1SYSTEMD.NETDEV(5) systemd.network SYSTEMD.NETDEV(5)
2
6 systemd.netdev - Virtual Network Device configuration
7
9 netdev.netdev
10
12 A plain ini-style text file that encodes configuration about a virtual
13 network device, used by systemd-networkd(8). See systemd.syntax(7) for
14 a general description of the syntax.
15 The main Virtual Network Device file must have the extension .netdev;
17 other extensions are ignored. Virtual network devices are created as
18 soon as networkd is started. If a netdev with the specified name
19 already exists, networkd will use that as-is rather than create its
20 own. Note that the settings of the pre-existing netdev will not be
21 changed by networkd.
22 The .netdev files are read from the files located in the system network
24 directory /usr/lib/systemd/network and /usr/local/lib/systemd/network,
25 the volatile runtime network directory /run/systemd/network and the
26 local administration network directory /etc/systemd/network. All
27 configuration files are collectively sorted and processed in
28 alphanumeric order, regardless of the directories in which they live.
29 However, files with identical filenames replace each other. It is
30 recommended that each filename is prefixed with a number (e.g.
31 10-vlan.netdev). Otherwise, .netdev files generated by systemd-network-
32 generator.service(8) may take precedence over user configured files.
33 Files in /etc/ have the highest priority, files in /run/ take
34 precedence over files with the same name in /usr/lib/. This can be used
35 to override a system-supplied configuration file with a local file if
36 needed. As a special case, an empty file (file size 0) or symlink with
37 the same name pointing to /dev/null disables the configuration file
38 entirely (it is "masked").
39 Along with the netdev file foo.netdev, a "drop-in" directory
41 foo.netdev.d/ may exist. All files with the suffix ".conf" from this
42 directory will be merged in the alphanumeric order and parsed after the
43 main file itself has been parsed. This is useful to alter or add
44 configuration settings, without having to modify the main configuration
45 file. Each drop-in file must have appropriate section headers.
46 In addition to /etc/systemd/network, drop-in ".d" directories can be
48 placed in /usr/lib/systemd/network or /run/systemd/network directories.
49 Drop-in files in /etc/ take precedence over those in /run/ which in
50 turn take precedence over those in /usr/lib/. Drop-in files under any
51 of these directories take precedence over the main netdev file wherever
52 located. (Of course, since /run/ is temporary and /usr/lib/ is for
53 vendors, it is unlikely drop-ins should be used in either of those
54 places.)
55
57 The following kinds of virtual network devices may be configured in
58 .netdev files:
59 Table 1. Supported kinds of virtual network devices
61 ┌──────────┬────────────────────────────┐
62 │Kind │ Description │
63 ├──────────┼────────────────────────────┤
64 │bond │ A bond device is an │
65 │ │ aggregation of all its │
66 │ │ slave devices. See Linux │
67 │ │ Ethernet Bonding Driver │
68 │ │ HOWTO[1] for details. │
69 ├──────────┼────────────────────────────┤
70 │bridge │ A bridge device is a │
71 │ │ software switch, and each │
72 │ │ of its slave devices and │
73 │ │ the bridge itself are │
74 │ │ ports of the switch. │
75 ├──────────┼────────────────────────────┤
76 │dummy │ A dummy device drops all │
77 │ │ packets sent to it. │
78 ├──────────┼────────────────────────────┤
79 │gre │ A Level 3 GRE tunnel over │
80 │ │ IPv4. See RFC 2784[2] for │
81 │ │ details. Name "gre0" │
82 │ │ should not be used, as the │
83 │ │ kernel creates a device │
84 │ │ with this name when the │
85 │ │ corresponding kernel │
86 │ │ module is loaded. │
87 ├──────────┼────────────────────────────┤
88 │gretap │ A Level 2 GRE tunnel over │
89 │ │ IPv4. Name "gretap0" │
90 │ │ should not be used, as the │
91 │ │ kernel creates a device │
92 │ │ with this name when the │
93 │ │ corresponding kernel │
94 │ │ module is loaded. │
95 ├──────────┼────────────────────────────┤
96 │erspan │ ERSPAN mirrors traffic on │
97 │ │ one or more source ports │
98 │ │ and delivers the mirrored │
99 │ │ traffic to one or more │
100 │ │ destination ports on │
101 │ │ another switch. The │
102 │ │ traffic is encapsulated in │
103 │ │ generic routing │
104 │ │ encapsulation (GRE) and is │
105 │ │ therefore routable across │
106 │ │ a layer 3 network between │
107 │ │ the source switch and the │
108 │ │ destination switch. Name │
109 │ │ "erspan0" should not be │
110 │ │ used, as the kernel │
111 │ │ creates a device with this │
112 │ │ name when the │
113 │ │ corresponding kernel │
114 │ │ module is loaded. │
115 ├──────────┼────────────────────────────┤
116 │ip6gre │ A Level 3 GRE tunnel over │
117 │ │ IPv6. │
118 ├──────────┼────────────────────────────┤
119 │ip6tnl │ An IPv4 or IPv6 tunnel │
120 │ │ over IPv6 │
121 ├──────────┼────────────────────────────┤
122 │ip6gretap │ A Level 2 GRE tunnel over │
123 │ │ IPv6. │
124 ├──────────┼────────────────────────────┤
125 │ipip │ An IPv4 over IPv4 tunnel. │
126 ├──────────┼────────────────────────────┤
127 │ipvlan │ An IPVLAN device is a │
128 │ │ stacked device which │
129 │ │ receives packets from its │
130 │ │ underlying device based on │
131 │ │ IP address filtering. │
132 ├──────────┼────────────────────────────┤
133 │ipvtap │ An IPVTAP device is a │
134 │ │ stacked device which │
135 │ │ receives packets from its │
136 │ │ underlying device based on │
137 │ │ IP address filtering and │
138 │ │ can be accessed using the │
139 │ │ tap user space interface. │
140 ├──────────┼────────────────────────────┤
141 │macvlan │ A macvlan device is a │
142 │ │ stacked device which │
143 │ │ receives packets from its │
144 │ │ underlying device based on │
145 │ │ MAC address filtering. │
146 ├──────────┼────────────────────────────┤
147 │macvtap │ A macvtap device is a │
148 │ │ stacked device which │
149 │ │ receives packets from its │
150 │ │ underlying device based on │
151 │ │ MAC address filtering. │
152 ├──────────┼────────────────────────────┤
153 │sit │ An IPv6 over IPv4 tunnel. │
154 ├──────────┼────────────────────────────┤
155 │tap │ A persistent Level 2 │
156 │ │ tunnel between a network │
157 │ │ device and a device node. │
158 ├──────────┼────────────────────────────┤
159 │tun │ A persistent Level 3 │
160 │ │ tunnel between a network │
161 │ │ device and a device node. │
162 ├──────────┼────────────────────────────┤
163 │veth │ An Ethernet tunnel between │
164 │ │ a pair of network devices. │
165 ├──────────┼────────────────────────────┤
166 │vlan │ A VLAN is a stacked device │
167 │ │ which receives packets │
168 │ │ from its underlying device │
169 │ │ based on VLAN tagging. See │
170 │ │ IEEE 802.1Q[3] for │
171 │ │ details. │
172 ├──────────┼────────────────────────────┤
173 │vti │ An IPv4 over IPSec tunnel. │
174 ├──────────┼────────────────────────────┤
175 │vti6 │ An IPv6 over IPSec tunnel. │
176 ├──────────┼────────────────────────────┤
177 │vxlan │ A virtual extensible LAN │
178 │ │ (vxlan), for connecting │
179 │ │ Cloud computing │
180 │ │ deployments. │
181 ├──────────┼────────────────────────────┤
182 │geneve │ A GEneric NEtwork │
183 │ │ Virtualization │
184 │ │ Encapsulation (GENEVE) │
185 │ │ netdev driver. │
186 ├──────────┼────────────────────────────┤
187 │l2tp │ A Layer 2 Tunneling │
188 │ │ Protocol (L2TP) is a │
189 │ │ tunneling protocol used to │
190 │ │ support virtual private │
191 │ │ networks (VPNs) or as part │
192 │ │ of the delivery of │
193 │ │ services by ISPs. It does │
194 │ │ not provide any encryption │
195 │ │ or confidentiality by │
196 │ │ itself │
197 ├──────────┼────────────────────────────┤
198 │macsec │ Media Access Control │
199 │ │ Security (MACsec) is an │
200 │ │ 802.1AE IEEE │
201 │ │ industry-standard security │
202 │ │ technology that provides │
203 │ │ secure communication for │
204 │ │ all traffic on Ethernet │
205 │ │ links. MACsec provides │
206 │ │ point-to-point security on │
