7 :abbr:`BGP` stands for Border Gateway Protocol. The latest BGP version is 4.
8 BGP-4 is one of the Exterior Gateway Protocols and the de facto standard
9 interdomain routing protocol. BGP-4 is described in :rfc:`1771` and updated by
10 :rfc:`4271`. :rfc:`2858` adds multiprotocol support to BGP-4.
17 The default configuration file of *bgpd* is :file:`bgpd.conf`. *bgpd* searches
18 the current directory first, followed by |INSTALL_PREFIX_ETC|/bgpd.conf. All of
19 *bgpd*'s commands must be configured in :file:`bgpd.conf` when the integrated
20 config is not being used.
22 *bgpd* specific invocation options are described below. Common options may also
23 be specified (:ref:`common-invocation-options`).
27 .. option:: -p, --bgp_port <port>
29 Set the bgp protocol's port number. When port number is 0, that means do not
32 .. option:: -l, --listenon
34 Specify specific IP addresses for bgpd to listen on, rather than its default
35 of ``0.0.0.0`` / ``::``. This can be useful to constrain bgpd to an internal
36 address, or to run multiple bgpd processes on one host. Multiple addresses
39 In the following example, bgpd is started listening for connections on the
40 addresses 100.0.1.2 and fd00::2:2. The options -d (runs in daemon mode) and
41 -f (uses specific configuration file) are also used in this example as we
42 are likely to run multiple bgpd instances, each one with different
43 configurations, when using -l option.
45 Note that this option implies the --no_kernel option, and no learned routes will be installed into the linux kernel.
49 # /usr/lib/frr/bgpd -d -f /some-folder/bgpd.conf -l 100.0.1.2 -l fd00::2:2
51 .. option:: -n, --no_kernel
53 Do not install learned routes into the linux kernel. This option is useful
54 for a route-reflector environment or if you are running multiple bgp
55 processes in the same namespace. This option is different than the --no_zebra
56 option in that a ZAPI connection is made.
58 This option can also be toggled during runtime by using the
59 ``[no] bgp no-rib`` commands in VTY shell.
61 Note that this option will persist after saving the configuration during
62 runtime, unless unset by the ``no bgp no-rib`` command in VTY shell prior to
63 a configuration write operation.
65 .. option:: -S, --skip_runas
67 Skip the normal process of checking capabilities and changing user and group
70 .. option:: -e, --ecmp
72 Run BGP with a limited ecmp capability, that is different than what BGP
73 was compiled with. The value specified must be greater than 0 and less
74 than or equal to the MULTIPATH_NUM specified on compilation.
76 .. option:: -Z, --no_zebra
78 Do not communicate with zebra at all. This is different than the --no_kernel
79 option in that we do not even open a ZAPI connection to the zebra process.
81 .. option:: -s, --socket_size
83 When opening tcp connections to our peers, set the socket send buffer
84 size that the kernel will use for the peers socket. This option
85 is only really useful at a very large scale. Experimentation should
86 be done to see if this is helping or not at the scale you are running
92 .. option:: -I, --int_num
94 Set zclient id. This is required when using Zebra label manager in proxy mode.
96 .. _bgp-basic-concepts:
101 .. _bgp-autonomous-systems:
108 An AS is a connected group of one or more IP prefixes run by one or more
109 network operators which has a SINGLE and CLEARLY DEFINED routing policy.
111 Each AS has an identifying number associated with it called an :abbr:`ASN
112 (Autonomous System Number)`. This is a two octet value ranging in value from 1
113 to 65535. The AS numbers 64512 through 65535 are defined as private AS numbers.
114 Private AS numbers must not be advertised on the global Internet.
116 The :abbr:`ASN (Autonomous System Number)` is one of the essential elements of
117 BGP. BGP is a distance vector routing protocol, and the AS-Path framework
118 provides distance vector metric and loop detection to BGP.
120 .. seealso:: :rfc:`1930`
122 .. _bgp-address-families:
127 Multiprotocol extensions enable BGP to carry routing information for multiple
128 network layer protocols. BGP supports an Address Family Identifier (AFI) for
129 IPv4 and IPv6. Support is also provided for multiple sets of per-AFI
130 information via the BGP Subsequent Address Family Identifier (SAFI). FRR
131 supports SAFIs for unicast information, labeled information (:rfc:`3107` and
132 :rfc:`8277`), and Layer 3 VPN information (:rfc:`4364` and :rfc:`4659`).
134 .. _bgp-route-selection:
139 The route selection process used by FRR's BGP implementation uses the following
140 decision criterion, starting at the top of the list and going towards the
141 bottom until one of the factors can be used.
145 Prefer higher local weight routes to lower routes.
147 2. **Local preference check**
149 Prefer higher local preference routes to lower.
151 If ``bgp bestpath aigp`` is enabled, and both paths that are compared have
152 AIGP attribute, BGP uses AIGP tie-breaking unless both of the paths have the
153 AIGP metric attribute. This means that the AIGP attribute is not evaluated
154 during the best path selection process between two paths when one path does
155 not have the AIGP attribute.
157 3. **Local route check**
159 Prefer local routes (statics, aggregates, redistributed) to received routes.
161 4. **AS path length check**
163 Prefer shortest hop-count AS_PATHs.
167 Prefer the lowest origin type route. That is, prefer IGP origin routes to
168 EGP, to Incomplete routes.
172 Where routes with a MED were received from the same AS, prefer the route
173 with the lowest MED. :ref:`bgp-med`.
175 7. **External check**
177 Prefer the route received from an external, eBGP peer over routes received
178 from other types of peers.
180 8. **IGP cost check**
182 Prefer the route with the lower IGP cost.
184 9. **Multi-path check**
186 If multi-pathing is enabled, then check whether the routes not yet
187 distinguished in preference may be considered equal. If
188 :clicmd:`bgp bestpath as-path multipath-relax` is set, all such routes are
189 considered equal, otherwise routes received via iBGP with identical AS_PATHs
190 or routes received from eBGP neighbours in the same AS are considered equal.
192 10. **Already-selected external check**
194 Where both routes were received from eBGP peers, then prefer the route
195 which is already selected. Note that this check is not applied if
196 :clicmd:`bgp bestpath compare-routerid` is configured. This check can
197 prevent some cases of oscillation.
199 11. **Router-ID check**
201 Prefer the route with the lowest `router-ID`. If the route has an
202 `ORIGINATOR_ID` attribute, through iBGP reflection, then that router ID is
203 used, otherwise the `router-ID` of the peer the route was received from is
206 12. **Cluster-List length check**
208 The route with the shortest cluster-list length is used. The cluster-list
209 reflects the iBGP reflection path the route has taken.
213 Prefer the route received from the peer with the higher transport layer
214 address, as a last-resort tie-breaker.
216 .. _bgp-capability-negotiation:
218 Capability Negotiation
219 ----------------------
221 When adding IPv6 routing information exchange feature to BGP. There were some
222 proposals. :abbr:`IETF (Internet Engineering Task Force)`
223 :abbr:`IDR (Inter Domain Routing)` adopted a proposal called Multiprotocol
224 Extension for BGP. The specification is described in :rfc:`2283`. The protocol
225 does not define new protocols. It defines new attributes to existing BGP. When
226 it is used exchanging IPv6 routing information it is called BGP-4+. When it is
227 used for exchanging multicast routing information it is called MBGP.
229 *bgpd* supports Multiprotocol Extension for BGP. So if a remote peer supports
230 the protocol, *bgpd* can exchange IPv6 and/or multicast routing information.
232 Traditional BGP did not have the feature to detect a remote peer's
233 capabilities, e.g. whether it can handle prefix types other than IPv4 unicast
234 routes. This was a big problem using Multiprotocol Extension for BGP in an
235 operational network. :rfc:`2842` adopted a feature called Capability
236 Negotiation. *bgpd* use this Capability Negotiation to detect the remote peer's
237 capabilities. If a peer is only configured as an IPv4 unicast neighbor, *bgpd*
238 does not send these Capability Negotiation packets (at least not unless other
239 optional BGP features require capability negotiation).
241 By default, FRR will bring up peering with minimal common capability for the
242 both sides. For example, if the local router has unicast and multicast
243 capabilities and the remote router only has unicast capability the local router
244 will establish the connection with unicast only capability. When there are no
245 common capabilities, FRR sends Unsupported Capability error and then resets the
248 .. _bgp-router-configuration:
250 BGP Router Configuration
251 ========================
256 First of all you must configure BGP router with the :clicmd:`router bgp ASN`
257 command. The AS number is an identifier for the autonomous system. The BGP
258 protocol uses the AS number for detecting whether the BGP connection is
259 internal or external.
261 .. clicmd:: router bgp ASN
263 Enable a BGP protocol process with the specified ASN. After
264 this statement you can input any `BGP Commands`.
266 .. clicmd:: bgp router-id A.B.C.D
268 This command specifies the router-ID. If *bgpd* connects to *zebra* it gets
269 interface and address information. In that case default router ID value is
270 selected as the largest IP Address of the interfaces. When `router zebra` is
271 not enabled *bgpd* can't get interface information so `router-id` is set to
272 0.0.0.0. So please set router-id by hand.
