DHCP server, instead of the machine's hostname. Currently only has an effect on the networkd back end.

  • use-routes (boolean)

Default: true. When true, the routes you get from the DHCP server get installed in the routing table like normal. When it's false, routes from the DHCP server are ignored — and in that case you're responsible for adding static routes if you need them for the network to work properly. This lets you avoid installing a default gateway on interfaces configured via DHCP. Available on both the networkd and NetworkManager back ends. 🔥

  • route-metric (scalar)

Use this value as the default metric for automatically-added routes. Handy for prioritising routes per device — just set a lower metric on the preferred interface. Available on both the networkd and NetworkManager back ends.

  • use-domains (scalar) — since 0.98

Takes a boolean, or the special value route. When true, the domain name you get from the DHCP server is used as a DNS search domain over this link — similar to what the Domains= setting does. If you set it to route, the domain name from the DHCP server is used for routing DNS queries only, not for searching, similar to how Domains= behaves when the argument is prefixed with ~ (tilde).

Requires feature: dhcp-use-domains

IPv6 Router Advertisement Overrides

Overrides for IPv6 Router Advertisement (RA) behaviour — only supported with the networkd back end. 🛰️

  • ra-overrides (mapping) — since 1.1

The ra-overrides mappings override the default IPv6 Router Advertisement behaviour.

  • use-dns (boolean)

Default: true. When true, the DNS servers you get from the Router Advertisement will be used. Currently only has an effect on the networkd back end.

  • use-domains (scalar)

Takes a boolean, or the special value route. When true, the domain name you get from the Router Advertisement is used as a DNS search domain over this link. If you set it to route, the domain name from the IPv6 RA is used for routing DNS queries only, not for searching. Defaults to false.

  • table (scalar)

Okay so this next bit is still finishing the thought from the option right before it: the routing table number for routes received in the IPv6 RA. Allowed values are positive integers starting from 1 💅. Some values are already in use to refer to specific routing tables: see {/etc,/usr/share}/iproute2/rt_tables — yes, both paths, bestie.

Routing 🧭

Complex routing is possible with Netplan. Standard static routes as well as policy routing using routing tables are supported via the networkd back end. These options are available for all types of interfaces. No cap, it's giving very capable. ✨

Default routes 🛣️

The most common need for routing concerns the definition of default routes to reach the wider internet. Those default routes can only be defined once per IP family and routing table — so don't try to sneak in doubles, bestie, it will not slay. 😭

A typical example would look like the following:

network:
  ethernets:
    eth0:
      [...]
      routes:
        - to: default # could be 0.0.0.0/0 optionally
          via: 10.0.0.1
          metric: 100
          on-link: true
          advertised-mss: 1400
        - to: default # could be ::/0 optionally
          via: cf02:de:ad:be:ef::2
    eth1:
      [...]
      routes:
        - to: default
          via: 172.134.67.1
          metric: 100
          on-link: true
          # Not on the main routing table,
          # does not conflict with the eth0 default route
      table: 76
  • routes (mapping) — this block defines standard static routes for an interface. At least to must be specified. If type is local or nat, a default scope of host is assumed. If type is unicast and no gateway (via) is given or type is broadcast, multicast or anycast, a default scope of link is assumed. Otherwise, a global scope is the default setting.
    • For from, to and via, both IPv4 and IPv6 addresses are recognised, and must be in the form addr/prefixlen or addr.
    • from (scalar) — set a source IP address for traffic going through the route. (NetworkManager: as of v1.8.0)
    • to (scalar) — destination address for the route.
    • via (scalar) — address to the gateway to use for this route.
    • on-link (boolean) — when set to true, specifies that the route is directly connected to the interface. (NetworkManager: as of v1.12.0 for IPv4 and v1.18.0 for IPv6)
    • metric (scalar) — the relative priority of the route. Must be a positive integer value.
    • type (scalar) — the type of route. Valid options are unicast (default), anycast, blackhole, broadcast, local, multicast, nat, prohibit, throw, unreachable or xresolve.
    • scope (scalar) — the route scope, how wide-ranging it is to the network. Possible values are global, link, or host. Applies to IPv4 only.
    • table (scalar) — the table number to use for the route. In some scenarios, it may be useful to set routes in a separate routing table. It may also be used to refer to routing policy rules which also accept a table parameter. Allowed values are positive integers starting from 1. Some values are already in use to refer to specific routing tables: see /etc/iproute2/rt_tables. (NetworkManager: as of v1.10.0)
    • mtu (scalar) – since 0.101 — the MTU to be used for the route, in bytes. Must be a positive integer value.
    • congestion-window (scalar) – since 0.102 — the congestion window to be used for the route, represented by number of segments. Must be a positive integer value.
    • advertised-receive-window (scalar) – since 0.102 — the receive window to be advertised for the route, represented by number of segments. Must be a positive integer value.
    • advertised-mss (scalar) – since 1.1 — the Maximum MSS ('Maximal Segment Size') to advertise to these destinations when establishing TCP connections. If it is not given, Linux uses a default value calculated from the first hop device MTU. Must be a positive integer.
  • routing-policy (mapping) — this block defines extra routing policy for a network, where traffic may be handled specially based on the source IP, firewall marking, etc.
    • For from, to, both IPv4 and IPv6 addresses are recognised, and must be in the form addr/prefixlen or addr.
    • from (scalar) — set a source IP address to match traffic for this policy rule.
    • to (scalar) — match on traffic going to the specified destination.
    • table (scalar) — the table number to match for the route. In some scenarios, it may be useful to set routes in a separate routing table. It may also be used to refer to routes which also accept a table parameter. Allowed values are positive integers starting from 1. Some values are already in use to refer to specific routing tables: see /etc/iproute2/rt_tables.
    • priority (scalar) — specify a priority for the routing policy rule, to influence the order in which routing rules are processed. A higher number means lower priority: rules are processed in order by increasing priority number. Specifying an explicit, unique, priority for each routing policy rule is strongly recommended and is mandatory on the NetworkManager back-end.
    • mark (scalar) — have this routing policy rule match on traffic that has been marked by the iptables firewall with this value. Allowed values are positive integers starting from 1.
    • type-of-service (scalar) — match this policy rule based on the type of service number applied to the traffic.