207 │ │ Ethernet links between │
208 │ │ directly connected nodes │
209 │ │ and is capable of │
210 │ │ identifying and preventing │
211 │ │ most security threats. │
212 ├──────────┼────────────────────────────┤
213 │vrf │ A Virtual Routing and │
214 │ │ Forwarding (VRF[4]) │
215 │ │ interface to create │
216 │ │ separate routing and │
217 │ │ forwarding domains. │
218 ├──────────┼────────────────────────────┤
219 │vcan │ The virtual CAN driver │
220 │ │ (vcan). Similar to the │
221 │ │ network loopback devices, │
222 │ │ vcan offers a virtual │
223 │ │ local CAN interface. │
224 ├──────────┼────────────────────────────┤
225 │vxcan │ The virtual CAN tunnel │
226 │ │ driver (vxcan). Similar to │
227 │ │ the virtual ethernet │
228 │ │ driver veth, vxcan │
229 │ │ implements a local CAN │
230 │ │ traffic tunnel between two │
231 │ │ virtual CAN network │
232 │ │ devices. When creating a │
233 │ │ vxcan, two vxcan devices │
234 │ │ are created as pair. When │
235 │ │ one end receives the │
236 │ │ packet it appears on its │
237 │ │ pair and vice versa. The │
238 │ │ vxcan can be used for │
239 │ │ cross namespace │
240 │ │ communication. │
241 ├──────────┼────────────────────────────┤
242 │wireguard │ WireGuard Secure Network │
243 │ │ Tunnel. │
244 ├──────────┼────────────────────────────┤
245 │nlmon │ A Netlink monitor device. │
246 │ │ Use an nlmon device when │
247 │ │ you want to monitor system │
248 │ │ Netlink messages. │
249 ├──────────┼────────────────────────────┤
250 │fou │ Foo-over-UDP tunneling. │
251 ├──────────┼────────────────────────────┤
252 │xfrm │ A virtual tunnel interface │
253 │ │ like vti/vti6 but with │
254 │ │ several advantages. │
255 ├──────────┼────────────────────────────┤
256 │ifb │ The Intermediate │
257 │ │ Functional Block (ifb) │
258 │ │ pseudo network interface │
259 │ │ acts as a QoS concentrator │
260 │ │ for multiple different │
261 │ │ sources of traffic. │
262 ├──────────┼────────────────────────────┤
263 │bareudp │ Bare UDP tunnels provide a │
264 │ │ generic L3 encapsulation │
265 │ │ support for tunnelling │
266 │ │ different L3 protocols │
267 │ │ like MPLS, IP etc. inside │
268 │ │ of an UDP tunnel. │
269 ├──────────┼────────────────────────────┤
270 │batadv │ B.A.T.M.A.N. Advanced[5] │
271 │ │ is a routing protocol for │
272 │ │ multi-hop mobile ad-hoc │
273 │ │ networks which operates on │
274 │ │ layer 2. │
275 ├──────────┼────────────────────────────┤
276 │ipoib │ An IP over Infiniband │
277 │ │ subinterface. │
278 ├──────────┼────────────────────────────┤
279 │wlan │ A virtual wireless network │
280 │ │ (WLAN) interface. │
281 └──────────┴────────────────────────────┘
282
284 A virtual network device is only created if the [Match] section matches
285 the current environment, or if the section is empty. The following keys
286 are accepted:
287 Host=
289 Matches against the hostname or machine ID of the host. See
290 ConditionHost= in systemd.unit(5) for details. When prefixed with
291 an exclamation mark ("!"), the result is negated. If an empty
292 string is assigned, then previously assigned value is cleared.
293 Virtualization=
295 Checks whether the system is executed in a virtualized environment
296 and optionally test whether it is a specific implementation. See
297 ConditionVirtualization= in systemd.unit(5) for details. When
298 prefixed with an exclamation mark ("!"), the result is negated. If
299 an empty string is assigned, then previously assigned value is
300 cleared.
301 KernelCommandLine=
303 Checks whether a specific kernel command line option is set. See
304 ConditionKernelCommandLine= in systemd.unit(5) for details. When
305 prefixed with an exclamation mark ("!"), the result is negated. If
306 an empty string is assigned, then previously assigned value is
307 cleared.
308 KernelVersion=
310 Checks whether the kernel version (as reported by uname -r) matches
311 a certain expression. See ConditionKernelVersion= in
312 systemd.unit(5) for details. When prefixed with an exclamation mark
313 ("!"), the result is negated. If an empty string is assigned, then
314 previously assigned value is cleared.
315 Architecture=
317 Checks whether the system is running on a specific architecture.
318 See ConditionArchitecture= in systemd.unit(5) for details. When
319 prefixed with an exclamation mark ("!"), the result is negated. If
320 an empty string is assigned, then previously assigned value is
321 cleared.
322 Firmware=
324 Checks whether the system is running on a machine with the
325 specified firmware. See ConditionFirmware= in systemd.unit(5) for
326 details. When prefixed with an exclamation mark ("!"), the result
327 is negated. If an empty string is assigned, then previously
328 assigned value is cleared.
329
331 The [NetDev] section accepts the following keys:
332 Description=
334 A free-form description of the netdev.
335 Name=
337 The interface name used when creating the netdev. This setting is
338 compulsory.
339 Kind=
341 The netdev kind. This setting is compulsory. See the "Supported
342 netdev kinds" section for the valid keys.
343 MTUBytes=
345 The maximum transmission unit in bytes to set for the device. The
346 usual suffixes K, M, G are supported and are understood to the base
347 of 1024. For "tun" or "tap" devices, MTUBytes= setting is not
348 currently supported in [NetDev] section. Please specify it in
349 [Link] section of corresponding systemd.network(5) files.
350 MACAddress=
352 Specifies the MAC address to use for the device, or takes the
353 special value "none". When "none", systemd-networkd does not
354 request the MAC address for the device, and the kernel will assign
355 a random MAC address. For "tun", "tap", or "l2tp" devices, the
356 MACAddress= setting in the [NetDev] section is not supported and
357 will be ignored. Please specify it in the [Link] section of the
358 corresponding systemd.network(5) file. If this option is not set,
359 "vlan" device inherits the MAC address of the master interface. For
360 other kind of netdevs, if this option is not set, then the MAC
361 address is generated based on the interface name and the machine-
362 id(5).
363 Note, even if "none" is specified, systemd-udevd will assign the
365 persistent MAC address for the device, as 99-default.link has
366 MACAddressPolicy=persistent. So, it is also necessary to create a
367 custom .link file for the device, if the MAC address assignment is
368 not desired.
369
371 The [Bridge] section only applies for netdevs of kind "bridge", and
372 accepts the following keys:
373 HelloTimeSec=
375 HelloTimeSec specifies the number of seconds between two hello
376 packets sent out by the root bridge and the designated bridges.
377 Hello packets are used to communicate information about the
378 topology throughout the entire bridged local area network.
379 MaxAgeSec=
381 MaxAgeSec specifies the number of seconds of maximum message age.
382 If the last seen (received) hello packet is more than this number
383 of seconds old, the bridge in question will start the takeover
384 procedure in attempt to become the Root Bridge itself.
385 ForwardDelaySec=
387 ForwardDelaySec specifies the number of seconds spent in each of
388 the Listening and Learning states before the Forwarding state is
389 entered.
390 AgeingTimeSec=
392 This specifies the number of seconds a MAC Address will be kept in
393 the forwarding database after having a packet received from this
394 MAC Address.
395 Priority=
397 The priority of the bridge. An integer between 0 and 65535. A lower
398 value means higher priority. The bridge having the lowest priority
399 will be elected as root bridge.
400 GroupForwardMask=
402 A 16-bit bitmask represented as an integer which allows forwarding
403 of link local frames with 802.1D reserved addresses
404 (01:80:C2:00:00:0X). A logical AND is performed between the
405 specified bitmask and the exponentiation of 2^X, the lower nibble
406 of the last octet of the MAC address. For example, a value of 8
407 would allow forwarding of frames addressed to 01:80:C2:00:00:03
408 (802.1X PAE).