275 .. _bgp-multiple-autonomous-systems:
277 Multiple Autonomous Systems
278 ---------------------------
280 FRR's BGP implementation is capable of running multiple autonomous systems at
281 once. Each configured AS corresponds to a :ref:`zebra-vrf`. In the past, to get
282 the same functionality the network administrator had to run a new *bgpd*
283 process; using VRFs allows multiple autonomous systems to be handled in a
286 When using multiple autonomous systems, all router config blocks after the
287 first one must specify a VRF to be the target of BGP's route selection. This
288 VRF must be unique within respect to all other VRFs being used for the same
289 purpose, i.e. two different autonomous systems cannot use the same VRF.
290 However, the same AS can be used with different VRFs.
294 The separated nature of VRFs makes it possible to peer a single *bgpd*
295 process to itself, on one machine. Note that this can be done fully within
296 BGP without a corresponding VRF in the kernel or Zebra, which enables some
297 practical use cases such as :ref:`route reflectors <bgp-route-reflector>`
300 Configuration of additional autonomous systems, or of a router that targets a
301 specific VRF, is accomplished with the following command:
303 .. clicmd:: router bgp ASN vrf VRFNAME
305 ``VRFNAME`` is matched against VRFs configured in the kernel. When ``vrf
306 VRFNAME`` is not specified, the BGP protocol process belongs to the default
309 An example configuration with multiple autonomous systems might look like this:
314 neighbor 10.0.0.1 remote-as 20
315 neighbor 10.0.0.2 remote-as 30
317 router bgp 2 vrf blue
318 neighbor 10.0.0.3 remote-as 40
319 neighbor 10.0.0.4 remote-as 50
322 neighbor 10.0.0.5 remote-as 60
323 neighbor 10.0.0.6 remote-as 70
326 .. seealso:: :ref:`bgp-vrf-route-leaking`
327 .. seealso:: :ref:`zebra-vrf`
335 In addition to supporting multiple autonomous systems, FRR's BGP implementation
336 also supports *views*.
338 BGP views are almost the same as normal BGP processes, except that routes
339 selected by BGP are not installed into the kernel routing table. Each BGP view
340 provides an independent set of routing information which is only distributed
341 via BGP. Multiple views can be supported, and BGP view information is always
342 independent from other routing protocols and Zebra/kernel routes. BGP views use
343 the core instance (i.e., default VRF) for communication with peers.
345 .. clicmd:: router bgp AS-NUMBER view NAME
347 Make a new BGP view. You can use an arbitrary word for the ``NAME``. Routes
348 selected by the view are not installed into the kernel routing table.
350 With this command, you can setup Route Server like below.
356 neighbor 10.0.0.1 remote-as 2
357 neighbor 10.0.0.2 remote-as 3
360 neighbor 10.0.0.3 remote-as 4
361 neighbor 10.0.0.4 remote-as 5
363 .. clicmd:: show [ip] bgp view NAME
365 Display the routing table of BGP view ``NAME``.
371 .. clicmd:: bgp bestpath as-path confed
373 This command specifies that the length of confederation path sets and
374 sequences should should be taken into account during the BGP best path
377 .. clicmd:: bgp bestpath as-path multipath-relax
379 This command specifies that BGP decision process should consider paths
380 of equal AS_PATH length candidates for multipath computation. Without
381 the knob, the entire AS_PATH must match for multipath computation.
383 .. clicmd:: bgp bestpath compare-routerid
385 Ensure that when comparing routes where both are equal on most metrics,
386 including local-pref, AS_PATH length, IGP cost, MED, that the tie is broken
389 If this option is enabled, then the already-selected check, where
390 already selected eBGP routes are preferred, is skipped.
392 If a route has an `ORIGINATOR_ID` attribute because it has been reflected,
393 that `ORIGINATOR_ID` will be used. Otherwise, the router-ID of the peer the
394 route was received from will be used.
396 The advantage of this is that the route-selection (at this point) will be
397 more deterministic. The disadvantage is that a few or even one lowest-ID
398 router may attract all traffic to otherwise-equal paths because of this
399 check. It may increase the possibility of MED or IGP oscillation, unless
400 other measures were taken to avoid these. The exact behaviour will be
401 sensitive to the iBGP and reflection topology.
403 .. clicmd:: bgp bestpath peer-type multipath-relax
405 This command specifies that BGP decision process should consider paths
406 from all peers for multipath computation. If this option is enabled,
407 paths learned from any of eBGP, iBGP, or confederation neighbors will
408 be multipath if they are otherwise considered equal cost.
410 .. clicmd:: bgp bestpath aigp
412 Use the bgp bestpath aigp command to evaluate the AIGP attribute during
413 the best path selection process between two paths that have the AIGP
416 When bgp bestpath aigp is disabled, BGP does not use AIGP tie-breaking
417 rules unless paths have the AIGP attribute.
421 .. clicmd:: maximum-paths (1-128)
423 Sets the maximum-paths value used for ecmp calculations for this
424 bgp instance in EBGP. The maximum value listed, 128, can be limited by
425 the ecmp cli for bgp or if the daemon was compiled with a lower
426 ecmp value. This value can also be set in ipv4/ipv6 unicast/labeled
427 unicast to only affect those particular afi/safi's.
429 .. clicmd:: maximum-paths ibgp (1-128) [equal-cluster-length]
431 Sets the maximum-paths value used for ecmp calculations for this
432 bgp instance in IBGP. The maximum value listed, 128, can be limited by
433 the ecmp cli for bgp or if the daemon was compiled with a lower
434 ecmp value. This value can also be set in ipv4/ipv6 unicast/labeled
435 unicast to only affect those particular afi/safi's.
439 Administrative Distance Metrics
440 -------------------------------
442 .. clicmd:: distance bgp (1-255) (1-255) (1-255)
444 This command changes distance value of BGP. The arguments are the distance
445 values for external routes, internal routes and local routes
448 .. clicmd:: distance (1-255) A.B.C.D/M
450 .. clicmd:: distance (1-255) A.B.C.D/M WORD
452 Sets the administrative distance for a particular route.
454 .. _bgp-requires-policy:
456 Require policy on EBGP
457 -------------------------------
459 .. clicmd:: bgp ebgp-requires-policy
461 This command requires incoming and outgoing filters to be applied
462 for eBGP sessions as part of RFC-8212 compliance. Without the incoming
463 filter, no routes will be accepted. Without the outgoing filter, no
464 routes will be announced.
466 This is enabled by default for the traditional configuration and
467 turned off by default for datacenter configuration.
469 When you enable/disable this option you MUST clear the session.
471 When the incoming or outgoing filter is missing you will see
472 "(Policy)" sign under ``show bgp summary``:
476 exit1# show bgp summary
478 IPv4 Unicast Summary (VRF default):
479 BGP router identifier 10.10.10.1, local AS number 65001 vrf-id 0
481 RIB entries 7, using 1344 bytes of memory
482 Peers 2, using 43 KiB of memory
484 Neighbor V AS MsgRcvd MsgSent TblVer InQ OutQ Up/Down State/PfxRcd PfxSnt Desc
485 192.168.0.2 4 65002 8 10 0 0 0 00:03:09 5 (Policy) N/A
486 fe80:1::2222 4 65002 9 11 0 0 0 00:03:09 (Policy) (Policy) N/A
488 Additionally a `show bgp neighbor` command would indicate in the `For address family:`
493 exit1# show bgp neighbor
495 For address family: IPv4 Unicast
496 Update group 1, subgroup 1
497 Packet Queue length 0
498 Inbound soft reconfiguration allowed
499 Community attribute sent to this neighbor(all)
500 Inbound updates discarded due to missing policy
501 Outbound updates discarded due to missing policy
504 Reject routes with AS_SET or AS_CONFED_SET types
505 ------------------------------------------------
507 .. clicmd:: bgp reject-as-sets
509 This command enables rejection of incoming and outgoing routes having AS_SET or AS_CONFED_SET type.
511 Suppress duplicate updates
512 --------------------------
514 .. clicmd:: bgp suppress-duplicates
516 For example, BGP routers can generate multiple identical announcements with
517 empty community attributes if stripped at egress. This is an undesired behavior.
518 Suppress duplicate updates if the route actually not changed.
521 Send Hard Reset CEASE Notification for Administrative Reset
522 -----------------------------------------------------------
524 .. clicmd:: bgp hard-administrative-reset
526 Send Hard Reset CEASE Notification for 'Administrative Reset' events.
528 When disabled, and Graceful Restart Notification capability is exchanged
529 between the peers, Graceful Restart procedures apply, and routes will be
534 Disable checking if nexthop is connected on EBGP sessions
535 ---------------------------------------------------------
537 .. clicmd:: bgp disable-ebgp-connected-route-check
539 This command is used to disable the connection verification process for EBGP peering sessions
540 that are reachable by a single hop but are configured on a loopback interface or otherwise
541 configured with a non-directly connected IP address.
543 .. _bgp-route-flap-dampening:
548 .. clicmd:: bgp dampening (1-45) (1-20000) (1-50000) (1-255)
550 This command enables BGP route-flap dampening and specifies dampening parameters.
553 Half-life time for the penalty
556 Value to start reusing a route
559 Value to start suppressing a route
562 Maximum duration to suppress a stable route
564 The route-flap damping algorithm is compatible with :rfc:`2439`. The use of
565 this command is not recommended nowadays.
567 At the moment, route-flap dampening is not working per VRF and is working only
568 for IPv4 unicast and multicast.