Authentication 🔐

Netplan supports advanced authentication settings for Ethernet and Wi-Fi interfaces, as well as individual Wi-Fi networks, by means of the auth block. It's giving security era, honestly. 🫶

  • auth (mapping) — specifies authentication settings for a device of type ethernets:, or an access-points: entry on a wifis: device. The auth block supports the following properties:
    • key-management (scalar) — the supported key management modes are none (no key management); psk (WPA with pre-shared key, common for home Wi-Fi); psk-sha256 (WPA2 with pre-shared key, common for home Wi-Fi); eap (WPA with EAP, common for enterprise Wi-Fi); eap-sha256 (used with WPA3-Enterprise); eap-suite-b-192 (used with WPA3-Enterprise); sae (used by WPA3); and 802.1x (used primarily for wired Ethernet connections).
    • password (scalar) — the password string for EAP, or the pre-shared key for WPA-PSK.
    • Conditional glow-up: the following properties can be used if key-management is eap or 802.1x:
      • method (scalar) — the EAP method to use. The supported EAP methods are tls (TLS), peap (Protected EAP), leap (Lightweight EAP), pwd (EAP Password) and ttls (Tunnelled TLS).
      • identity (scalar) — the identity to use for EAP.
      • anonymous-identity (scalar) — the identity to pass over the unencrypted channel if the chosen EAP method supports passing a different tunnelled identity.
      • ca-certificate (scalar) — path to a file with one or more trusted certificate authority (CA) certificates.
      • client-certificate (scalar) — path to a file containing the certificate to be used by the client during authentication.
      • client-key (scalar) — path to a file containing the private key corresponding to client-certificate.
      • client-key-password (scalar) — password to use to decrypt the private key specified in client-key if it is encrypted.
      • phase2-auth (scalar) – since 0.99 — phase 2 authentication mechanism.

Properties for device type ethernets 🌐

Status: Optional.

Purpose: Use the ethernets key to configure Ethernet interfaces.

Structure: The key consists of a mapping of Ethernet interface IDs. Each ethernet has a number of configuration options. You don't need to define each interface by their name inside the ethernets mapping. You can use any ID that describes the interface and match the actual network card using the match key. The general configuration structure for Ethernet is shown below.

network:
  ethernets:
    device-id:
      ...

device-id is the interface identifier. If you use the interface name as the ID, Netplan will match that interface.

Consider the example below. In this case, an interface called eth0 will be configured with DHCP.

network:
  ethernets:
    eth0:
      dhcp4: true

The device-id can be any descriptive name you find meaningful. Although, if it doesn't match a real interface name, you must use the property match to identify the device you want to configure.

The example below defines an Ethernet connection called isp-interface (supposedly an external interface connected to the Internet Service Provider) and uses match to apply the configuration to the physical device with MAC address aa:bb:cc:00:11:22.

network:
  ethernets:
    isp-interface:
      match:
        macaddress: aa:bb:cc:00:11:22
      dhcp4: true

Ethernet device definitions, beyond common ones described above, also support some additional properties that can be used for SR-IOV devices. Fancy, fancy. 💅

  • link (scalar) – since 0.99 — (SR-IOV devices only) The link property declares the device as a Virtual Function of the selected Physical Function device, as identified by the given Netplan ID.

Example:

network:
  ethernets:
    enp1: {...}
    enp1s16f1:
      link: enp1
  • virtual-function-count (scalar) – since 0.99
  • (still on virtual-function-count, from back up there 💅) — (SR-IOV devices only) In certain special cases VFs might need to be configured outside of Netplan. For such configurations virtual-function-count can be optionally used to set an explicit number of Virtual Functions for the given Physical Function. If unset, the default is to create only as many VFs as are defined in the Netplan configuration. This should be used for special cases only.

Requires feature: sriov

  • embedded-switch-mode (scalar) – since 0.104

(SR-IOV devices only) Change the operational mode of the embedded switch of a supported SmartNIC PCI device (e.g. Mellanox ConnectX-5). Possible values are switchdev or legacy, if unspecified the vendor's default configuration is used.

Requires feature: eswitch-mode

  • delay-virtual-functions-rebind (boolean) – since 0.104

(SR-IOV devices only) Delay rebinding of SR-IOV virtual functions to its driver after changing the embedded-switch-mode setting to a later stage. Can be enabled when bonding/VF LAG is in use. Defaults to false.