409 DefaultPVID=
411 This specifies the default port VLAN ID of a newly attached bridge
412 port. Set this to an integer in the range 1...4094 or "none" to
413 disable the PVID.
414 MulticastQuerier=
416 Takes a boolean. This setting controls the IFLA_BR_MCAST_QUERIER
417 option in the kernel. If enabled, the kernel will send general ICMP
418 queries from a zero source address. This feature should allow
419 faster convergence on startup, but it causes some multicast-aware
420 switches to misbehave and disrupt forwarding of multicast packets.
421 When unset, the kernel's default will be used.
422 MulticastSnooping=
424 Takes a boolean. This setting controls the IFLA_BR_MCAST_SNOOPING
425 option in the kernel. If enabled, IGMP snooping monitors the
426 Internet Group Management Protocol (IGMP) traffic between hosts and
427 multicast routers. When unset, the kernel's default will be used.
428 VLANFiltering=
430 Takes a boolean. This setting controls the IFLA_BR_VLAN_FILTERING
431 option in the kernel. If enabled, the bridge will be started in
432 VLAN-filtering mode. When unset, the kernel's default will be used.
433 VLANProtocol=
435 Allows setting the protocol used for VLAN filtering. Takes 802.1q
436 or, 802.1ad, and defaults to unset and kernel's default is used.
437 STP=
439 Takes a boolean. This enables the bridge's Spanning Tree Protocol
440 (STP). When unset, the kernel's default will be used.
441 MulticastIGMPVersion=
443 Allows changing bridge's multicast Internet Group Management
444 Protocol (IGMP) version. Takes an integer 2 or 3. When unset, the
445 kernel's default will be used.
446
448 The [VLAN] section only applies for netdevs of kind "vlan", and accepts
449 the following key:
450 Id=
452 The VLAN ID to use. An integer in the range 0...4094. This setting
453 is compulsory.
454 Protocol=
456 Allows setting the protocol used for the VLAN interface. Takes
457 "802.1q" or, "802.1ad", and defaults to unset and kernel's default
458 is used.
459 GVRP=
461 Takes a boolean. The Generic VLAN Registration Protocol (GVRP) is a
462 protocol that allows automatic learning of VLANs on a network. When
463 unset, the kernel's default will be used.
464 MVRP=
466 Takes a boolean. Multiple VLAN Registration Protocol (MVRP)
467 formerly known as GARP VLAN Registration Protocol (GVRP) is a
468 standards-based Layer 2 network protocol, for automatic
469 configuration of VLAN information on switches. It was defined in
470 the 802.1ak amendment to 802.1Q-2005. When unset, the kernel's
471 default will be used.
472 LooseBinding=
474 Takes a boolean. The VLAN loose binding mode, in which only the
475 operational state is passed from the parent to the associated
476 VLANs, but the VLAN device state is not changed. When unset, the
477 kernel's default will be used.
478 ReorderHeader=
480 Takes a boolean. When enabled, the VLAN reorder header is used and
481 VLAN interfaces behave like physical interfaces. When unset, the
482 kernel's default will be used.
483 EgressQOSMaps=, IngressQOSMaps=
485 Defines a mapping of Linux internal packet priority (SO_PRIORITY)
486 to VLAN header PCP field for outgoing and incoming frames,
487 respectively. Takes a whitespace-separated list of integer pairs,
488 where each integer must be in the range 1...4294967294, in the
489 format "from"-"to", e.g., "21-7 45-5". Note that "from" must be
490 greater than or equal to "to". When unset, the kernel's default
491 will be used.
492
494 The [MACVLAN] section only applies for netdevs of kind "macvlan", and
495 accepts the following key:
496 Mode=
498 The MACVLAN mode to use. The supported options are "private",
499 "vepa", "bridge", "passthru", and "source".
500 SourceMACAddress=
502 A whitespace-separated list of remote hardware addresses allowed on
503 the MACVLAN. This option only has an effect in source mode. Use
504 full colon-, hyphen- or dot-delimited hexadecimal. This option may
505 appear more than once, in which case the lists are merged. If the
506 empty string is assigned to this option, the list of hardware
507 addresses defined prior to this is reset. Defaults to unset.
508 BroadcastMulticastQueueLength=
510 Specifies the length of the receive queue for broadcast/multicast
511 packets. An unsigned integer in the range 0...4294967294. Defaults
512 to unset.
513
515 The [MACVTAP] section applies for netdevs of kind "macvtap" and accepts
516 the same keys as [MACVLAN].
517
519 The [IPVLAN] section only applies for netdevs of kind "ipvlan", and
520 accepts the following key:
521 Mode=
523 The IPVLAN mode to use. The supported options are "L2","L3" and
524 "L3S".
525 Flags=
527 The IPVLAN flags to use. The supported options are
528 "bridge","private" and "vepa".
529
531 The [IPVTAP] section only applies for netdevs of kind "ipvtap" and
532 accepts the same keys as [IPVLAN].
533
535 The [VXLAN] section only applies for netdevs of kind "vxlan", and
536 accepts the following keys:
537 VNI=
539 The VXLAN Network Identifier (or VXLAN Segment ID). Takes a number
540 in the range 1...16777215.
541 Remote=
543 Configures destination IP address.
544 Local=
546 Configures local IP address. It must be an address on the
547 underlying interface of the VXLAN interface, or one of the special
548 values "ipv4_link_local", "ipv6_link_local", "dhcp4", "dhcp6", and
549 "slaac". If one of the special values is specified, an address
550 which matches the corresponding type on the underlying interface
551 will be used. Defaults to unset.
552 Group=
554 Configures VXLAN multicast group IP address. All members of a VXLAN
555 must use the same multicast group address.
556 TOS=
558 The Type Of Service byte value for a vxlan interface.
559 TTL=
561 A fixed Time To Live N on Virtual eXtensible Local Area Network
562 packets. Takes "inherit" or a number in the range 0...255. 0 is a
563 special value meaning inherit the inner protocol's TTL value.
564 "inherit" means that it will inherit the outer protocol's TTL
565 value.
566 MacLearning=
568 Takes a boolean. When true, enables dynamic MAC learning to
569 discover remote MAC addresses.
570 FDBAgeingSec=
572 The lifetime of Forwarding Database entry learnt by the kernel, in
573 seconds.
574 MaximumFDBEntries=
576 Configures maximum number of FDB entries.
577 ReduceARPProxy=
579 Takes a boolean. When true, bridge-connected VXLAN tunnel endpoint
580 answers ARP requests from the local bridge on behalf of remote
581 Distributed Overlay Virtual Ethernet (DOVE)[6] clients. Defaults to
582 false.
583 L2MissNotification=
585 Takes a boolean. When true, enables netlink LLADDR miss
586 notifications.
587 L3MissNotification=
589 Takes a boolean. When true, enables netlink IP address miss
590 notifications.
591 RouteShortCircuit=
593 Takes a boolean. When true, route short circuiting is turned on.
594 UDPChecksum=
596 Takes a boolean. When true, transmitting UDP checksums when doing
597 VXLAN/IPv4 is turned on.
598 UDP6ZeroChecksumTx=
600 Takes a boolean. When true, sending zero checksums in VXLAN/IPv6 is
601 turned on.
602 UDP6ZeroChecksumRx=
604 Takes a boolean. When true, receiving zero checksums in VXLAN/IPv6
605 is turned on.
606 RemoteChecksumTx=
608 Takes a boolean. When true, remote transmit checksum offload of
609 VXLAN is turned on.
610 RemoteChecksumRx=
612 Takes a boolean. When true, remote receive checksum offload in
613 VXLAN is turned on.
614 GroupPolicyExtension=
616 Takes a boolean. When true, it enables Group Policy VXLAN extension
617 security label mechanism across network peers based on VXLAN. For
618 details about the Group Policy VXLAN, see the VXLAN Group Policy[7]
619 document. Defaults to false.
620 GenericProtocolExtension=
622 Takes a boolean. When true, Generic Protocol Extension extends the
623 existing VXLAN protocol to provide protocol typing, OAM, and
624 versioning capabilities. For details about the VXLAN GPE Header,
625 see the Generic Protocol Extension for VXLAN[8] document. If
626 destination port is not specified and Generic Protocol Extension is
627 set then default port of 4790 is used. Defaults to false.
628 DestinationPort=
630 Configures the default destination UDP port. If the destination
631 port is not specified then Linux kernel default will be used. Set
632 to 4789 to get the IANA assigned value.