571 https://www.ripe.net/publications/docs/ripe-378
575 Multi-Exit Discriminator
576 ------------------------
578 The BGP :abbr:`MED (Multi-Exit Discriminator)` attribute has properties which
579 can cause subtle convergence problems in BGP. These properties and problems
580 have proven to be hard to understand, at least historically, and may still not
581 be widely understood. The following attempts to collect together and present
582 what is known about MED, to help operators and FRR users in designing and
583 configuring their networks.
585 The BGP :abbr:`MED` attribute is intended to allow one AS to indicate its
586 preferences for its ingress points to another AS. The MED attribute will not be
587 propagated on to another AS by the receiving AS - it is 'non-transitive' in the
590 E.g., if AS X and AS Y have 2 different BGP peering points, then AS X might set
591 a MED of 100 on routes advertised at one and a MED of 200 at the other. When AS
592 Y selects between otherwise equal routes to or via AS X, AS Y should prefer to
593 take the path via the lower MED peering of 100 with AS X. Setting the MED
594 allows an AS to influence the routing taken to it within another, neighbouring
597 In this use of MED it is not really meaningful to compare the MED value on
598 routes where the next AS on the paths differs. E.g., if AS Y also had a route
599 for some destination via AS Z in addition to the routes from AS X, and AS Z had
600 also set a MED, it wouldn't make sense for AS Y to compare AS Z's MED values to
601 those of AS X. The MED values have been set by different administrators, with
602 different frames of reference.
604 The default behaviour of BGP therefore is to not compare MED values across
605 routes received from different neighbouring ASes. In FRR this is done by
606 comparing the neighbouring, left-most AS in the received AS_PATHs of the routes
607 and only comparing MED if those are the same.
609 Unfortunately, this behaviour of MED, of sometimes being compared across routes
610 and sometimes not, depending on the properties of those other routes, means MED
611 can cause the order of preference over all the routes to be undefined. That is,
612 given routes A, B, and C, if A is preferred to B, and B is preferred to C, then
613 a well-defined order should mean the preference is transitive (in the sense of
614 orders [#med-transitivity-rant]_) and that A would be preferred to C.
616 However, when MED is involved this need not be the case. With MED it is
617 possible that C is actually preferred over A. So A is preferred to B, B is
618 preferred to C, but C is preferred to A. This can be true even where BGP
619 defines a deterministic 'most preferred' route out of the full set of A,B,C.
620 With MED, for any given set of routes there may be a deterministically
621 preferred route, but there need not be any way to arrange them into any order
622 of preference. With unmodified MED, the order of preference of routes literally
625 That MED can induce non-transitive preferences over routes can cause issues.
626 Firstly, it may be perceived to cause routing table churn locally at speakers;
627 secondly, and more seriously, it may cause routing instability in iBGP
628 topologies, where sets of speakers continually oscillate between different
631 The first issue arises from how speakers often implement routing decisions.
632 Though BGP defines a selection process that will deterministically select the
633 same route as best at any given speaker, even with MED, that process requires
634 evaluating all routes together. For performance and ease of implementation
635 reasons, many implementations evaluate route preferences in a pair-wise fashion
636 instead. Given there is no well-defined order when MED is involved, the best
637 route that will be chosen becomes subject to implementation details, such as
638 the order the routes are stored in. That may be (locally) non-deterministic,
639 e.g.: it may be the order the routes were received in.
641 This indeterminism may be considered undesirable, though it need not cause
642 problems. It may mean additional routing churn is perceived, as sometimes more
643 updates may be produced than at other times in reaction to some event .
645 This first issue can be fixed with a more deterministic route selection that
646 ensures routes are ordered by the neighbouring AS during selection.
647 :clicmd:`bgp deterministic-med`. This may reduce the number of updates as routes
648 are received, and may in some cases reduce routing churn. Though, it could
649 equally deterministically produce the largest possible set of updates in
650 response to the most common sequence of received updates.
652 A deterministic order of evaluation tends to imply an additional overhead of
653 sorting over any set of n routes to a destination. The implementation of
654 deterministic MED in FRR scales significantly worse than most sorting
655 algorithms at present, with the number of paths to a given destination. That
656 number is often low enough to not cause any issues, but where there are many
657 paths, the deterministic comparison may quickly become increasingly expensive
660 Deterministic local evaluation can *not* fix the second, more major, issue of
661 MED however. Which is that the non-transitive preference of routes MED can
662 cause may lead to routing instability or oscillation across multiple speakers
663 in iBGP topologies. This can occur with full-mesh iBGP, but is particularly
664 problematic in non-full-mesh iBGP topologies that further reduce the routing
665 information known to each speaker. This has primarily been documented with iBGP
666 :ref:`route-reflection <bgp-route-reflector>` topologies. However, any
667 route-hiding technologies potentially could also exacerbate oscillation with MED.
669 This second issue occurs where speakers each have only a subset of routes, and
670 there are cycles in the preferences between different combinations of routes -
671 as the undefined order of preference of MED allows - and the routes are
672 distributed in a way that causes the BGP speakers to 'chase' those cycles. This
673 can occur even if all speakers use a deterministic order of evaluation in route
676 E.g., speaker 4 in AS A might receive a route from speaker 2 in AS X, and from
677 speaker 3 in AS Y; while speaker 5 in AS A might receive that route from
678 speaker 1 in AS Y. AS Y might set a MED of 200 at speaker 1, and 100 at speaker
679 3. I.e, using ASN:ID:MED to label the speakers:
685 X:2------|--A:4-------A:5--|-Y:1:200
691 Assuming all other metrics are equal (AS_PATH, ORIGIN, 0 IGP costs), then based
692 on the RFC4271 decision process speaker 4 will choose X:2 over Y:3:100, based
693 on the lower ID of 2. Speaker 4 advertises X:2 to speaker 5. Speaker 5 will
694 continue to prefer Y:1:200 based on the ID, and advertise this to speaker 4.
695 Speaker 4 will now have the full set of routes, and the Y:1:200 it receives
696 from 5 will beat X:2, but when speaker 4 compares Y:1:200 to Y:3:100 the MED
697 check now becomes active as the ASes match, and now Y:3:100 is preferred.
698 Speaker 4 therefore now advertises Y:3:100 to 5, which will also agrees that
699 Y:3:100 is preferred to Y:1:200, and so withdraws the latter route from 4.
700 Speaker 4 now has only X:2 and Y:3:100, and X:2 beats Y:3:100, and so speaker 4
701 implicitly updates its route to speaker 5 to X:2. Speaker 5 sees that Y:1:200
702 beats X:2 based on the ID, and advertises Y:1:200 to speaker 4, and the cycle
705 The root cause is the lack of a clear order of preference caused by how MED
706 sometimes is and sometimes is not compared, leading to this cycle in the
707 preferences between the routes:
712 /---> X:2 ---beats---> Y:3:100 --\\
715 \\---beats--- Y:1:200 <---beats---/
719 This particular type of oscillation in full-mesh iBGP topologies can be
720 avoided by speakers preferring already selected, external routes rather than
721 choosing to update to new a route based on a post-MED metric (e.g. router-ID),
722 at the cost of a non-deterministic selection process. FRR implements this, as
723 do many other implementations, so long as it is not overridden by setting
724 :clicmd:`bgp bestpath compare-routerid`, and see also
725 :ref:`bgp-route-selection`.
727 However, more complex and insidious cycles of oscillation are possible with
728 iBGP route-reflection, which are not so easily avoided. These have been
729 documented in various places. See, e.g.:
731 - [bgp-route-osci-cond]_
732 - [stable-flexible-ibgp]_
733 - [ibgp-correctness]_
735 for concrete examples and further references.
737 There is as of this writing *no* known way to use MED for its original purpose;
738 *and* reduce routing information in iBGP topologies; *and* be sure to avoid the
739 instability problems of MED due the non-transitive routing preferences it can
740 induce; in general on arbitrary networks.
742 There may be iBGP topology specific ways to reduce the instability risks, even
743 while using MED, e.g.: by constraining the reflection topology and by tuning
744 IGP costs between route-reflector clusters, see :rfc:`3345` for details. In the
745 near future, the Add-Path extension to BGP may also solve MED oscillation while
746 still allowing MED to be used as intended, by distributing "best-paths per
747 neighbour AS". This would be at the cost of distributing at least as many
748 routes to all speakers as a full-mesh iBGP would, if not more, while also
749 imposing similar CPU overheads as the "Deterministic MED" feature at each
752 More generally, the instability problems that MED can introduce on more
753 complex, non-full-mesh, iBGP topologies may be avoided either by:
755 - Setting :clicmd:`bgp always-compare-med`, however this allows MED to be compared
756 across values set by different neighbour ASes, which may not produce
757 coherent desirable results, of itself.
758 - Effectively ignoring MED by setting MED to the same value (e.g.: 0) using
759 :clicmd:`set metric METRIC` on all received routes, in combination with
760 setting :clicmd:`bgp always-compare-med` on all speakers. This is the simplest
761 and most performant way to avoid MED oscillation issues, where an AS is happy
762 not to allow neighbours to inject this problematic metric.
764 As MED is evaluated after the AS_PATH length check, another possible use for
765 MED is for intra-AS steering of routes with equal AS_PATH length, as an
766 extension of the last case above. As MED is evaluated before IGP metric, this
767 can allow cold-potato routing to be implemented to send traffic to preferred
768 hand-offs with neighbours, rather than the closest hand-off according to the
771 Note that even if action is taken to address the MED non-transitivity issues,
772 other oscillations may still be possible. E.g., on IGP cost if iBGP and IGP
773 topologies are at cross-purposes with each other - see the Flavel and Roughan
774 paper above for an example. Hence the guideline that the iBGP topology should
775 follow the IGP topology.