Requires feature: eswitch-mode

  • infiniband-mode (scalar) – since 0.105

(InfiniBand devices only) Change the operational mode of a IPoIB device. Possible values are datagram or connected. If unspecified the kernel's default configuration is used.

Requires feature: infiniband

🎀 Properties for device type modems 🎀

Status: Optional.

Purpose: Use the modems key to configure modem interfaces. GSM/CDMA modem configuration is only supported for the NetworkManager back end. systemd-networkd does not support modems. Like, at all. Don't ask it to, bestie.

Structure: The key consists of a mapping of modem IDs. Each modem has a number of configuration options. The general configuration structure for Modems is shown below.

network:
  version: 2
  renderer: NetworkManager
  modems:
    cdc-wdm1:
      mtu: 1600
      apn: ISP.CINGULAR
      username: ISP@CINGULARGPRS.COM
      password: CINGULAR1
      number: "*99#"
      network-id: 24005
      device-id: da812de91eec16620b06cd0ca5cbc7ea25245222
      pin: 2345
      sim-id: 89148000000060671234
      sim-operator-id: 310260

Requires feature: modems

  • apn (scalar) – since 0.99

Set the carrier APN (Access Point Name). This can be omitted if auto-config is enabled.

  • auto-config (boolean) – since 0.99

Specify whether to try and auto-configure the modem by doing a lookup of the carrier against the Mobile Broadband Provider database. Heads up though: this may not work for all carriers 🫠.

  • device-id (scalar) – since 0.99

Specify the device ID (as given by the WWAN management service) of the modem to match. You can find this using mmcli.

  • network-id (scalar) – since 0.99

Specify the Network ID (GSM LAI format). If this is specified, the device will not roam networks. Periodt, it's staying home.

  • number (scalar) – since 0.99

The number to dial to establish the connection to the mobile broadband network. (Deprecated for GSM 👀)

  • password (scalar) – since 0.99

Specify the password used to authenticate with the carrier network. This can be omitted if auto-config is enabled.

  • pin (scalar) – since 0.99

Specify the SIM PIN to allow it to operate if a PIN is set.

  • sim-id (scalar) – since 0.99

Specify the SIM unique identifier (as given by the WWAN management service) which this connection applies to. If given, the connection will apply to any device also allowed by device-id which contains a SIM card matching the given identifier.

  • sim-operator-id (scalar) – since 0.99

Specify the MCC/MNC string (such as 310260 or 21601) which identifies the carrier that this connection should apply to. If given, the connection will apply to any device also allowed by device-id and sim-id which contains a SIM card provisioned by the given operator.

  • username (scalar) – since 0.99

Specify the username used to authenticate with the carrier network. This can be omitted if auto-config is enabled.

🌸 Properties for device type wifis 🌸

Status: Optional.

Purpose: Use the wifis key to configure Wi-Fi access points. Delulu-free Wi-Fi, we love to see it ✨

Structure: The key consists of a mapping of Wi-Fi IDs. Each wifi has a number of configuration options. The general configuration structure for Wi-Fi is shown below.

network:
  version: 2
  wifis:
    wlp0s1:
      access-points:
        "network_ssid_name":
          password: "**********"
Note that systemd-networkd does not have native support Wi-Fi, so you need wpasupplicant installed if you let the networkd renderer handle Wi-Fi. Iykyk 🧚
  • access-points (mapping)

This provides pre-configured connections to NetworkManager. Note that users can of course select other access points/SSIDs. The keys of the mapping are the SSIDs, and the values are mappings with the following supported properties:

  • password (scalar)

Enable WPA/WPA2 authentication and set the passphrase for it. If neither this nor an auth block are given, the network is assumed to be open. The setting

password: "S3kr1t"

is equivalent to

auth:
  key-management: psk
  password: "S3kr1t"
  • mode (scalar)

Possible access point modes are infrastructure (the default), ap (create an access point to which other devices can connect), and adhoc (peer to peer networks without a central access point). ap is only supported with NetworkManager.

  • bssid (scalar) – since 0.99

If specified, directs the device to only associate with the given access point. Loyalty 💅

  • band (scalar) – since 0.99

Possible bands are 5GHz (for 5GHz 802.11a) and 2.4GHz (for 2.4GHz 802.11), do not restrict the 802.11 frequency band of the network if unset (the default).

  • channel (scalar) – since 0.99

Wireless channel to use for the Wi-Fi connection. Because channel numbers overlap between bands, this property takes effect only if the band property is also set.

  • hidden (boolean) – since 0.100

Set to true to change the SSID scan technique for connecting to hidden Wi-Fi networks. Note this may have slower performance compared to false (the default) when connecting to publicly broadcast SSIDs. Tradeoffs, bestie 😭

  • wakeonwlan (sequence of scalars) – since 0.99

This enables WakeOnWLan on supported devices. Not all drivers support all options. May be any combination of any, disconnect, magic_pkt, gtk_rekey_failure, eap_identity_req, four_way_handshake, rfkill_release or tcp (NetworkManager only). Or the exclusive default flag (the default).