633 PortRange=
635 Configures the source port range for the VXLAN. The kernel assigns
636 the source UDP port based on the flow to help the receiver to do
637 load balancing. When this option is not set, the normal range of
638 local UDP ports is used.
639 FlowLabel=
641 Specifies the flow label to use in outgoing packets. The valid
642 range is 0-1048575.
643 IPDoNotFragment=
645 Allows setting the IPv4 Do not Fragment (DF) bit in outgoing
646 packets, or to inherit its value from the IPv4 inner header. Takes
647 a boolean value, or "inherit". Set to "inherit" if the encapsulated
648 protocol is IPv6. When unset, the kernel's default will be used.
649 Independent=
651 Takes a boolean. When true, the vxlan interface is created without
652 any underlying network interface. Defaults to false, which means
653 that a .network file that requests this VXLAN interface using
654 VXLAN= is required for the VXLAN to be created.
655
657 The [GENEVE] section only applies for netdevs of kind "geneve", and
658 accepts the following keys:
659 Id=
661 Specifies the Virtual Network Identifier (VNI) to use, a number
662 between 0 and 16777215. This field is mandatory.
663 Remote=
665 Specifies the unicast destination IP address to use in outgoing
666 packets.
667 TOS=
669 Specifies the TOS value to use in outgoing packets. Takes a number
670 between 1 and 255.
671 TTL=
673 Accepts the same values as in the [VXLAN] section, except that when
674 unset or set to 0, the kernel's default will be used, meaning that
675 packet TTL will be set from /proc/sys/net/ipv4/ip_default_ttl.
676 UDPChecksum=
678 Takes a boolean. When true, specifies that UDP checksum is
679 calculated for transmitted packets over IPv4.
680 UDP6ZeroChecksumTx=
682 Takes a boolean. When true, skip UDP checksum calculation for
683 transmitted packets over IPv6.
684 UDP6ZeroChecksumRx=
686 Takes a boolean. When true, allows incoming UDP packets over IPv6
687 with zero checksum field.
688 DestinationPort=
690 Specifies destination port. Defaults to 6081. If not set or
691 assigned the empty string, the default port of 6081 is used.
692 FlowLabel=
694 Specifies the flow label to use in outgoing packets.
695 IPDoNotFragment=
697 Accepts the same key as in [VXLAN] section.
698
700 The [BareUDP] section only applies for netdevs of kind "bareudp", and
701 accepts the following keys:
702 DestinationPort=
704 Specifies the destination UDP port (in range 1...65535). This is
705 mandatory.
706 EtherType=
708 Specifies the L3 protocol. Takes one of "ipv4", "ipv6", "mpls-uc"
709 or "mpls-mc". This is mandatory.
710
712 The [L2TP] section only applies for netdevs of kind "l2tp", and accepts
713 the following keys:
714 TunnelId=
716 Specifies the tunnel identifier. Takes an number in the range
717 1...4294967295. The value used must match the "PeerTunnelId=" value
718 being used at the peer. This setting is compulsory.
719 PeerTunnelId=
721 Specifies the peer tunnel id. Takes a number in the range
722 1...4294967295. The value used must match the "TunnelId=" value
723 being used at the peer. This setting is compulsory.
724 Remote=
726 Specifies the IP address of the remote peer. This setting is
727 compulsory.
728 Local=
730 Specifies the IP address of a local interface. Takes an IP address,
731 or the special values "auto", "static", or "dynamic". Optionally a
732 name of a local interface can be specified after "@", e.g.
733 "192.168.0.1@eth0" or "auto@eth0". When an address is specified,
734 then a local or specified interface must have the address, and the
735 remote address must be accessible through the local address. If
736 "auto", then one of the addresses on a local or specified interface
737 which is accessible to the remote address will be used. Similarly,
738 if "static" or "dynamic" is set, then one of the static or dynamic
739 addresses will be used. Defaults to "auto".
740 EncapsulationType=
742 Specifies the encapsulation type of the tunnel. Takes one of "udp"
743 or "ip".
744 UDPSourcePort=
746 Specifies the UDP source port to be used for the tunnel. When UDP
747 encapsulation is selected it's mandatory. Ignored when IP
748 encapsulation is selected.
749 UDPDestinationPort=
751 Specifies destination port. When UDP encapsulation is selected it's
752 mandatory. Ignored when IP encapsulation is selected.
753 UDPChecksum=
755 Takes a boolean. When true, specifies that UDP checksum is
756 calculated for transmitted packets over IPv4.
757 UDP6ZeroChecksumTx=
759 Takes a boolean. When true, skip UDP checksum calculation for
760 transmitted packets over IPv6.
761 UDP6ZeroChecksumRx=
763 Takes a boolean. When true, allows incoming UDP packets over IPv6
764 with zero checksum field.
765
767 The [L2TPSession] section only applies for netdevs of kind "l2tp", and
768 accepts the following keys:
769 Name=
771 Specifies the name of the session. This setting is compulsory.
772 SessionId=
774 Specifies the session identifier. Takes an number in the range
775 1...4294967295. The value used must match the "SessionId=" value
776 being used at the peer. This setting is compulsory.
777 PeerSessionId=
779 Specifies the peer session identifier. Takes an number in the range
780 1...4294967295. The value used must match the "PeerSessionId="
781 value being used at the peer. This setting is compulsory.
782 Layer2SpecificHeader=
784 Specifies layer2specific header type of the session. One of "none"
785 or "default". Defaults to "default".
786
788 The [MACsec] section only applies for network devices of kind "macsec",
789 and accepts the following keys:
790 Port=
792 Specifies the port to be used for the MACsec transmit channel. The
793 port is used to make secure channel identifier (SCI). Takes a value
794 between 1 and 65535. Defaults to unset.
795 Encrypt=
797 Takes a boolean. When true, enable encryption. Defaults to unset.
798
800 The [MACsecReceiveChannel] section only applies for network devices of
801 kind "macsec", and accepts the following keys:
802 Port=
804 Specifies the port to be used for the MACsec receive channel. The
805 port is used to make secure channel identifier (SCI). Takes a value
806 between 1 and 65535. This option is compulsory, and is not set by
807 default.
808 MACAddress=
810 Specifies the MAC address to be used for the MACsec receive
811 channel. The MAC address used to make secure channel identifier
812 (SCI). This setting is compulsory, and is not set by default.
813
815 The [MACsecTransmitAssociation] section only applies for network
816 devices of kind "macsec", and accepts the following keys:
817 PacketNumber=
819 Specifies the packet number to be used for replay protection and
820 the construction of the initialization vector (along with the
821 secure channel identifier [SCI]). Takes a value between
822 1-4,294,967,295. Defaults to unset.
823 KeyId=
825 Specifies the identification for the key. Takes a number between
826 0-255. This option is compulsory, and is not set by default.
827 Key=
829 Specifies the encryption key used in the transmission channel. The
830 same key must be configured on the peer’s matching receive channel.
831 This setting is compulsory, and is not set by default. Takes a
832 128-bit key encoded in a hexadecimal string, for example
833 "dffafc8d7b9a43d5b9a3dfbbf6a30c16".
834 KeyFile=
836 Takes an absolute path to a file which contains a 128-bit key
837 encoded in a hexadecimal string, which will be used in the
838 transmission channel. When this option is specified, Key= is
839 ignored. Note that the file must be readable by the user
840 "systemd-network", so it should be, e.g., owned by
841 "root:systemd-network" with a "0640" file mode. If the path refers
842 to an AF_UNIX stream socket in the file system a connection is made
843 to it and the key read from it.
844 Activate=
846 Takes a boolean. If enabled, then the security association is
847 activated. Defaults to unset.
848 UseForEncoding=
850 Takes a boolean. If enabled, then the security association is used
851 for encoding. Only one [MACsecTransmitAssociation] section can
852 enable this option. When enabled, Activate=yes is implied. Defaults
853 to unset.