777 .. clicmd:: bgp deterministic-med
779 Carry out route-selection in way that produces deterministic answers
780 locally, even in the face of MED and the lack of a well-defined order of
781 preference it can induce on routes. Without this option the preferred route
782 with MED may be determined largely by the order that routes were received
785 Setting this option will have a performance cost that may be noticeable when
786 there are many routes for each destination. Currently in FRR it is
787 implemented in a way that scales poorly as the number of routes per
788 destination increases.
790 The default is that this option is not set.
792 Note that there are other sources of indeterminism in the route selection
793 process, specifically, the preference for older and already selected routes
794 from eBGP peers, :ref:`bgp-route-selection`.
796 .. clicmd:: bgp always-compare-med
798 Always compare the MED on routes, even when they were received from
799 different neighbouring ASes. Setting this option makes the order of
800 preference of routes more defined, and should eliminate MED induced
803 If using this option, it may also be desirable to use
804 :clicmd:`set metric METRIC` to set MED to 0 on routes received from external
807 This option can be used, together with :clicmd:`set metric METRIC` to use
808 MED as an intra-AS metric to steer equal-length AS_PATH routes to, e.g.,
812 .. _bgp-graceful-restart:
817 BGP graceful restart functionality as defined in
818 `RFC-4724 <https://tools.ietf.org/html/rfc4724/>`_ defines the mechanisms that
819 allows BGP speaker to continue to forward data packets along known routes
820 while the routing protocol information is being restored.
823 Usually, when BGP on a router restarts, all the BGP peers detect that the
824 session went down and then came up. This "down/up" transition results in a
825 "routing flap" and causes BGP route re-computation, generation of BGP routing
826 updates, and unnecessary churn to the forwarding tables.
828 The following functionality is provided by graceful restart:
830 1. The feature allows the restarting router to indicate to the helping peer the
831 routes it can preserve in its forwarding plane during control plane restart
832 by sending graceful restart capability in the OPEN message sent during
833 session establishment.
834 2. The feature allows helping router to advertise to all other peers the routes
835 received from the restarting router which are preserved in the forwarding
836 plane of the restarting router during control plane restart.
843 (R1)-----------------------------------------------------------------(R2)
845 1. BGP Graceful Restart Capability exchanged between R1 & R2.
847 <--------------------------------------------------------------------->
849 2. Kill BGP Process at R1.
851 ---------------------------------------------------------------------->
853 3. R2 Detects the above BGP Restart & verifies BGP Restarting
856 4. Start BGP Process at R1.
858 5. Re-establish the BGP session between R1 & R2.
860 <--------------------------------------------------------------------->
862 6. R2 Send initial route updates, followed by End-Of-Rib.
864 <----------------------------------------------------------------------
866 7. R1 was waiting for End-Of-Rib from R2 & which has been received
869 8. R1 now runs BGP Best-Path algorithm. Send Initial BGP Update,
870 followed by End-Of Rib
872 <--------------------------------------------------------------------->
875 .. _bgp-GR-preserve-forwarding-state:
877 BGP-GR Preserve-Forwarding State
878 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
880 BGP OPEN message carrying optional capabilities for Graceful Restart has
881 8 bit “Flags for Address Family” for given AFI and SAFI. This field contains
882 bit flags relating to routes that were advertised with the given AFI and SAFI.
891 The most significant bit is defined as the Forwarding State (F) bit, which
892 can be used to indicate whether the forwarding state for routes that were
893 advertised with the given AFI and SAFI has indeed been preserved during the
894 previous BGP restart. When set (value 1), the bit indicates that the
895 forwarding state has been preserved.
896 The remaining bits are reserved and MUST be set to zero by the sender and
897 ignored by the receiver.
899 .. clicmd:: bgp graceful-restart preserve-fw-state
901 FRR gives us the option to enable/disable the "F" flag using this specific
902 vty command. However, it doesn't have the option to enable/disable
903 this flag only for specific AFI/SAFI i.e. when this command is used, it
904 applied to all the supported AFI/SAFI combinations for this peer.
906 .. _bgp-end-of-rib-message:
908 End-of-RIB (EOR) message
909 ^^^^^^^^^^^^^^^^^^^^^^^^
911 An UPDATE message with no reachable Network Layer Reachability Information
912 (NLRI) and empty withdrawn NLRI is specified as the End-of-RIB marker that can
913 be used by a BGP speaker to indicate to its peer the completion of the initial
914 routing update after the session is established.
916 For the IPv4 unicast address family, the End-of-RIB marker is an UPDATE message
917 with the minimum length. For any other address family, it is an UPDATE message
918 that contains only the MP_UNREACH_NLRI attribute with no withdrawn routes for
921 Although the End-of-RIB marker is specified for the purpose of BGP graceful
922 restart, it is noted that the generation of such a marker upon completion of
923 the initial update would be useful for routing convergence in general, and thus
924 the practice is recommended.
926 .. _bgp-route-selection-deferral-timer:
928 Route Selection Deferral Timer
929 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
931 Specifies the time the restarting router defers the route selection process
934 Restarting Router : The usage of route election deferral timer is specified
935 in https://tools.ietf.org/html/rfc4724#section-4.1
937 Once the session between the Restarting Speaker and the Receiving Speaker is
938 re-established, the Restarting Speaker will receive and process BGP messages
941 However, it MUST defer route selection for an address family until it either.
943 1. Receives the End-of-RIB marker from all its peers (excluding the ones with
944 the "Restart State" bit set in the received capability and excluding the ones
945 that do not advertise the graceful restart capability).
946 2. The Selection_Deferral_Timer timeout.
948 .. clicmd:: bgp graceful-restart select-defer-time (0-3600)
950 This is command, will set deferral time to value specified.
953 .. clicmd:: bgp graceful-restart rib-stale-time (1-3600)
955 This is command, will set the time for which stale routes are kept in RIB.
957 .. clicmd:: bgp graceful-restart restart-time (0-4095)
959 Set the time to wait to delete stale routes before a BGP open message
962 Using with Long-lived Graceful Restart capability, this is recommended
963 setting this timer to 0 and control stale routes with
964 ``bgp long-lived-graceful-restart stale-time``.
966 Default value is 120.
968 .. clicmd:: bgp graceful-restart stalepath-time (1-4095)
970 This is command, will set the max time (in seconds) to hold onto
971 restarting peer's stale paths.
973 It also controls Enhanced Route-Refresh timer.
975 If this command is configured and the router does not receive a Route-Refresh EoRR
976 message, the router removes the stale routes from the BGP table after the timer
977 expires. The stale path timer is started when the router receives a Route-Refresh
980 .. clicmd:: bgp graceful-restart notification
982 Indicate Graceful Restart support for BGP NOTIFICATION messages.
984 After changing this parameter, you have to reset the peers in order to advertise
985 N-bit in Graceful Restart capability.
987 Without Graceful-Restart Notification capability (N-bit not set), GR is not
988 activated when receiving CEASE/HOLDTIME expire notifications.
990 When sending ``CEASE/Administrative Reset`` (``clear bgp``), the session is closed
991 and routes are not retained. When N-bit is set and ``bgp hard-administrative-reset``
992 is turned off Graceful-Restart is activated and routes are retained.
996 .. _bgp-per-peer-graceful-restart:
998 BGP Per Peer Graceful Restart
999 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1001 Ability to enable and disable graceful restart, helper and no GR at all mode
1002 functionality at peer level.
1004 So bgp graceful restart can be enabled at modes global BGP level or at per
1005 peer level. There are two FSM, one for BGP GR global mode and other for peer
1008 Default global mode is helper and default peer per mode is inherit from global.
1009 If per peer mode is configured, the GR mode of this particular peer will
1010 override the global mode.
1012 .. _bgp-GR-global-mode-cmd:
1014 BGP GR Global Mode Commands
1015 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
1017 .. clicmd:: bgp graceful-restart
1019 This command will enable BGP graceful restart functionality at the global
1022 .. clicmd:: bgp graceful-restart disable
1024 This command will disable both the functionality graceful restart and helper
1028 .. _bgp-GR-peer-mode-cmd:
1030 BGP GR Peer Mode Commands
1031 ^^^^^^^^^^^^^^^^^^^^^^^^^
1033 .. clicmd:: neighbor A.B.C.D graceful-restart
1035 This command will enable BGP graceful restart functionality at the peer
1038 .. clicmd:: neighbor A.B.C.D graceful-restart-helper
1040 This command will enable BGP graceful restart helper only functionality
1043 .. clicmd:: neighbor A.B.C.D graceful-restart-disable
1045 This command will disable the entire BGP graceful restart functionality
1049 Long-lived Graceful Restart
1050 ---------------------------
1052 Currently, only restarter mode is supported. This capability is advertised only
1053 if graceful restart capability is negotiated.
1055 .. clicmd:: bgp long-lived-graceful-restart stale-time (1-16777215)
1057 Specifies the maximum time to wait before purging long-lived stale routes for
1060 Default is 0, which means the feature is off by default. Only graceful
1061 restart takes into account.
1065 Administrative Shutdown
1066 -----------------------
1068 .. clicmd:: bgp shutdown [message MSG...]