  • regulatory-domain (scalar) – since 0.105

This can be used to define the radio's regulatory domain, to make use of additional Wi-Fi channels outside the "world domain". Takes an ISO/ IEC 3166 country code (like GB) or 00 to reset to the "world domain". See wireless-regdb for available values.

Requires dependency: iw, if it is to be used outside the networkd (wpasupplicant) back end.

🦋 Properties for device type bridges 🦋

Status: Optional.

Purpose: Use the bridges key to create Bridge interfaces. Group project but, like, a functional one 🫶

Structure: The key consists of a mapping of Bridge interface names. Each bridge has an optional list of interfaces that will be bridged together. The interfaces listed in the interfaces key (enp5s0 and enp5s1 below) must also be defined in your Netplan configuration. The general configuration structure for Bridges is shown below.

network:
  bridges:
    br0:
      interfaces:
        - enp5s0
        - enp5s1
      dhcp4: true
      ...

When applied, a virtual interface of type bridge called br0 will be created in the system. Main character interface, honestly 👑

The specific settings for bridges are defined below.

  • interfaces (sequence of scalars)

All devices matching this ID list will be added to the bridge. This may be an empty list, in which case the bridge will be brought online with no member interfaces.

Example:

network:
  ethernets:
    switchports:
      match: {name: "enp2*"}
  [...]
  bridges:
    br0:
      interfaces: [switchports]
  • parameters (mapping)

Customisation parameters for special bridging options. Time intervals may need to be expressed as a number of seconds or milliseconds: the default value type is specified below. If necessary, time intervals can be qualified using a time suffix (such as s for seconds, ms for milliseconds) to allow for more control over its behaviour.

  • ageing-time, aging-time (scalar)

Set the period of time to keep a MAC address in the forwarding database after a packet is received. This maps to the AgeingTimeSec= property when the networkd renderer is used. If no time suffix is specified, the value will be interpreted as seconds.

  • priority (scalar)

Set the priority value for the bridge. This value should be a number between 0 and 65535. Lower values mean higher priority. The bridge with the higher priority will be elected as the root bridge. Root bridge energy ✨

  • port-priority (mapping)

Set the port priority per interface. The priority value is a number between 0 and 63. This metric is used in the designated port and root port selection algorithms.

Example:

network:
  ethernets:
    eth0:
      dhcp4: false
    eth1:
      dhcp4: false
  bridges:
    br0:
      interfaces: [eth0, eth1]
      parameters:
        port-priority:
          eth0: 10
          eth1: 20
  • forward-delay (scalar) — bestie this is the bridge's "let me think about it" era 💭. It's the amount of time the bridge chills in the Listening and Learning states before it finally reaches the Forwarding state. This field maps to the ForwardDelaySec= property for the networkd renderer. And if you don't give it a time suffix, it just gets read as seconds, no cap.
  • hello-time (scalar) — the gap between two hello packets getting sent out from the root and designated bridges. Those hello packets are literally gossiping about the network topology, iykyk. When the networkd renderer is used, this maps to the HelloTimeSec= property. No time suffix? It's seconds, periodt.
  • max-age (scalar) — sets the maximum age of a hello packet. If the last hello packet is older than that value, the bridge is like "okay I guess I'm the main character now" and tries to become the root bridge. Maps to the MaxAgeSec= property when the networkd renderer is used. No time suffix → interpreted as seconds.
  • path-cost (mapping) — sets the per-interface cost of a path on the bridge. Faster interfaces should get a lower cost, obviously — it's giving meritocracy. This gives you finer control over the network topology so the fastest paths are available whenever possible. Slay.

Example:

network:
  ethernets:
    eth0:
      dhcp4: false
    eth1:
      dhcp4: false
  bridges:
    br0:
      interfaces: [eth0, eth1]
      parameters:
        path-cost:
          eth0: 100
          eth1: 200
  • stp (boolean) — do we use Spanning Tree Protocol on this bridge, yes or yes? The default value is true, which means Spanning Tree should be used. 🎀

🌈 Properties for device type dummy-devices 🌈

Status: Optional.

Purpose: use the dummy-devices key to create virtual interfaces, bestie.

Structure: the key consists of a mapping of interface names. Dummy devices are virtual devices you can route packets to without actually transmitting them — lowkey phantom vibes, highkey useful 👻.

network:
  dummy-devices:
    dm0:
      addresses:
        - 192.168.0.123/24
      ...

When applied, a virtual interface called dm0 will be created in the system.

See the "Properties for all device types" section for the list of properties that can be used with this type of interface.

🔗 Properties for device type bonds 🔗

Status: Optional.

Purpose: use the bonds key to create Bond (Link Aggregation) interfaces — multiple links holding hands so they count as one 🫶.

Structure: the key consists of a mapping of Bond interface names. Each bond has an optional list of interfaces that will be part of the aggregation. The interfaces listed in the interfaces key must also be defined in your Netplan configuration — no ghost members, bestie. The general configuration structure for Bonds is shown below.

network:
  bonds:
    bond0:
      interfaces:
        - enp5s0
        - enp5s1
        - enp5s2
      parameters:
        mode: active-backup
      ...

When applied, a virtual interface of type bond called bond0 will be created in the system.

The specific settings for bonds are defined below.

  • interfaces (sequence of scalars) — all devices matching this ID list will be added to the bond. All of them. The whole friend group.