854
856 The [MACsecReceiveAssociation] section only applies for network devices
857 of kind "macsec", and accepts the following keys:
858 Port=
860 Accepts the same key as in [MACsecReceiveChannel] section.
861 MACAddress=
863 Accepts the same key as in [MACsecReceiveChannel] section.
864 PacketNumber=
866 Accepts the same key as in [MACsecTransmitAssociation] section.
867 KeyId=
869 Accepts the same key as in [MACsecTransmitAssociation] section.
870 Key=
872 Accepts the same key as in [MACsecTransmitAssociation] section.
873 KeyFile=
875 Accepts the same key as in [MACsecTransmitAssociation] section.
876 Activate=
878 Accepts the same key as in [MACsecTransmitAssociation] section.
879
881 The [Tunnel] section only applies for netdevs of kind "ipip", "sit",
882 "gre", "gretap", "ip6gre", "ip6gretap", "vti", "vti6", "ip6tnl", and
883 "erspan" and accepts the following keys:
884 External=
886 Takes a boolean value. When true, then the tunnel is externally
887 controlled, which is also known as collect metadata mode, and most
888 settings below like Local= or Remote= are ignored. This implies
889 Independent=. Defaults to false.
890 Local=
892 A static local address for tunneled packets. It must be an address
893 on another interface of this host, or one of the special values
894 "any", "ipv4_link_local", "ipv6_link_local", "dhcp4", "dhcp6", and
895 "slaac". If one of the special values except for "any" is
896 specified, an address which matches the corresponding type on the
897 underlying interface will be used. Defaults to "any".
898 Remote=
900 The remote endpoint of the tunnel. Takes an IP address or the
901 special value "any".
902 TOS=
904 The Type Of Service byte value for a tunnel interface. For details
905 about the TOS, see the Type of Service in the Internet Protocol
906 Suite[9] document.
907 TTL=
909 A fixed Time To Live N on tunneled packets. N is a number in the
910 range 1...255. 0 is a special value meaning that packets inherit
911 the TTL value. The default value for IPv4 tunnels is 0 (inherit).
912 The default value for IPv6 tunnels is 64.
913 DiscoverPathMTU=
915 Takes a boolean. When true, enables Path MTU Discovery on the
916 tunnel.
917 IPv6FlowLabel=
919 Configures the 20-bit flow label (see RFC 6437[10]) field in the
920 IPv6 header (see RFC 2460[11]), which is used by a node to label
921 packets of a flow. It is only used for IPv6 tunnels. A flow label
922 of zero is used to indicate packets that have not been labeled. It
923 can be configured to a value in the range 0...0xFFFFF, or be set to
924 "inherit", in which case the original flowlabel is used.
925 CopyDSCP=
927 Takes a boolean. When true, the Differentiated Service Code Point
928 (DSCP) field will be copied to the inner header from outer header
929 during the decapsulation of an IPv6 tunnel packet. DSCP is a field
930 in an IP packet that enables different levels of service to be
931 assigned to network traffic. Defaults to "no".
932 EncapsulationLimit=
934 The Tunnel Encapsulation Limit option specifies how many additional
935 levels of encapsulation are permitted to be prepended to the
936 packet. For example, a Tunnel Encapsulation Limit option containing
937 a limit value of zero means that a packet carrying that option may
938 not enter another tunnel before exiting the current tunnel. (see
939 RFC 2473[12]). The valid range is 0...255 and "none". Defaults to
940 4.
941 Key=
943 The Key= parameter specifies the same key to use in both directions
944 (InputKey= and OutputKey=). The Key= is either a number or an IPv4
945 address-like dotted quad. It is used as mark-configured SAD/SPD
946 entry as part of the lookup key (both in data and control path) in
947 IP XFRM (framework used to implement IPsec protocol). See ip-xfrm —
948 transform configuration[13] for details. It is only used for
949 VTI/VTI6, GRE, GRETAP, and ERSPAN tunnels.
950 InputKey=
952 The InputKey= parameter specifies the key to use for input. The
953 format is same as Key=. It is only used for VTI/VTI6, GRE, GRETAP,
954 and ERSPAN tunnels.
955 OutputKey=
957 The OutputKey= parameter specifies the key to use for output. The
958 format is same as Key=. It is only used for VTI/VTI6, GRE, GRETAP,
959 and ERSPAN tunnels.
960 Mode=
962 An "ip6tnl" tunnel can be in one of three modes "ip6ip6" for IPv6
963 over IPv6, "ipip6" for IPv4 over IPv6 or "any" for either.
964 Independent=
966 Takes a boolean. When false (the default), the tunnel is always
967 created over some network device, and a .network file that requests
968 this tunnel using Tunnel= is required for the tunnel to be created.
969 When true, the tunnel is created independently of any network as
970 "tunnel@NONE".
971 AssignToLoopback=
973 Takes a boolean. If set to "yes", the loopback interface "lo" is
974 used as the underlying device of the tunnel interface. Defaults to
975 "no".
976 AllowLocalRemote=
978 Takes a boolean. When true allows tunnel traffic on ip6tnl devices
979 where the remote endpoint is a local host address. When unset, the
980 kernel's default will be used.
981 FooOverUDP=
983 Takes a boolean. Specifies whether FooOverUDP= tunnel is to be
984 configured. Defaults to false. This takes effects only for IPIP,
985 SIT, GRE, and GRETAP tunnels. For more detail information see Foo
986 over UDP[14]
987 FOUDestinationPort=
989 This setting specifies the UDP destination port for encapsulation.
990 This field is mandatory when FooOverUDP=yes, and is not set by
991 default.
992 FOUSourcePort=
994 This setting specifies the UDP source port for encapsulation.
995 Defaults to 0 — that is, the source port for packets is left to the
996 network stack to decide.
997 Encapsulation=
999 Accepts the same key as in the [FooOverUDP] section.
1000 IPv6RapidDeploymentPrefix=
1002 Reconfigure the tunnel for IPv6 Rapid Deployment[15], also known as
1003 6rd. The value is an ISP-specific IPv6 prefix with a non-zero
1004 length. Only applicable to SIT tunnels.
1005 ISATAP=
1007 Takes a boolean. If set, configures the tunnel as Intra-Site
1008 Automatic Tunnel Addressing Protocol (ISATAP) tunnel. Only
1009 applicable to SIT tunnels. When unset, the kernel's default will be
1010 used.
1011 SerializeTunneledPackets=
1013 Takes a boolean. If set to yes, then packets are serialized. Only
1014 applies for GRE, GRETAP, and ERSPAN tunnels. When unset, the
1015 kernel's default will be used.
1016 ERSPANIndex=
1018 Specifies the ERSPAN index field for the interface, an integer in
1019 the range 1...1048575 associated with the ERSPAN traffic's source
1020 port and direction. This field is mandatory.
1021
1023 The [FooOverUDP] section only applies for netdevs of kind "fou" and
1024 accepts the following keys:
1025 Encapsulation=
1027 Specifies the encapsulation mechanism used to store networking
1028 packets of various protocols inside the UDP packets. Supports the
1029 following values: "FooOverUDP" provides the simplest no-frills
1030 model of UDP encapsulation, it simply encapsulates packets directly
1031 in the UDP payload. "GenericUDPEncapsulation" is a generic and
1032 extensible encapsulation, it allows encapsulation of packets for
1033 any IP protocol and optional data as part of the encapsulation. For
1034 more detailed information see Generic UDP Encapsulation[16].
1035 Defaults to "FooOverUDP".
1036 Port=
1038 Specifies the port number where the encapsulated packets will
1039 arrive. Those packets will be removed and manually fed back into
1040 the network stack with the encapsulation removed to be sent to the
1041 real destination. This option is mandatory.
1042 PeerPort=
1044 Specifies the peer port number. Defaults to unset. Note that when
1045 peer port is set "Peer=" address is mandatory.
1046 Protocol=
1048 The Protocol= specifies the protocol number of the packets arriving
1049 at the UDP port. When Encapsulation=FooOverUDP, this field is
1050 mandatory and is not set by default. Takes an IP protocol name such
1051 as "gre" or "ipip", or an integer within the range 1...255. When
1052 Encapsulation=GenericUDPEncapsulation, this must not be specified.
1053 Peer=
1055 Configures peer IP address. Note that when peer address is set
1056 "PeerPort=" is mandatory.
1057 Local=
1059 Configures local IP address.
1060
1062 The [Peer] section only applies for netdevs of kind "veth" and accepts
1063 the following keys:
1064 Name=
1066 The interface name used when creating the netdev. This setting is
1067 compulsory.
1068 MACAddress=
1070 The peer MACAddress, if not set, it is generated in the same way as
1071 the MAC address of the main interface.
1072
1074 The [VXCAN] section only applies for netdevs of kind "vxcan" and
1075 accepts the following key:
1076 Peer=
1078 The peer interface name used when creating the netdev. This setting
1079 is compulsory.