1070 Administrative shutdown of all peers of a bgp instance. Drop all BGP peers,
1071 but preserve their configurations. The peers are notified in accordance with
1072 `RFC 8203 <https://tools.ietf.org/html/rfc8203/>`_ by sending a
1073 ``NOTIFICATION`` message with error code ``Cease`` and subcode
1074 ``Administrative Shutdown`` prior to terminating connections. This global
1075 shutdown is independent of the neighbor shutdown, meaning that individually
1076 shut down peers will not be affected by lifting it.
1078 An optional shutdown message `MSG` can be specified.
1086 .. clicmd:: network A.B.C.D/M
1088 This command adds the announcement network.
1093 address-family ipv4 unicast
1097 This configuration example says that network 10.0.0.0/8 will be
1098 announced to all neighbors. Some vendors' routers don't advertise
1099 routes if they aren't present in their IGP routing tables; `bgpd`
1100 doesn't care about IGP routes when announcing its routes.
1103 .. clicmd:: bgp network import-check
1105 This configuration modifies the behavior of the network statement.
1106 If you have this configured the underlying network must exist in
1107 the rib. If you have the [no] form configured then BGP will not
1108 check for the networks existence in the rib. For versions 7.3 and
1109 before frr defaults for datacenter were the network must exist,
1110 traditional did not check for existence. For versions 7.4 and beyond
1111 both traditional and datacenter the network must exist.
1113 .. _bgp-ipv6-support:
1118 .. clicmd:: neighbor A.B.C.D activate
1120 This configuration modifies whether to enable an address family for a
1121 specific neighbor. By default only the IPv4 unicast address family is
1127 address-family ipv6 unicast
1128 neighbor 2001:0DB8::1 activate
1129 network 2001:0DB8:5009::/64
1132 This configuration example says that network 2001:0DB8:5009::/64 will be
1133 announced and enables the neighbor 2001:0DB8::1 to receive this announcement.
1135 By default, only the IPv4 unicast address family is announced to all
1136 neighbors. Using the 'no bgp default ipv4-unicast' configuration overrides
1137 this default so that all address families need to be enabled explicitly.
1142 no bgp default ipv4-unicast
1143 neighbor 10.10.10.1 remote-as 2
1144 neighbor 2001:0DB8::1 remote-as 3
1145 address-family ipv4 unicast
1146 neighbor 10.10.10.1 activate
1147 network 192.168.1.0/24
1149 address-family ipv6 unicast
1150 neighbor 2001:0DB8::1 activate
1151 network 2001:0DB8:5009::/64
1154 This configuration demonstrates how the 'no bgp default ipv4-unicast' might
1155 be used in a setup with two upstreams where each of the upstreams should only
1156 receive either IPv4 or IPv6 announcements.
1158 Using the ``bgp default ipv6-unicast`` configuration, IPv6 unicast
1159 address family is enabled by default for all new neighbors.
1162 .. _bgp-route-aggregation:
1167 .. _bgp-route-aggregation-ipv4:
1169 Route Aggregation-IPv4 Address Family
1170 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1172 .. clicmd:: aggregate-address A.B.C.D/M
1174 This command specifies an aggregate address.
1176 In order to advertise an aggregated prefix, a more specific (longer) prefix
1177 MUST exist in the BGP table. For example, if you want to create an
1178 ``aggregate-address 10.0.0.0/24``, you should make sure you have something
1179 like ``10.0.0.5/32`` or ``10.0.0.0/26``, or any other smaller prefix in the
1180 BGP table. The routing information table (RIB) is not enough, you have to
1181 redistribute them into the BGP table.
1183 .. clicmd:: aggregate-address A.B.C.D/M route-map NAME
1185 Apply a route-map for an aggregated prefix.
1187 .. clicmd:: aggregate-address A.B.C.D/M origin <egp|igp|incomplete>
1189 Override ORIGIN for an aggregated prefix.
1191 .. clicmd:: aggregate-address A.B.C.D/M as-set
1193 This command specifies an aggregate address. Resulting routes include
1196 .. clicmd:: aggregate-address A.B.C.D/M summary-only
1198 This command specifies an aggregate address.
1200 Longer prefixes advertisements of more specific routes to all neighbors are suppressed.
1202 .. clicmd:: aggregate-address A.B.C.D/M matching-MED-only
1204 Configure the aggregated address to only be created when the routes MED
1205 match, otherwise no aggregated route will be created.
1207 .. clicmd:: aggregate-address A.B.C.D/M suppress-map NAME
1209 Similar to `summary-only`, but will only suppress more specific routes that
1210 are matched by the selected route-map.
1213 This configuration example sets up an ``aggregate-address`` under the ipv4
1219 address-family ipv4 unicast
1220 aggregate-address 10.0.0.0/8
1221 aggregate-address 20.0.0.0/8 as-set
1222 aggregate-address 40.0.0.0/8 summary-only
1223 aggregate-address 50.0.0.0/8 route-map aggr-rmap
1227 .. _bgp-route-aggregation-ipv6:
1229 Route Aggregation-IPv6 Address Family
1230 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1232 .. clicmd:: aggregate-address X:X::X:X/M
1234 This command specifies an aggregate address.
1236 .. clicmd:: aggregate-address X:X::X:X/M route-map NAME
1238 Apply a route-map for an aggregated prefix.
1240 .. clicmd:: aggregate-address X:X::X:X/M origin <egp|igp|incomplete>
1242 Override ORIGIN for an aggregated prefix.
1244 .. clicmd:: aggregate-address X:X::X:X/M as-set
1246 This command specifies an aggregate address. Resulting routes include
1249 .. clicmd:: aggregate-address X:X::X:X/M summary-only
1251 This command specifies an aggregate address.
1253 Longer prefixes advertisements of more specific routes to all neighbors are suppressed
1255 .. clicmd:: aggregate-address X:X::X:X/M matching-MED-only
1257 Configure the aggregated address to only be created when the routes MED
1258 match, otherwise no aggregated route will be created.
1260 .. clicmd:: aggregate-address X:X::X:X/M suppress-map NAME
1262 Similar to `summary-only`, but will only suppress more specific routes that
1263 are matched by the selected route-map.
1266 This configuration example sets up an ``aggregate-address`` under the ipv6
1272 address-family ipv6 unicast
1273 aggregate-address 10::0/64
1274 aggregate-address 20::0/64 as-set
1275 aggregate-address 40::0/64 summary-only
1276 aggregate-address 50::0/64 route-map aggr-rmap
1280 .. _bgp-redistribute-to-bgp:
1285 Redistribution configuration should be placed under the ``address-family``
1286 section for the specific AF to redistribute into. Protocol availability for
1287 redistribution is determined by BGP AF; for example, you cannot redistribute
1288 OSPFv3 into ``address-family ipv4 unicast`` as OSPFv3 supports IPv6.
1290 .. clicmd:: redistribute <babel|connected|eigrp|isis|kernel|openfabric|ospf|ospf6|rip|ripng|sharp|static|table> [metric (0-4294967295)] [route-map WORD]
1292 Redistribute routes from other protocols into BGP.
1294 .. clicmd:: redistribute vnc-direct
1296 Redistribute VNC direct (not via zebra) routes to BGP process.
1298 .. clicmd:: bgp update-delay MAX-DELAY
1300 .. clicmd:: bgp update-delay MAX-DELAY ESTABLISH-WAIT
1302 This feature is used to enable read-only mode on BGP process restart or when
1303 a BGP process is cleared using 'clear ip bgp \*'. Note that this command is
1304 configured at the global level and applies to all bgp instances/vrfs. It
1305 cannot be used at the same time as the "update-delay" command described below,
1306 which is entered in each bgp instance/vrf desired to delay update installation
1307 and advertisements. The global and per-vrf approaches to defining update-delay
1308 are mutually exclusive.
1310 When applicable, read-only mode would begin as soon as the first peer reaches
1311 Established status and a timer for max-delay seconds is started. During this
1312 mode BGP doesn't run any best-path or generate any updates to its peers. This
1313 mode continues until:
1315 1. All the configured peers, except the shutdown peers, have sent explicit EOR
1316 (End-Of-RIB) or an implicit-EOR. The first keep-alive after BGP has reached
1317 Established is considered an implicit-EOR.
1318 If the establish-wait optional value is given, then BGP will wait for
1319 peers to reach established from the beginning of the update-delay till the
1320 establish-wait period is over, i.e. the minimum set of established peers for
1321 which EOR is expected would be peers established during the establish-wait
1322 window, not necessarily all the configured neighbors.
1323 2. max-delay period is over.
1325 On hitting any of the above two conditions, BGP resumes the decision process
1326 and generates updates to its peers.
1328 Default max-delay is 0, i.e. the feature is off by default.
1331 .. clicmd:: update-delay MAX-DELAY
1333 .. clicmd:: update-delay MAX-DELAY ESTABLISH-WAIT
1335 This feature is used to enable read-only mode on BGP process restart or when
1336 a BGP process is cleared using 'clear ip bgp \*'. Note that this command is
1337 configured under the specific bgp instance/vrf that the feature is enabled for.
1338 It cannot be used at the same time as the global "bgp update-delay" described
1339 above, which is entered at the global level and applies to all bgp instances.