Example:

network:
  ethernets:
    switchports:
      match: {name: "enp2*"}
  [...]
  bonds:
    bond0:
      interfaces: [switchports]
  • parameters (mapping) — customisation parameters for special bonding options. Time intervals may need to be expressed as a number of seconds or milliseconds: the default value type is specified below for each one. If necessary, time intervals can be qualified using a time suffix (such as s for seconds, ms for milliseconds) so you get more control over its behaviour. Fancy ✨
  • mode (scalar) — sets the bonding mode used for the interfaces. The default is balance-rr (round robin). Possible values are balance-rr, active-backup, balance-xor, broadcast, 802.3ad, balance-tlb and balance-alb. For Open vSwitch, active-backup and the additional modes balance-tcp and balance-slb are supported.
  • lacp-rate (scalar) — sets the rate at which LACPDUs are transmitted. This is only useful in 802.3ad mode. Possible values are slow (30 seconds, default), and fast (every second). Pick your tempo 🎶
  • mii-monitor-interval (scalar) — specifies the interval for MII monitoring (verifying if an interface of the bond has carrier). The default is 0, which disables MII monitoring entirely. This is equivalent to the MIIMonitorSec= field for the networkd back end. If no time suffix is specified, the value will be interpreted as milliseconds.
  • min-links (scalar) — the minimum number of links up in a bond to consider the bond interface to be up. Like a group project where you need at least this many people to show up 😭
  • transmit-hash-policy (scalar) — specifies the transmit hash policy for the selection of ports. This is only useful in balance-xor, 802.3ad and balance-tlb modes. Possible values are layer2, layer3+4, layer2+3, encap2+3 and encap3+4.
  • ad-select (scalar) — sets the aggregation selection mode. Possible values are stable, bandwidth and count. This option is only used in 802.3ad mode.
  • all-members-active (boolean) – since 0.106 — if the bond should drop duplicate frames received on inactive ports, set this option to false. If they should be delivered, set it to true. The default value is false and is the desirable behaviour in most situations.

Alias: all-slaves-active

  • arp-interval (scalar) — sets the interval value for how frequently ARP link monitoring should happen. The default value is 0, which disables ARP monitoring. For the networkd back end, this maps to the ARPIntervalSec= property. If no time suffix is specified, the value will be interpreted as milliseconds.
  • arp-ip-targets (sequence of scalars) — IP addresses of other hosts on the link which should be sent ARP requests in order to validate that a port is up. This option is only used when arp-interval is set to a value other than 0. At least one IP address must be given for ARP link monitoring to function. Only IPv4 addresses are supported. You can specify up to 16 IP addresses. The default value is an empty list.
  • arp-validate (scalar) — configure how ARP replies are to be validated when using ARP link monitoring. Possible values are none, active, backup, and all.
  • arp-all-targets (scalar) — specify whether to use any ARP IP target being up as sufficient for a port to be considered up; or if all the targets must be up. This is only used for active-backup mode when arp-validate is enabled. Possible values are any and all.
  • up-delay (scalar) — specifies the delay before enabling a link once the link is physically up. The default value is 0. This maps to the UpDelaySec= property for the networkd renderer. This option is only valid for the miimon link monitor. If no time suffix is specified, the value will be interpreted as milliseconds.
  • down-delay (scalar) — specifies the delay before disabling a link once the link has been lost. The default value is 0. This maps to the DownDelaySec= property for the networkd renderer. This option is only valid for the miimon link monitor. If no time suffix is specified, the value will be interpreted as milliseconds.
  • fail-over-mac-policy (scalar) — sets whether to set all ports to the same MAC address when adding them to the bond, or how else the system should handle MAC addresses. The possible values are none, active and follow.
  • gratuitous-arp (scalar) — specifies how many ARP packets to send after failover. Once a link is up on a new port, a notification is sent and possibly repeated if this value is set to a number greater than 1. The default value is 1 and valid values are between 1 and 255. This only affects active-backup mode. Announcing yourself that many times, iconic 💅

For historical reasons, the misspelling gratuitious-arp is also accepted and has the same function. (Not a typo we made up, bestie — it's legit legacy 👑)

  • packets-per-member (scalar) – since 0.106 — in balance-rr mode, specifies the number of packets to transmit on a port before switching to the next. When this value is set to 0, ports are chosen at random. Allowable values are between 0 and 65535. The default value is 1. This setting is only used in balance-rr mode.

Alias: packets-per-slave

  • primary-reselect-policy (scalar) — sets the reselection policy for the primary port. On failure of the active port, the system will use this policy to decide how the new active port will be chosen and how recovery will be handled. The possible values are always, better and failure.
  • resend-igmp (scalar) — in modes balance-rr, active-backup, balance-tlb and balance-alb, a failover can switch IGMP traffic from one port to another.

This parameter is the “how many IGMP membership reports do we fire off on a failover?” knob. Values go from 0 to 255. A big fat 0 disables sending membership reports. Otherwise, the first membership report goes out on failover, and the follow-up reports go out at 200ms intervals.