1080
1082 The [Tun] section only applies for netdevs of kind "tun", and accepts
1083 the following keys:
1084 MultiQueue=
1086 Takes a boolean. Configures whether to use multiple file
1087 descriptors (queues) to parallelize packets sending and receiving.
1088 Defaults to "no".
1089 PacketInfo=
1091 Takes a boolean. Configures whether packets should be prepended
1092 with four extra bytes (two flag bytes and two protocol bytes). If
1093 disabled, it indicates that the packets will be pure IP packets.
1094 Defaults to "no".
1095 VNetHeader=
1097 Takes a boolean. Configures IFF_VNET_HDR flag for a tun or tap
1098 device. It allows sending and receiving larger Generic Segmentation
1099 Offload (GSO) packets. This may increase throughput significantly.
1100 Defaults to "no".
1101 User=
1103 User to grant access to the /dev/net/tun device.
1104 Group=
1106 Group to grant access to the /dev/net/tun device.
1107
1109 The [Tap] section only applies for netdevs of kind "tap", and accepts
1110 the same keys as the [Tun] section.
1111
1113 The [WireGuard] section accepts the following keys:
1114 PrivateKey=
1116 The Base64 encoded private key for the interface. It can be
1117 generated using the wg genkey command (see wg(8)). This option or
1118 PrivateKeyFile= is mandatory to use WireGuard. Note that because
1119 this information is secret, you may want to set the permissions of
1120 the .netdev file to be owned by "root:systemd-network" with a
1121 "0640" file mode.
1122 PrivateKeyFile=
1124 Takes an absolute path to a file which contains the Base64 encoded
1125 private key for the interface. When this option is specified, then
1126 PrivateKey= is ignored. Note that the file must be readable by the
1127 user "systemd-network", so it should be, e.g., owned by
1128 "root:systemd-network" with a "0640" file mode. If the path refers
1129 to an AF_UNIX stream socket in the file system a connection is made
1130 to it and the key read from it.
1131 ListenPort=
1133 Sets UDP port for listening. Takes either value between 1 and 65535
1134 or "auto". If "auto" is specified, the port is automatically
1135 generated based on interface name. Defaults to "auto".
1136 FirewallMark=
1138 Sets a firewall mark on outgoing WireGuard packets from this
1139 interface. Takes a number between 1 and 4294967295.
1140 RouteTable=
1142 The table identifier for the routes to the addresses specified in
1143 the AllowedIPs=. Takes a negative boolean value, one of the
1144 predefined names "default", "main", and "local", names defined in
1145 RouteTable= in networkd.conf(5), or a number in the range
1146 1...4294967295. When "off" the routes to the addresses specified in
1147 the AllowedIPs= setting will not be configured. Defaults to false.
1148 This setting will be ignored when the same setting is specified in
1149 the [WireGuardPeer] section.
1150 RouteMetric=
1152 The priority of the routes to the addresses specified in the
1153 AllowedIPs=. Takes an integer in the range 0...4294967295. Defaults
1154 to 0 for IPv4 addresses, and 1024 for IPv6 addresses. This setting
1155 will be ignored when the same setting is specified in the
1156 [WireGuardPeer] section.
1157
1159 The [WireGuardPeer] section accepts the following keys:
1160 PublicKey=
1162 Sets a Base64 encoded public key calculated by wg pubkey (see
1163 wg(8)) from a private key, and usually transmitted out of band to
1164 the author of the configuration file. This option is mandatory for
1165 this section.
1166 PresharedKey=
1168 Optional preshared key for the interface. It can be generated by
1169 the wg genpsk command. This option adds an additional layer of
1170 symmetric-key cryptography to be mixed into the already existing
1171 public-key cryptography, for post-quantum resistance. Note that
1172 because this information is secret, you may want to set the
1173 permissions of the .netdev file to be owned by
1174 "root:systemd-network" with a "0640" file mode.
1175 PresharedKeyFile=
1177 Takes an absolute path to a file which contains the Base64 encoded
1178 preshared key for the peer. When this option is specified, then
1179 PresharedKey= is ignored. Note that the file must be readable by
1180 the user "systemd-network", so it should be, e.g., owned by
1181 "root:systemd-network" with a "0640" file mode. If the path refers
1182 to an AF_UNIX stream socket in the file system a connection is made
1183 to it and the key read from it.
1184 AllowedIPs=
1186 Sets a comma-separated list of IP (v4 or v6) addresses with CIDR
1187 masks from which this peer is allowed to send incoming traffic and
1188 to which outgoing traffic for this peer is directed.
1189 The catch-all 0.0.0.0/0 may be specified for matching all IPv4
1191 addresses, and ::/0 may be specified for matching all IPv6
1192 addresses.
1193 Note that this only affects routing inside the network interface
1195 itself, i.e. the packets that pass through the tunnel itself. To
1196 cause packets to be sent via the tunnel in the first place, an
1197 appropriate route needs to be added as well — either in the
1198 "[Routes]" section on the ".network" matching the wireguard
1199 interface, or externally to systemd-networkd.
1200 Endpoint=
1202 Sets an endpoint IP address or hostname, followed by a colon, and
1203 then a port number. This endpoint will be updated automatically
1204 once to the most recent source IP address and port of correctly
1205 authenticated packets from the peer at configuration time.
1206 PersistentKeepalive=
1208 Sets a seconds interval, between 1 and 65535 inclusive, of how
1209 often to send an authenticated empty packet to the peer for the
1210 purpose of keeping a stateful firewall or NAT mapping valid
1211 persistently. For example, if the interface very rarely sends
1212 traffic, but it might at anytime receive traffic from a peer, and
1213 it is behind NAT, the interface might benefit from having a
1214 persistent keepalive interval of 25 seconds. If set to 0 or "off",
1215 this option is disabled. By default or when unspecified, this
1216 option is off. Most users will not need this.
1217 RouteTable=
1219 The table identifier for the routes to the addresses specified in
1220 the AllowedIPs=. Takes a negative boolean value, one of the
1221 predefined names "default", "main", and "local", names defined in
1222 RouteTable= in networkd.conf(5), or a number in the range
1223 1...4294967295. Defaults to unset, and the value specified in the
1224 same setting in the [WireGuard] section will be used.
1225 RouteMetric=
1227 The priority of the routes to the addresses specified in the
1228 AllowedIPs=. Takes an integer in the range 0...4294967295. Defaults
1229 to unset, and the value specified in the same setting in the
1230 [WireGuard] section will be used.
1231
1233 The [Bond] section accepts the following key:
1234 Mode=
1236 Specifies one of the bonding policies. The default is "balance-rr"
1237 (round robin). Possible values are "balance-rr", "active-backup",
1238 "balance-xor", "broadcast", "802.3ad", "balance-tlb", and
1239 "balance-alb".
1240 TransmitHashPolicy=
1242 Selects the transmit hash policy to use for slave selection in
1243 balance-xor, 802.3ad, and tlb modes. Possible values are "layer2",
1244 "layer3+4", "layer2+3", "encap2+3", and "encap3+4".
1245 LACPTransmitRate=
1247 Specifies the rate with which link partner transmits Link
1248 Aggregation Control Protocol Data Unit packets in 802.3ad mode.
1249 Possible values are "slow", which requests partner to transmit
1250 LACPDUs every 30 seconds, and "fast", which requests partner to
1251 transmit LACPDUs every second. The default value is "slow".
1252 MIIMonitorSec=
1254 Specifies the frequency that Media Independent Interface link
1255 monitoring will occur. A value of zero disables MII link
1256 monitoring. This value is rounded down to the nearest millisecond.
1257 The default value is 0.
1258 UpDelaySec=
1260 Specifies the delay before a link is enabled after a link up status
1261 has been detected. This value is rounded down to a multiple of
1262 MIIMonitorSec=. The default value is 0.
1263 DownDelaySec=
1265 Specifies the delay before a link is disabled after a link down
1266 status has been detected. This value is rounded down to a multiple
1267 of MIIMonitorSec=. The default value is 0.
1268 LearnPacketIntervalSec=
1270 Specifies the number of seconds between instances where the bonding
1271 driver sends learning packets to each slave peer switch. The valid
1272 range is 1...0x7fffffff; the default value is 1. This option has an
1273 effect only for the balance-tlb and balance-alb modes.