1340 The global and per-vrf approaches to defining update-delay are mutually
1343 When applicable, read-only mode would begin as soon as the first peer reaches
1344 Established status and a timer for max-delay seconds is started. During this
1345 mode BGP doesn't run any best-path or generate any updates to its peers. This
1346 mode continues until:
1348 1. All the configured peers, except the shutdown peers, have sent explicit EOR
1349 (End-Of-RIB) or an implicit-EOR. The first keep-alive after BGP has reached
1350 Established is considered an implicit-EOR.
1351 If the establish-wait optional value is given, then BGP will wait for
1352 peers to reach established from the beginning of the update-delay till the
1353 establish-wait period is over, i.e. the minimum set of established peers for
1354 which EOR is expected would be peers established during the establish-wait
1355 window, not necessarily all the configured neighbors.
1356 2. max-delay period is over.
1358 On hitting any of the above two conditions, BGP resumes the decision process
1359 and generates updates to its peers.
1361 Default max-delay is 0, i.e. the feature is off by default.
1363 .. clicmd:: table-map ROUTE-MAP-NAME
1365 This feature is used to apply a route-map on route updates from BGP to
1366 Zebra. All the applicable match operations are allowed, such as match on
1367 prefix, next-hop, communities, etc. Set operations for this attach-point are
1368 limited to metric and next-hop only. Any operation of this feature does not
1369 affect BGPs internal RIB.
1371 Supported for ipv4 and ipv6 address families. It works on multi-paths as
1372 well, however, metric setting is based on the best-path only.
1379 .. _bgp-defining-peers:
1384 .. clicmd:: neighbor PEER remote-as ASN
1386 Creates a new neighbor whose remote-as is ASN. PEER can be an IPv4 address
1387 or an IPv6 address or an interface to use for the connection.
1392 neighbor 10.0.0.1 remote-as 2
1394 In this case my router, in AS-1, is trying to peer with AS-2 at 10.0.0.1.
1396 This command must be the first command used when configuring a neighbor. If
1397 the remote-as is not specified, *bgpd* will complain like this: ::
1399 can't find neighbor 10.0.0.1
1401 .. clicmd:: neighbor PEER remote-as internal
1403 Create a peer as you would when you specify an ASN, except that if the
1404 peers ASN is different than mine as specified under the :clicmd:`router bgp ASN`
1405 command the connection will be denied.
1407 .. clicmd:: neighbor PEER remote-as external
1409 Create a peer as you would when you specify an ASN, except that if the
1410 peers ASN is the same as mine as specified under the :clicmd:`router bgp ASN`
1411 command the connection will be denied.
1413 .. clicmd:: bgp listen range <A.B.C.D/M|X:X::X:X/M> peer-group PGNAME
1415 Accept connections from any peers in the specified prefix. Configuration
1416 from the specified peer-group is used to configure these peers.
1420 When using BGP listen ranges, if the associated peer group has TCP MD5
1421 authentication configured, your kernel must support this on prefixes. On
1422 Linux, this support was added in kernel version 4.14. If your kernel does
1423 not support this feature you will get a warning in the log file, and the
1424 listen range will only accept connections from peers without MD5 configured.
1426 Additionally, we have observed that when using this option at scale (several
1427 hundred peers) the kernel may hit its option memory limit. In this situation
1428 you will see error messages like:
1430 ``bgpd: sockopt_tcp_signature: setsockopt(23): Cannot allocate memory``
1432 In this case you need to increase the value of the sysctl
1433 ``net.core.optmem_max`` to allow the kernel to allocate the necessary option
1436 .. clicmd:: bgp listen limit <1-65535>
1438 Define the maximum number of peers accepted for one BGP instance. This
1439 limit is set to 100 by default. Increasing this value will really be
1440 possible if more file descriptors are available in the BGP process. This
1441 value is defined by the underlying system (ulimit value), and can be
1442 overridden by `--limit-fds`. More information is available in chapter
1443 (:ref:`common-invocation-options`).
1445 .. clicmd:: coalesce-time (0-4294967295)
1447 The time in milliseconds that BGP will delay before deciding what peers
1448 can be put into an update-group together in order to generate a single
1449 update for them. The default time is 1000.
1451 .. _bgp-configuring-peers:
1456 .. clicmd:: neighbor PEER shutdown [message MSG...] [rtt (1-65535) [count (1-255)]]
1458 Shutdown the peer. We can delete the neighbor's configuration by
1459 ``no neighbor PEER remote-as ASN`` but all configuration of the neighbor
1460 will be deleted. When you want to preserve the configuration, but want to
1461 drop the BGP peer, use this syntax.
1463 Optionally you can specify a shutdown message `MSG`.
1465 Also, you can specify optionally ``rtt`` in milliseconds to automatically
1466 shutdown the peer if round-trip-time becomes higher than defined.
1468 Additional ``count`` parameter is the number of keepalive messages to count
1469 before shutdown the peer if round-trip-time becomes higher than defined.
1471 .. clicmd:: neighbor PEER disable-connected-check
1473 Allow peerings between directly connected eBGP peers using loopback
1476 .. clicmd:: neighbor PEER disable-link-bw-encoding-ieee
1478 By default bandwidth in extended communities is carried encoded as IEEE
1479 floating-point format, which is according to the draft.
1481 Older versions have the implementation where extended community bandwidth
1482 value is carried encoded as uint32. To enable backward compatibility we
1483 need to disable IEEE floating-point encoding option per-peer.
1485 .. clicmd:: neighbor PEER extended-optional-parameters
1487 Force Extended Optional Parameters Length format to be used for OPEN messages.
1489 By default, it's disabled. If the standard optional parameters length is
1490 higher than one-octet (255), then extended format is enabled automatically.
1492 For testing purposes, extended format can be enabled with this command.
1494 .. clicmd:: neighbor PEER ebgp-multihop
1496 Specifying ``ebgp-multihop`` allows sessions with eBGP neighbors to
1497 establish when they are multiple hops away. When the neighbor is not
1498 directly connected and this knob is not enabled, the session will not
1501 If the peer's IP address is not in the RIB and is reachable via the
1502 default route, then you have to enable ``ip nht resolve-via-default``.
1504 .. clicmd:: neighbor PEER description ...
1506 Set description of the peer.
1508 .. clicmd:: neighbor PEER interface IFNAME
1510 When you connect to a BGP peer over an IPv6 link-local address, you have to
1511 specify the IFNAME of the interface used for the connection. To specify
1512 IPv4 session addresses, see the ``neighbor PEER update-source`` command
1515 .. clicmd:: neighbor PEER interface remote-as <internal|external|ASN>
1517 Configure an unnumbered BGP peer. ``PEER`` should be an interface name. The
1518 session will be established via IPv6 link locals. Use ``internal`` for iBGP
1519 and ``external`` for eBGP sessions, or specify an ASN if you wish.
1521 .. clicmd:: neighbor PEER next-hop-self [force]
1523 This command specifies an announced route's nexthop as being equivalent to
1524 the address of the bgp router if it is learned via eBGP. This will also
1525 bypass third-party next-hops in favor of the local bgp address. If the
1526 optional keyword ``force`` is specified the modification is done also for
1527 routes learned via iBGP.
1529 .. clicmd:: neighbor PEER attribute-unchanged [{as-path|next-hop|med}]
1531 This command specifies attributes to be left unchanged for advertisements
1532 sent to a peer. Use this to leave the next-hop unchanged in ipv6
1533 configurations, as the route-map directive to leave the next-hop unchanged
1534 is only available for ipv4.
1536 .. clicmd:: neighbor PEER update-source <IFNAME|ADDRESS>
1538 Specify the IPv4 source address to use for the :abbr:`BGP` session to this
1539 neighbour, may be specified as either an IPv4 address directly or as an
1540 interface name (in which case the *zebra* daemon MUST be running in order
1541 for *bgpd* to be able to retrieve interface state).
1546 neighbor foo update-source 192.168.0.1
1547 neighbor bar update-source lo0
1550 .. clicmd:: neighbor PEER default-originate [route-map WORD]
1552 *bgpd*'s default is to not announce the default route (0.0.0.0/0) even if it
1553 is in routing table. When you want to announce default routes to the peer,
1556 If ``route-map`` keyword is specified, then the default route will be
1557 originated only if route-map conditions are met. For example, announce
1558 the default route only if ``10.10.10.10/32`` route exists and set an
1559 arbitrary community for a default route.
1564 address-family ipv4 unicast
1565 neighbor 192.168.255.1 default-originate route-map default
1567 ip prefix-list p1 seq 5 permit 10.10.10.10/32
1569 route-map default permit 10
1570 match ip address prefix-list p1
1571 set community 123:123
1574 .. clicmd:: neighbor PEER port PORT
1576 .. clicmd:: neighbor PEER password PASSWORD
1578 Set a MD5 password to be used with the tcp socket that is being used
1579 to connect to the remote peer. Please note if you are using this
1580 command with a large number of peers on linux you should consider
1581 modifying the `net.core.optmem_max` sysctl to a larger value to
1582 avoid out of memory errors from the linux kernel.
1584 .. clicmd:: neighbor PEER send-community
1586 .. clicmd:: neighbor PEER weight WEIGHT
1588 This command specifies a default `weight` value for the neighbor's routes.
1590 .. clicmd:: neighbor PEER maximum-prefix NUMBER [force]
1592 Sets a maximum number of prefixes we can receive from a given peer. If this
1593 number is exceeded, the BGP session will be destroyed.