  • learn-packet-interval (scalar) — Sets the gap between learning packets sent to each port. The value range is between 1 and 0x7fffffff. The default value is 1. This option only affects balance-tlb and balance-alb modes. Using the networkd renderer, this field maps to the LearnPacketIntervalSec= property. If no time suffix is specified, the value will be interpreted as seconds.
  • primary (scalar) — Name a device to be the primary port, or the preferred device to use as a port for the bond (i.e. the preferred device to send data through), whenever it is available. This only affects active-backup, balance-alb and balance-tlb modes.

🌈 Properties for device type tunnels 🫶

Status: Optional. ✨

Purpose: Use the tunnels key to create virtual tunnel interfaces.

Structure: The key consists of a mapping of tunnel interface names. Each tunnel requires the identification of the tunnel mode (see the section mode below for the list of supported modes). The general configuration structure for Tunnels is shown below.

network:
  tunnels:
    tunnel0:
      mode: SCALAR
      ...

When applied, a virtual interface called tunnel0 will be created in the system. Its operation mode is defined by the property mode.

Tunnels let traffic pass as if it was between systems on the same local network, even though those systems may be far from each other but reachable via the Internet. They may be used to support IPv6 traffic on a network where the ISP does not provide the service, or to extend and “connect” separate local networks. See Tunneling_protocol for more general information about tunnels.

The specific settings for tunnels are defined below.

  • mode (scalar) — Defines the tunnel mode. Valid options are sit, gre, ip6gre, ipip, ipip6, ip6ip6, vti, vti6, wireguard, vxlan, gretap and ip6gretap modes. In addition, the NetworkManager back end supports isatap tunnels.
  • local (scalar) — Defines the address of the local endpoint of the tunnel. (For VXLAN) This should match one of the parent’s IP addresses or make use of the networkd special values.
  • remote (scalar) — Defines the address of the remote endpoint of the tunnel or multicast group IP address for VXLAN.
  • ttl (scalar) – since 0.103 — Defines the Time To Live (TTL) of the tunnel. Takes a number in the range 1..255.
  • key (scalar or mapping) — Define keys to use for the tunnel. The key can be a number or a dotted quad (an IPv4 address). For wireguard it can be a base64-encoded private key or (as of networkd v242+) an absolute path to a file, containing the private key (since 0.100). It is used for identification of IP transforms. This is only required for vti and vti6 when using the networkd back end.

This field may be used as a scalar (meaning that a single key is specified and to be used for input, output and private key), or as a mapping, where you can further specify input/output/private.

  • input (scalar) — The input key for the tunnel.
  • output (scalar) — The output key for the tunnel.
  • private (scalar) – since 0.100 — A base64-encoded private key required for WireGuard tunnels. When the systemd-networkd back end (v242+) is used, this can also be an absolute path to a file containing the private key.
  • private-key-flags (sequence of scalars) – since 0.107 — Private key flags used by NetworkManager. Possible values are: agent-owned, not-saved and not-required.
    • agent-owned: a user-session secret agent is responsible for providing and storing this secret.
    • not-saved: this secret should not be saved but should be requested from the user each time it is required.
    • not-required: this flag hints that the secret is not required and should not be requested from the user.

Example:

network:
  renderer: NetworkManager
  tunnels:
    wg0:
      mode: wireguard
      port: 5182
      key:
        private-key-flags:
          - agent-owned
      peers:
        - keys:
            public: rlbInAj0qV69CysWPQY7KEBnKxpYCpaWqOs/dLevdWc=
          allowed-ips: [0.0.0.0/0, "2001:fe:ad:de:ad:be:ef:1/24"]
          keepalive: 23
          endpoint: 1.2.3.4:5
  • keys (scalar or mapping) — Alternate name for the key field. See above.

Examples:

network:
  tunnels:
    tun0:
      mode: gre
      local: ...
      remote: ...
      keys:
        input: 1234
        output: 5678
network:
  tunnels:
    tun0:
      mode: vti6
      local: ...
      remote: ...
      key: 59568549
network:
  tunnels:
    wg0:
      mode: wireguard
      addresses: [...]
      peers:
        - keys:
            public: rlbInAj0qV69CysWPQY7KEBnKxpYCpaWqOs/dLevdWc=
            shared: /path/to/shared.key
          ...
      key: mNb7OIIXTdgW4khM7OFlzJ+UPs7lmcWHV7xjPgakMkQ=
network:
  tunnels:
    wg0:
      mode: wireguard
      addresses: [...]
      peers:
        - keys:
            public: rlbInAj0qV69CysWPQY7KEBnKxpYCpaWqOs/dLevdWc=
          ...
      keys:
        private: /path/to/priv.key

💜 WireGuard specific keys 💜

  • mark (scalar) – since 0.100 — Firewall mark for outgoing WireGuard packets from this interface, optional.
  • port (scalar) – since 0.100 — UDP port to listen at or auto. Optional, defaults to auto.
  • peers (sequence of mappings) – since 0.100 — A list of peers, each having keys documented below.