1274 AdSelect=
1276 Specifies the 802.3ad aggregation selection logic to use. Possible
1277 values are "stable", "bandwidth" and "count".
1278 AdActorSystemPriority=
1280 Specifies the 802.3ad actor system priority. Takes a number in the
1281 range 1...65535.
1282 AdUserPortKey=
1284 Specifies the 802.3ad user defined portion of the port key. Takes a
1285 number in the range 0...1023.
1286 AdActorSystem=
1288 Specifies the 802.3ad system MAC address. This cannot be a null or
1289 multicast address.
1290 FailOverMACPolicy=
1292 Specifies whether the active-backup mode should set all slaves to
1293 the same MAC address at the time of enslavement or, when enabled,
1294 to perform special handling of the bond's MAC address in accordance
1295 with the selected policy. The default policy is none. Possible
1296 values are "none", "active" and "follow".
1297 ARPValidate=
1299 Specifies whether or not ARP probes and replies should be validated
1300 in any mode that supports ARP monitoring, or whether non-ARP
1301 traffic should be filtered (disregarded) for link monitoring
1302 purposes. Possible values are "none", "active", "backup" and "all".
1303 ARPIntervalSec=
1305 Specifies the ARP link monitoring frequency. A value of 0 disables
1306 ARP monitoring. The default value is 0, and the default unit
1307 seconds.
1308 ARPIPTargets=
1310 Specifies the IP addresses to use as ARP monitoring peers when
1311 ARPIntervalSec= is greater than 0. These are the targets of the ARP
1312 request sent to determine the health of the link to the targets.
1313 Specify these values in IPv4 dotted decimal format. At least one IP
1314 address must be given for ARP monitoring to function. The maximum
1315 number of targets that can be specified is 16. The default value is
1316 no IP addresses.
1317 ARPAllTargets=
1319 Specifies the quantity of ARPIPTargets= that must be reachable in
1320 order for the ARP monitor to consider a slave as being up. This
1321 option affects only active-backup mode for slaves with ARPValidate
1322 enabled. Possible values are "any" and "all".
1323 PrimaryReselectPolicy=
1325 Specifies the reselection policy for the primary slave. This
1326 affects how the primary slave is chosen to become the active slave
1327 when failure of the active slave or recovery of the primary slave
1328 occurs. This option is designed to prevent flip-flopping between
1329 the primary slave and other slaves. Possible values are "always",
1330 "better" and "failure".
1331 ResendIGMP=
1333 Specifies the number of IGMP membership reports to be issued after
1334 a failover event. One membership report is issued immediately after
1335 the failover, subsequent packets are sent in each 200ms interval.
1336 The valid range is 0...255. Defaults to 1. A value of 0 prevents
1337 the IGMP membership report from being issued in response to the
1338 failover event.
1339 PacketsPerSlave=
1341 Specify the number of packets to transmit through a slave before
1342 moving to the next one. When set to 0, then a slave is chosen at
1343 random. The valid range is 0...65535. Defaults to 1. This option
1344 only has effect when in balance-rr mode.
1345 GratuitousARP=
1347 Specify the number of peer notifications (gratuitous ARPs and
1348 unsolicited IPv6 Neighbor Advertisements) to be issued after a
1349 failover event. As soon as the link is up on the new slave, a peer
1350 notification is sent on the bonding device and each VLAN
1351 sub-device. This is repeated at each link monitor interval
1352 (ARPIntervalSec or MIIMonitorSec, whichever is active) if the
1353 number is greater than 1. The valid range is 0...255. The default
1354 value is 1. These options affect only the active-backup mode.
1355 AllSlavesActive=
1357 Takes a boolean. Specifies that duplicate frames (received on
1358 inactive ports) should be dropped when false, or delivered when
1359 true. Normally, bonding will drop duplicate frames (received on
1360 inactive ports), which is desirable for most users. But there are
1361 some times it is nice to allow duplicate frames to be delivered.
1362 The default value is false (drop duplicate frames received on
1363 inactive ports).
1364 DynamicTransmitLoadBalancing=
1366 Takes a boolean. Specifies if dynamic shuffling of flows is
1367 enabled. Applies only for balance-tlb mode. Defaults to unset.
1368 MinLinks=
1370 Specifies the minimum number of links that must be active before
1371 asserting carrier. The default value is 0.
1372 For more detail information see Linux Ethernet Bonding Driver HOWTO[1]
1374
1376 The [Xfrm] section accepts the following keys:
1377 InterfaceId=
1379 Sets the ID/key of the xfrm interface which needs to be associated
1380 with a SA/policy. Can be decimal or hexadecimal, valid range is
1381 1-0xffffffff. This is mandatory.
1382 Independent=
1384 Takes a boolean. If false (the default), the xfrm interface must
1385 have an underlying device which can be used for hardware
1386 offloading.
1387 For more detail information see Virtual XFRM Interfaces[17].
1389
1391 The [VRF] section only applies for netdevs of kind "vrf" and accepts
1392 the following key:
1393 Table=
1395 The numeric routing table identifier. This setting is compulsory.
1396
1398 The [BatmanAdvanced] section only applies for netdevs of kind "batadv"
1399 and accepts the following keys:
1400 GatewayMode=
1402 Takes one of "off", "server", or "client". A batman-adv node can
1403 either run in server mode (sharing its internet connection with the
1404 mesh) or in client mode (searching for the most suitable internet
1405 connection in the mesh) or having the gateway support turned off
1406 entirely (which is the default setting).
1407 Aggregation=
1409 Takes a boolean value. Enables or disables aggregation of
1410 originator messages. Defaults to true.
1411 BridgeLoopAvoidance=
1413 Takes a boolean value. Enables or disables avoidance of loops on
1414 bridges. Defaults to true.
1415 DistributedArpTable=
1417 Takes a boolean value. Enables or disables the distributed ARP
1418 table. Defaults to true.
1419 Fragmentation=
1421 Takes a boolean value. Enables or disables fragmentation. Defaults
1422 to true.
1423 HopPenalty=
1425 The hop penalty setting allows one to modify batctl(8) preference
1426 for multihop routes vs. short routes. This integer value is applied
1427 to the TQ (Transmit Quality) of each forwarded OGM (Originator
1428 Message), thereby propagating the cost of an extra hop (the packet
1429 has to be received and retransmitted which costs airtime). A higher
1430 hop penalty will make it more unlikely that other nodes will choose
1431 this node as intermediate hop towards any given destination. The
1432 default hop penalty of '15' is a reasonable value for most setups
1433 and probably does not need to be changed. However, mobile nodes
1434 could choose a value of 255 (maximum value) to avoid being chosen
1435 as a router by other nodes. The minimum value is 0.
1436 OriginatorIntervalSec=
1438 The value specifies the interval in seconds, unless another time
1439 unit is specified in which batman-adv floods the network with its
1440 protocol information. See systemd.time(7) for more information.
1441 GatewayBandwidthDown=
1443 If the node is a server, this parameter is used to inform other
1444 nodes in the network about this node's internet connection download
1445 bandwidth in bits per second. Just enter any number suffixed with
1446 K, M, G or T (base 1000) and the batman-adv module will propagate
1447 the entered value in the mesh.
1448 GatewayBandwidthUp=
1450 If the node is a server, this parameter is used to inform other
1451 nodes in the network about this node's internet connection upload
1452 bandwidth in bits per second. Just enter any number suffixed with
1453 K, M, G or T (base 1000) and the batman-adv module will propagate
1454 the entered value in the mesh.
1455 RoutingAlgorithm=
1457 This can be either "batman-v" or "batman-iv" and describes which
1458 routing_algo of batctl(8) to use. The algorithm cannot be changed
1459 after interface creation. Defaults to "batman-v".
1460
1462 The [IPoIB] section only applies for netdevs of kind "ipoib" and
1463 accepts the following keys:
1464 PartitionKey=
1466 Takes an integer in the range 1...0xffff, except for 0x8000.
1467 Defaults to unset, and the kernel's default is used.
1468 Mode=
1470 Takes one of the special values "datagram" or "connected". Defaults
1471 to unset, and the kernel's default is used.
1472 When "datagram", the Infiniband unreliable datagram (UD) transport
1474 is used, and so the interface MTU is equal to the IB L2 MTU minus
1475 the IPoIB encapsulation header (4 bytes). For example, in a typical
1476 IB fabric with a 2K MTU, the IPoIB MTU will be 2048 - 4 = 2044
1477 bytes.