1595 In practice, it is generally preferable to use a prefix-list to limit what
1596 prefixes are received from the peer instead of using this knob. Tearing down
1597 the BGP session when a limit is exceeded is far more destructive than merely
1598 rejecting undesired prefixes. The prefix-list method is also much more
1599 granular and offers much smarter matching criterion than number of received
1600 prefixes, making it more suited to implementing policy.
1602 If ``force`` is set, then ALL prefixes are counted for maximum instead of
1603 accepted only. This is useful for cases where an inbound filter is applied,
1604 but you want maximum-prefix to act on ALL (including filtered) prefixes. This
1605 option requires `soft-reconfiguration inbound` to be enabled for the peer.
1607 .. clicmd:: neighbor PEER maximum-prefix-out NUMBER
1609 Sets a maximum number of prefixes we can send to a given peer.
1611 Since sent prefix count is managed by update-groups, this option
1612 creates a separate update-group for outgoing updates.
1614 .. clicmd:: neighbor PEER local-as AS-NUMBER [no-prepend] [replace-as]
1616 Specify an alternate AS for this BGP process when interacting with the
1617 specified peer. With no modifiers, the specified local-as is prepended to
1618 the received AS_PATH when receiving routing updates from the peer, and
1619 prepended to the outgoing AS_PATH (after the process local AS) when
1620 transmitting local routes to the peer.
1622 If the no-prepend attribute is specified, then the supplied local-as is not
1623 prepended to the received AS_PATH.
1625 If the replace-as attribute is specified, then only the supplied local-as is
1626 prepended to the AS_PATH when transmitting local-route updates to this peer.
1628 Note that replace-as can only be specified if no-prepend is.
1630 This command is only allowed for eBGP peers.
1632 .. clicmd:: neighbor <A.B.C.D|X:X::X:X|WORD> as-override
1634 Override AS number of the originating router with the local AS number.
1636 Usually this configuration is used in PEs (Provider Edge) to replace
1637 the incoming customer AS number so the connected CE (Customer Edge)
1638 can use the same AS number as the other customer sites. This allows
1639 customers of the provider network to use the same AS number across
1642 This command is only allowed for eBGP peers.
1644 .. clicmd:: neighbor <A.B.C.D|X:X::X:X|WORD> allowas-in [<(1-10)|origin>]
1646 Accept incoming routes with AS path containing AS number with the same value
1647 as the current system AS.
1649 This is used when you want to use the same AS number in your sites, but you
1650 can't connect them directly. This is an alternative to
1651 `neighbor WORD as-override`.
1653 The parameter `(1-10)` configures the amount of accepted occurrences of the
1654 system AS number in AS path.
1656 The parameter `origin` configures BGP to only accept routes originated with
1657 the same AS number as the system.
1659 This command is only allowed for eBGP peers.
1661 .. clicmd:: neighbor <A.B.C.D|X:X::X:X|WORD> addpath-tx-all-paths
1663 Configure BGP to send all known paths to neighbor in order to preserve multi
1664 path capabilities inside a network.
1666 .. clicmd:: neighbor <A.B.C.D|X:X::X:X|WORD> addpath-tx-bestpath-per-AS
1668 Configure BGP to send best known paths to neighbor in order to preserve multi
1669 path capabilities inside a network.
1671 .. clicmd:: neighbor <A.B.C.D|X:X::X:X|WORD> disable-addpath-rx
1673 Do not accept additional paths from this neighbor.
1675 .. clicmd:: neighbor PEER ttl-security hops NUMBER
1677 This command enforces Generalized TTL Security Mechanism (GTSM), as
1678 specified in RFC 5082. With this command, only neighbors that are the
1679 specified number of hops away will be allowed to become neighbors. This
1680 command is mutually exclusive with *ebgp-multihop*.
1682 .. clicmd:: neighbor PEER capability extended-nexthop
1684 Allow bgp to negotiate the extended-nexthop capability with it's peer.
1685 If you are peering over a v6 LL address then this capability is turned
1686 on automatically. If you are peering over a v6 Global Address then
1687 turning on this command will allow BGP to install v4 routes with
1688 v6 nexthops if you do not have v4 configured on interfaces.
1690 .. clicmd:: neighbor <A.B.C.D|X:X::X:X|WORD> accept-own
1692 Enable handling of self-originated VPN routes containing ``accept-own`` community.
1694 This feature allows you to handle self-originated VPN routes, which a BGP speaker
1695 receives from a route-reflector. A 'self-originated' route is one that was
1696 originally advertised by the speaker itself. As per :rfc:`4271`, a BGP speaker rejects
1697 advertisements that originated the speaker itself. However, the BGP ACCEPT_OWN
1698 mechanism enables a router to accept the prefixes it has advertised, when reflected
1699 from a route-reflector that modifies certain attributes of the prefix.
1701 A special community called ``accept-own`` is attached to the prefix by the
1702 route-reflector, which is a signal to the receiving router to bypass the ORIGINATOR_ID
1703 and NEXTHOP/MP_REACH_NLRI check.
1707 .. clicmd:: neighbor <A.B.C.D|X:X::X:X|WORD> path-attribute discard (1-255)...
1709 Drops specified path attributes from BGP UPDATE messages from the specified neighbor.
1711 If you do not want specific attributes, you can drop them using this command, and
1712 let the BGP proceed by ignoring those attributes.
1714 .. clicmd:: neighbor <A.B.C.D|X:X::X:X|WORD> graceful-shutdown
1716 Mark all routes from this neighbor as less preferred by setting ``graceful-shutdown``
1717 community, and local-preference to 0.
1719 .. clicmd:: bgp fast-external-failover
1721 This command causes bgp to take down ebgp peers immediately
1722 when a link flaps. `bgp fast-external-failover` is the default
1723 and will not be displayed as part of a `show run`. The no form
1724 of the command turns off this ability.
1726 .. clicmd:: bgp default ipv4-unicast
1728 This command allows the user to specify that the IPv4 Unicast address
1729 family is turned on by default or not. This command defaults to on
1730 and is not displayed.
1731 The `no bgp default ipv4-unicast` form of the command is displayed.
1733 .. clicmd:: bgp default ipv4-multicast
1735 This command allows the user to specify that the IPv4 Multicast address
1736 family is turned on by default or not. This command defaults to off
1737 and is not displayed.
1738 The `bgp default ipv4-multicast` form of the command is displayed.
1740 .. clicmd:: bgp default ipv4-vpn
1742 This command allows the user to specify that the IPv4 MPLS VPN address
1743 family is turned on by default or not. This command defaults to off
1744 and is not displayed.
1745 The `bgp default ipv4-vpn` form of the command is displayed.
1747 .. clicmd:: bgp default ipv4-flowspec
1749 This command allows the user to specify that the IPv4 Flowspec address
1750 family is turned on by default or not. This command defaults to off
1751 and is not displayed.
1752 The `bgp default ipv4-flowspec` form of the command is displayed.
1754 .. clicmd:: bgp default ipv6-unicast
1756 This command allows the user to specify that the IPv6 Unicast address
1757 family is turned on by default or not. This command defaults to off
1758 and is not displayed.
1759 The `bgp default ipv6-unicast` form of the command is displayed.
1761 .. clicmd:: bgp default ipv6-multicast
1763 This command allows the user to specify that the IPv6 Multicast address
1764 family is turned on by default or not. This command defaults to off
1765 and is not displayed.
1766 The `bgp default ipv6-multicast` form of the command is displayed.
1768 .. clicmd:: bgp default ipv6-vpn
1770 This command allows the user to specify that the IPv6 MPLS VPN address
1771 family is turned on by default or not. This command defaults to off
1772 and is not displayed.
1773 The `bgp default ipv6-vpn` form of the command is displayed.
1775 .. clicmd:: bgp default ipv6-flowspec
1777 This command allows the user to specify that the IPv6 Flowspec address
1778 family is turned on by default or not. This command defaults to off
1779 and is not displayed.
1780 The `bgp default ipv6-flowspec` form of the command is displayed.
1782 .. clicmd:: bgp default l2vpn-evpn
1784 This command allows the user to specify that the L2VPN EVPN address
1785 family is turned on by default or not. This command defaults to off
1786 and is not displayed.
1787 The `bgp default l2vpn-evpn` form of the command is displayed.
1789 .. clicmd:: bgp default show-hostname
1791 This command shows the hostname of the peer in certain BGP commands
1792 outputs. It's easier to troubleshoot if you have a number of BGP peers.
1794 .. clicmd:: bgp default show-nexthop-hostname
1796 This command shows the hostname of the next-hop in certain BGP commands
1797 outputs. It's easier to troubleshoot if you have a number of BGP peers
1798 and a number of routes to check.
1800 .. clicmd:: neighbor PEER advertisement-interval (0-600)
1802 Setup the minimum route advertisement interval(mrai) for the
1803 peer in question. This number is between 0 and 600 seconds,
1804 with the default advertisement interval being 0.
1806 .. clicmd:: neighbor PEER timers (0-65535) (0-65535)
1808 Set keepalive and hold timers for a neighbor. The first value is keepalive
1809 and the second is hold time.
1811 .. clicmd:: neighbor PEER timers connect (1-65535)
1813 Set connect timer for a neighbor. The connect timer controls how long BGP
1814 waits between connection attempts to a neighbor.
1816 .. clicmd:: neighbor PEER timers delayopen (1-240)
1818 This command allows the user enable the
1819 `RFC 4271 <https://tools.ietf.org/html/rfc4271/>` DelayOpenTimer with the
1820 specified interval or disable it with the negating command for the peer. By
1821 default, the DelayOpenTimer is disabled. The timer interval may be set to a
1822 duration of 1 to 240 seconds.