Example:

network:
  tunnels:
    wg0:
      mode: wireguard
      key: /path/to/private.key
      mark: 42
      port: 5182
      peers:
        - keys:
            public: rlbInAj0qV69CysWPQY7KEBnKxpYCpaWqOs/dLevdWc=
          allowed-ips: [0.0.0.0/0, "2001:fe:ad:de:ad:be:ef:1/24"]
          keepalive: 23
          endpoint: 1.2.3.4:5
        - keys:
            public: M9nt4YujIOmNrRmpIRTmYSfMdrpvE7u6WkG8FY8WjG4=
            shared: /some/shared.key
          allowed-ips: [10.10.10.20/24]
          keepalive: 22
          endpoint: 5.4.3.2:1
  • endpoint (scalar) – since 0.100 — Remote endpoint IPv4/IPv6 address or a hostname, followed by a colon and a port number.
  • allowed-ips (sequence of scalars) – since 0.100 — A list of IP (v4 or v6) addresses with CIDR masks from which this peer is allowed to send incoming traffic and to which outgoing traffic for this peer is directed. The catch-all 0.0.0.0/0 may be specified for matching all IPv4 addresses, and ::/0 may be specified for matching all IPv6 addresses.
  • keepalive (scalar) – since 0.100 — An interval in seconds, between 1 and 65535 inclusive, of how often to send an authenticated empty packet to the peer for the purpose of keeping a stateful firewall or NAT mapping valid persistently. Optional.
  • keys (mapping) – since 0.100 — Define keys to use for the WireGuard peers.

This field can be used as a mapping, where you can further specify the public and shared keys.

  • public (scalar) – since 0.100 — A base64-encoded public key, required for WireGuard peers.
  • shared (scalar) – since 0.100 — A base64-encoded pre-shared key. Optional for WireGuard peers. When the systemd-networkd back end (v242+) is used, this can also be an absolute path to a file containing the pre-shared key.

🦋 VXLAN specific keys 🦋

  • id (scalar) – since 0.105 — The VXLAN Network Identifier (VNI or VXLAN Segment ID). Takes a number in the range 1..16777215.
  • link (scalar) – since 0.105 — Netplan ID of the parent device definition to which this VXLAN gets connected.
  • type-of-service (scalar) – since 0.105 — The Type Of Service byte value for a VXLAN interface.
  • mac-learning (scalar) – since 0.105 — Takes a boolean. When true, enables dynamic MAC learning to discover remote MAC addresses.
  • ageing, aging (scalar) – since 0.105 — The lifetime of Forwarding Database entry learned by the kernel, in seconds.
  • limit (scalar) – since 0.105 — Configures maximum number of FDB entries.
  • arp-proxy (scalar) – since 0.105 — Takes a boolean. When true, bridge-connected VXLAN tunnel endpoint answers ARP requests from the local bridge on behalf of remote Distributed Overlay Virtual Ethernet (DOVE) clients. Defaults to false.
  • notifications (sequence of scalars) – since 0.105 — Takes the flags l2-miss and l3-miss to enable netlink LLADDR and/or netlink IP address miss notifications.
  • short-circuit (scalar) – since 0.105 — Takes a boolean. When true, route short circuiting is turned on.
  • checksums (sequence of scalars) – since 0.105

…and for the VXLAN checksum vibes, it takes the flags udp, zero-udp6-tx, zero-udp6-rx, remote-tx and remote-rx to enable transmitting UDP checksums in VXLAN/IPv4, send/receive zero checksums in VXLAN/IPv6 and enable sending/receiving checksum offloading in VXLAN. slay, networking bestie ✨

  • extensions (sequence of scalars) – since 0.105 — Takes the flags group-policy and generic-protocol to enable the "Group Policy" and/or "Generic Protocol" VXLAN extensions.
  • port (scalar) – since 0.105 — Configures the default destination UDP port. If the destination port is not specified then the Linux kernel default will be used. Set to 4789 to get the IANA assigned value.
  • port-range (sequence of scalars) – since 0.105 — Configures the source port range for the VXLAN. The kernel assigns the source UDP port based on the flow to help the receiver to do load balancing. When this option is not set, the normal range of local UDP ports is used. Uses the form [LOWER, UPPER].
  • flow-label (scalar) – since 0.105 — Specifies the flow label to use in outgoing packets. The valid range is 0-1048575.
  • do-not-fragment (scalar) – since 0.105 — Allows setting the IPv4 Do not Fragment (DF) bit in outgoing packets. Takes a boolean value. When unset, the kernel default will be used.

🦋 Properties for device type virtual-ethernets 🦋

Status: Optional.

Purpose: Use the virtual-ethernets key to create virtual Ethernet interfaces. iconic little tunnel gremlins 💅

Structure: The key consists of a mapping of veth interface names. Each veth requires a peer. In order to have a fully working veth pair, both devices must be defined, i.e., only setting the peer key with the peer name is not enough — the peer interface must also be defined and set the first one as its peer. The general configuration structure for virtual Ethernet is shown below.

network:
  virtual-ethernets:
    veth0:
      peer: veth1
    veth1:
      peer: veth0

When applied, two virtual interfaces called veth0 and veth1 will be created in the system.

Virtual Ethernet devices act as tunnels forwarding traffic from one interface to the other. They can be used to connect two separate virtual networks such as network namespaces and bridges. It's not possible to move virtual-ethernets to different namespaces through Netplan at the present moment. noted, bestie 🫶

The specific settings for virtual-ethernets are defined below.

  • peer (scalar) — Defines the virtual-ethernet peer. The peer interface must also be a virtual-ethernet device.