1478 When "connected", the Infiniband reliable connected (RC) transport
1480 is used. Connected mode takes advantage of the connected nature of
1481 the IB transport and allows an MTU up to the maximal IP packet size
1482 of 64K, which reduces the number of IP packets needed for handling
1483 large UDP datagrams, TCP segments, etc and increases the
1484 performance for large messages.
1485 IgnoreUserspaceMulticastGroup=
1487 Takes an boolean value. When true, the kernel ignores multicast
1488 groups handled by userspace. Defaults to unset, and the kernel's
1489 default is used.
1490
1492 The [WLAN] section only applies to WLAN interfaces, and accepts the
1493 following keys:
1494 PhysicalDevice=
1496 Specifies the name or index of the physical WLAN device (e.g. "0"
1497 or "phy0"). The list of the physical WLAN devices that exist on the
1498 host can be obtained by iw phy command. This option is mandatory.
1499 Type=
1501 Specifies the type of the interface. Takes one of the "ad-hoc",
1502 "station", "ap", "ap-vlan", "wds", "monitor", "mesh-point",
1503 "p2p-client", "p2p-go", "p2p-device", "ocb", and "nan". This option
1504 is mandatory.
1505 WDS=
1507 Enables the Wireless Distribution System (WDS) mode on the
1508 interface. The mode is also known as the "4 address mode". Takes a
1509 boolean value. Defaults to unset, and the kernel's default will be
1510 used.
1511
1513 Example 1. /etc/systemd/network/25-bridge.netdev
1514 [NetDev]
1516 Name=bridge0
1517 Kind=bridge
1518 Example 2. /etc/systemd/network/25-vlan1.netdev
1520 [Match]
1522 Virtualization=no
1523 [NetDev]
1525 Name=vlan1
1526 Kind=vlan
1527 [VLAN]
1529 Id=1
1530 Example 3. /etc/systemd/network/25-ipip.netdev
1532 [NetDev]
1534 Name=ipip-tun
1535 Kind=ipip
1536 MTUBytes=1480
1537 [Tunnel]
1539 Local=192.168.223.238
1540 Remote=192.169.224.239
1541 TTL=64
1542 Example 4. /etc/systemd/network/1-fou-tunnel.netdev
1544 [NetDev]
1546 Name=fou-tun
1547 Kind=fou
1548 [FooOverUDP]
1550 Port=5555
1551 Protocol=4
1552 Example 5. /etc/systemd/network/25-fou-ipip.netdev
1555 [NetDev]
1557 Name=ipip-tun
1558 Kind=ipip
1559 [Tunnel]
1561 Independent=yes
1562 Local=10.65.208.212
1563 Remote=10.65.208.211
1564 FooOverUDP=yes
1565 FOUDestinationPort=5555
1566 Example 6. /etc/systemd/network/25-tap.netdev
1569 [NetDev]
1571 Name=tap-test
1572 Kind=tap
1573 [Tap]
1575 MultiQueue=yes
1576 PacketInfo=yes
1577 Example 7. /etc/systemd/network/25-sit.netdev
1579 [NetDev]
1581 Name=sit-tun
1582 Kind=sit
1583 MTUBytes=1480
1584 [Tunnel]
1586 Local=10.65.223.238
1587 Remote=10.65.223.239
1588 Example 8. /etc/systemd/network/25-6rd.netdev
1590 [NetDev]
1592 Name=6rd-tun
1593 Kind=sit
1594 MTUBytes=1480
1595 [Tunnel]
1597 Local=10.65.223.238
1598 IPv6RapidDeploymentPrefix=2602::/24
1599 Example 9. /etc/systemd/network/25-gre.netdev
1601 [NetDev]
1603 Name=gre-tun
1604 Kind=gre
1605 MTUBytes=1480
1606 [Tunnel]
1608 Local=10.65.223.238
1609 Remote=10.65.223.239
1610 Example 10. /etc/systemd/network/25-ip6gre.netdev
1612 [NetDev]
1614 Name=ip6gre-tun
1615 Kind=ip6gre
1616 [Tunnel]
1618 Key=123
1619 Example 11. /etc/systemd/network/25-vti.netdev
1621 [NetDev]
1623 Name=vti-tun
1624 Kind=vti
1625 MTUBytes=1480
1626 [Tunnel]
1628 Local=10.65.223.238
1629 Remote=10.65.223.239
1630 Example 12. /etc/systemd/network/25-veth.netdev
1632 [NetDev]
1634 Name=veth-test
1635 Kind=veth
1636 [Peer]
1638 Name=veth-peer
1639 Example 13. /etc/systemd/network/25-bond.netdev
1641 [NetDev]
1643 Name=bond1
1644 Kind=bond
1645 [Bond]
1647 Mode=802.3ad
1648 TransmitHashPolicy=layer3+4
1649 MIIMonitorSec=1s
1650 LACPTransmitRate=fast
1651 Example 14. /etc/systemd/network/25-dummy.netdev
1653 [NetDev]
1655 Name=dummy-test
1656 Kind=dummy
1657 MACAddress=12:34:56:78:9a:bc
1658 Example 15. /etc/systemd/network/25-vrf.netdev
1660 Create a VRF interface with table 42.
1662 [NetDev]
1664 Name=vrf-test
1665 Kind=vrf
1666 [VRF]
1668 Table=42
1669 Example 16. /etc/systemd/network/25-macvtap.netdev
1671 Create a MacVTap device.
1673 [NetDev]
1675 Name=macvtap-test
1676 Kind=macvtap
1677 Example 17. /etc/systemd/network/25-wireguard.netdev
1680 [NetDev]
1682 Name=wg0
1683 Kind=wireguard
1684 [WireGuard]
1686 PrivateKey=EEGlnEPYJV//kbvvIqxKkQwOiS+UENyPncC4bF46ong=
1687 ListenPort=51820
1688 [WireGuardPeer]
1690 PublicKey=RDf+LSpeEre7YEIKaxg+wbpsNV7du+ktR99uBEtIiCA=
1691 AllowedIPs=fd31:bf08:57cb::/48,192.168.26.0/24
1692 Endpoint=wireguard.example.com:51820
1693 Example 18. /etc/systemd/network/27-xfrm.netdev
1695 [NetDev]
1697 Name=xfrm0
1698 Kind=xfrm
1699 [Xfrm]
1701 Independent=yes
1702
1704 systemd(1), systemd-networkd(8), systemd.link(5), systemd.network(5),
1705 systemd-network-generator.service(8)
1706
1708 1. Linux Ethernet Bonding Driver HOWTO
1709 https://www.kernel.org/doc/Documentation/networking/bonding.txt
1710 2. RFC 2784
1712 https://tools.ietf.org/html/rfc2784
1713 3. IEEE 802.1Q
1715 http://www.ieee802.org/1/pages/802.1Q.html
1716 4. VRF
1718 https://www.kernel.org/doc/Documentation/networking/vrf.txt
1719 5. B.A.T.M.A.N. Advanced
1721 https://www.open-mesh.org/projects/open-mesh/wiki
1722 6. (DOVE)
1724 https://en.wikipedia.org/wiki/Distributed_Overlay_Virtual_Ethernet
1725 7. VXLAN Group Policy
1727 https://tools.ietf.org/html/draft-smith-vxlan-group-policy
1728 8. Generic Protocol Extension for VXLAN
1730 https://tools.ietf.org/html/draft-ietf-nvo3-vxlan-gpe-07
1731 9. Type of Service in the Internet Protocol Suite
1733 http://tools.ietf.org/html/rfc1349
1734 10. RFC 6437
1736 https://tools.ietf.org/html/rfc6437
1737 11. RFC 2460
1739 https://tools.ietf.org/html/rfc2460
1740 12. RFC 2473
1742 https://tools.ietf.org/html/rfc2473#section-4.1.1
1743 13. ip-xfrm — transform configuration
1745 http://man7.org/linux/man-pages/man8/ip-xfrm.8.html
1746 14. Foo over UDP
1748 https://lwn.net/Articles/614348
1749 15. IPv6 Rapid Deployment
1751 https://tools.ietf.org/html/rfc5569
1752 16. Generic UDP Encapsulation
1754 https://lwn.net/Articles/615044
1755 17. Virtual XFRM Interfaces
1757 https://lwn.net/Articles/757391
1758systemd 251 SYSTEMD.NETDEV(5)