Below is a complete example that uses a pair of virtual Ethernet devices to create a link between two bridges:

network:
  version: 2
  renderer: networkd
  virtual-ethernets:
    veth0-peer1:
      peer: veth0-peer2
    veth0-peer2:
      peer: veth0-peer1

  bridges:
    br0:
      interfaces:
        - veth0-peer1
    br1:
      interfaces:
        - veth0-peer2

🌸 Properties for device type vlans 🌸

Status: Optional.

Purpose: Use the vlans key to create VLAN interfaces. VLAN girl summer 💜

Structure: The key consists of a mapping of VLAN interface names. The interface used in the link option (enp5s0 in the example below) must also be defined in the Netplan configuration. The general configuration structure for VLANs is shown below.

network:
  vlans:
    vlan123:
      id: 123
      link: enp5s0
      dhcp4: yes

The specific settings for VLANs are defined below.

  • id (scalar) — VLAN ID, a number between 0 and 4094.
  • link (scalar) — Netplan ID of the underlying device definition on which this VLAN gets created.

Example:

network:
  ethernets:
    eno1: {...}
  vlans:
    en-intra:
      id: 1
      link: eno1
      dhcp4: yes
    en-vpn:
      id: 2
      link: eno1
      addresses: [...]

🧚 Properties for device type vrfs 🧚

Status: Optional.

Purpose: Use the vrfs key to create Virtual Routing and Forwarding (VRF) interfaces. routing with main character energy 👑

Structure: The key consists of a mapping of VRF interface names. The interface used in the link option (enp5s0 in the example below) must also be defined in the Netplan configuration. The general configuration structure for VRFs is shown below.

network:
  renderer: networkd
  vrfs:
    vrf1:
      table: 1
      interfaces:
        - enp5s0
      routes:
        - to: default
          via: 10.10.10.4
      routing-policy:
        - from: 10.10.10.42
  • table (scalar) – since 0.105 — The numeric routing table identifier. This setting is compulsory.
  • interfaces (sequence of scalars) – since 0.105 — All devices matching this ID list will be added to the VRF. This may be an empty list, in which case the VRF will be brought online with no member interfaces.
  • routes (sequence of mappings) – since 0.105 — Configure static routing for the device; see the Routing section. The table value is implicitly set to the VRF table.
  • routing-policy (sequence of mappings) – since 0.105 — Configure policy routing for the device; see the Routing section. The table value is implicitly set to the VRF table.

Example:

network:
  vrfs:
    vrf20:
      table: 20
      interfaces: [ br0 ]
      routes:
        - to: default
          via: 10.10.10.3
      routing-policy:
        - from: 10.10.10.42
    [...]
  bridges:
    br0:
      interfaces: []

💀 Properties for device type nm-devices 💀

Status: Optional. Its use is not recommended.

Bestie, listen: this one is for internal fallback energy, not your normal config file. 🚨

Purpose: Use the nm-devices key to configure device types that are not supported by Netplan. This is NetworkManager specific configuration.

Structure: The key consists of a mapping of NetworkManager connections. The nm-devices device type is for internal use only and should not be used in normal configuration files. It enables a fallback mode for unsupported settings, using the passthrough mapping. The general configuration structure for NM connections is shown below.

network:
  version: 2
  nm-devices:
    NM-db5f0f67-1f4c-4d59-8ab8-3d278389cf87:
      renderer: NetworkManager
      networkmanager:
        uuid: "db5f0f67-1f4c-4d59-8ab8-3d278389cf87"
        name: "myvpnconnection"
        passthrough:
          connection.type: "vpn"
          vpn.ca: "path to ca.crt"
          vpn.cert: "path to client.crt"
          vpn.cipher: "AES-256-GCM"
          vpn.connection-type: "tls"
          vpn.dev: "tun"
          vpn.key: "path to client.key"
          vpn.remote: "1.2.3.4:1194"
          vpn.service-type: "org.freedesktop.NetworkManager.openvpn"

🌈 Back end-specific configuration parameters 🌈

In addition to the other fields available to configure interfaces, some back ends may require to record some of their own parameters in Netplan, especially if the Netplan definitions are generated automatically by the consumer of that back end. Currently, this is only used with NetworkManager. iykyk ✨

  • networkmanager (mapping) – since 0.99 — Keeps the NetworkManager-specific configuration parameters used by the daemon to recognise connections.
    • name (scalar) – since 0.99 — Set the display name for the connection.
    • uuid (scalar) – since 0.99 — Defines the UUID (unique identifier) for this connection, as generated by NetworkManager itself.
    • stable-id (scalar) – since 0.99 — Defines the stable ID (a different form of a connection name) used by NetworkManager in case the name of the connection might otherwise change, such as when sharing connections between users.
    • device (scalar) – since 0.99 — Defines the interface name for which this connection applies.
    • passthrough (mapping) – since 0.102 — Can be used as a fallback mechanism to missing key-file settings.

👀 SEE ALSO 👀

Go read these next, bestie — the whole Netplan cinematic universe is right here:

💖 AUTHORS 💖

  • Mathieu Trudel-Lapierre (cyphermox@ubuntu.com)
  • Martin Pitt (martin.pitt@ubuntu.com)
  • Lukas Märdian (slyon@ubuntu.com)

That’s the whole page, bestie — the footer carries the source/version and date, so peek right there for the exact release deets. 💅💖✨

📦 where this page comes from