1824 .. clicmd:: bgp minimum-holdtime (1-65535)
1826 This command allows user to prevent session establishment with BGP peers
1827 with lower holdtime less than configured minimum holdtime.
1828 When this command is not set, minimum holdtime does not work.
1830 .. clicmd:: bgp tcp-keepalive (1-65535) (1-65535) (1-30)
1832 This command allows user to configure TCP keepalive with new BGP peers.
1833 Each parameter respectively stands for TCP keepalive idle timer (seconds),
1834 interval (seconds), and maximum probes. By default, TCP keepalive is
1837 Displaying Information about Peers
1838 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1840 .. clicmd:: show bgp <afi> <safi> neighbors WORD bestpath-routes [detail] [json] [wide]
1842 For the given neighbor, WORD, that is specified list the routes selected
1843 by BGP as having the best path.
1845 If ``detail`` option is specified, the detailed version of all routes
1846 will be displayed. The same format as ``show [ip] bgp [afi] [safi] PREFIX``
1847 will be used, but for the whole table of received, advertised or filtered
1850 If ``json`` option is specified, output is displayed in JSON format.
1852 If ``wide`` option is specified, then the prefix table's width is increased
1853 to fully display the prefix and the nexthop.
1855 .. _bgp-peer-filtering:
1860 .. clicmd:: neighbor PEER distribute-list NAME [in|out]
1862 This command specifies a distribute-list for the peer. `direct` is
1865 .. clicmd:: neighbor PEER prefix-list NAME [in|out]
1867 .. clicmd:: neighbor PEER filter-list NAME [in|out]
1869 .. clicmd:: neighbor PEER route-map NAME [in|out]
1871 Apply a route-map on the neighbor. `direct` must be `in` or `out`.
1873 .. clicmd:: bgp route-reflector allow-outbound-policy
1875 By default, attribute modification via route-map policy out is not reflected
1876 on reflected routes. This option allows the modifications to be reflected as
1877 well. Once enabled, it affects all reflected routes.
1879 .. clicmd:: neighbor PEER sender-as-path-loop-detection
1881 Enable the detection of sender side AS path loops and filter the
1882 bad routes before they are sent.
1884 This setting is disabled by default.
1891 Peer groups are used to help improve scaling by generating the same
1892 update information to all members of a peer group. Note that this means
1893 that the routes generated by a member of a peer group will be sent back
1894 to that originating peer with the originator identifier attribute set to
1895 indicated the originating peer. All peers not associated with a
1896 specific peer group are treated as belonging to a default peer group,
1897 and will share updates.
1899 .. clicmd:: neighbor WORD peer-group
1901 This command defines a new peer group.
1903 .. clicmd:: neighbor PEER peer-group PGNAME
1905 This command bind specific peer to peer group WORD.
1907 .. clicmd:: neighbor PEER solo
1909 This command is used to indicate that routes advertised by the peer
1910 should not be reflected back to the peer. This command only is only
1911 meaningful when there is a single peer defined in the peer-group.
1913 .. clicmd:: show [ip] bgp peer-group [json]
1915 This command displays configured BGP peer-groups.
1919 exit1-debian-9# show bgp peer-group
1921 BGP peer-group test1, remote AS 65001
1922 Peer-group type is external
1923 Configured address-families: IPv4 Unicast; IPv6 Unicast;
1924 1 IPv4 listen range(s)
1926 2 IPv6 listen range(s)
1930 192.168.200.1 Active
1933 BGP peer-group test2
1934 Peer-group type is external
1935 Configured address-families: IPv4 Unicast;
1937 Optional ``json`` parameter is used to display JSON output.
1945 "addressFamiliesConfigured":[
1975 "addressFamiliesConfigured":[
1981 Capability Negotiation
1982 ^^^^^^^^^^^^^^^^^^^^^^
1984 .. clicmd:: neighbor PEER strict-capability-match
1987 Strictly compares remote capabilities and local capabilities. If
1988 capabilities are different, send Unsupported Capability error then reset
1991 You may want to disable sending Capability Negotiation OPEN message optional
1992 parameter to the peer when remote peer does not implement Capability
1993 Negotiation. Please use *dont-capability-negotiate* command to disable the
1996 .. clicmd:: neighbor PEER dont-capability-negotiate
1998 Suppress sending Capability Negotiation as OPEN message optional parameter
1999 to the peer. This command only affects the peer is configured other than
2000 IPv4 unicast configuration.
2002 When remote peer does not have capability negotiation feature, remote peer
2003 will not send any capabilities at all. In that case, bgp configures the peer
2004 with configured capabilities.
2006 You may prefer locally configured capabilities more than the negotiated
2007 capabilities even though remote peer sends capabilities. If the peer is
2008 configured by *override-capability*, *bgpd* ignores received capabilities
2009 then override negotiated capabilities with configured values.
2011 Additionally the operator should be reminded that this feature fundamentally
2012 disables the ability to use widely deployed BGP features. BGP unnumbered,
2013 hostname support, AS4, Addpath, Route Refresh, ORF, Dynamic Capabilities,
2014 and graceful restart.
2016 .. clicmd:: neighbor PEER override-capability
2019 Override the result of Capability Negotiation with local configuration.
2020 Ignore remote peer's capability value.
2022 .. _bgp-as-path-access-lists:
2024 AS Path Access Lists
2025 --------------------
2027 AS path access list is user defined AS path.
2029 .. clicmd:: bgp as-path access-list WORD [seq (0-4294967295)] permit|deny LINE
2031 This command defines a new AS path access list.
2033 .. clicmd:: show bgp as-path-access-list [json]
2035 Display all BGP AS Path access lists.
2037 If the ``json`` option is specified, output is displayed in JSON format.
2039 .. clicmd:: show bgp as-path-access-list WORD [json]
2041 Display the specified BGP AS Path access list.
2043 If the ``json`` option is specified, output is displayed in JSON format.
2045 .. _bgp-bogon-filter-example:
2047 Bogon ASN filter policy configuration example
2048 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2052 bgp as-path access-list 99 permit _0_
2053 bgp as-path access-list 99 permit _23456_
2054 bgp as-path access-list 99 permit _1310[0-6][0-9]_|_13107[0-1]_
2055 bgp as-path access-list 99 seq 20 permit ^65
2057 .. _bgp-using-as-path-in-route-map:
2059 Using AS Path in Route Map
2060 --------------------------
2062 .. clicmd:: match as-path WORD
2064 For a given as-path, WORD, match it on the BGP as-path given for the prefix
2065 and if it matches do normal route-map actions. The no form of the command
2066 removes this match from the route-map.
2068 .. clicmd:: set as-path prepend AS-PATH
2070 Prepend the given string of AS numbers to the AS_PATH of the BGP path's NLRI.
2071 The no form of this command removes this set operation from the route-map.
2073 .. clicmd:: set as-path prepend last-as NUM
2075 Prepend the existing last AS number (the leftmost ASN) to the AS_PATH.
2076 The no form of this command removes this set operation from the route-map.
2078 .. clicmd:: set as-path replace <any|ASN>
2080 Replace a specific AS number to local AS number. ``any`` replaces each
2081 AS number in the AS-PATH with the local AS number.
2083 .. _bgp-communities-attribute:
2085 Communities Attribute
2086 ---------------------
2088 The BGP communities attribute is widely used for implementing policy routing.
2089 Network operators can manipulate BGP communities attribute based on their
2090 network policy. BGP communities attribute is defined in :rfc:`1997` and
2091 :rfc:`1998`. It is an optional transitive attribute, therefore local policy can
2092 travel through different autonomous system.
2094 The communities attribute is a set of communities values. Each community value
2095 is 4 octet long. The following format is used to define the community value.
2098 This format represents 4 octet communities value. ``AS`` is high order 2
2099 octet in digit format. ``VAL`` is low order 2 octet in digit format. This
2100 format is useful to define AS oriented policy value. For example,
2101 ``7675:80`` can be used when AS 7675 wants to pass local policy value 80 to
2105 ``internet`` represents well-known communities value 0.
2107 ``graceful-shutdown``
2108 ``graceful-shutdown`` represents well-known communities value
2109 ``GRACEFUL_SHUTDOWN`` ``0xFFFF0000`` ``65535:0``. :rfc:`8326` implements
2110 the purpose Graceful BGP Session Shutdown to reduce the amount of
2111 lost traffic when taking BGP sessions down for maintenance. The use
2112 of the community needs to be supported from your peers side to
2113 actually have any effect.
2116 ``accept-own`` represents well-known communities value ``ACCEPT_OWN``
2117 ``0xFFFF0001`` ``65535:1``. :rfc:`7611` implements a way to signal
2118 to a router to accept routes with a local nexthop address. This
2119 can be the case when doing policing and having traffic having a
2120 nexthop located in another VRF but still local interface to the
2121 router. It is recommended to read the RFC for full details.
2123 ``route-filter-translated-v4``
2124 ``route-filter-translated-v4`` represents well-known communities value
2125 ``ROUTE_FILTER_TRANSLATED_v4`` ``0xFFFF0002`` ``65535:2``.
2128 ``route-filter-v4`` represents well-known communities value
2129 ``ROUTE_FILTER_v4`` ``0xFFFF0003`` ``65535:3``.
2131 ``route-filter-translated-v6``
2132 ``route-filter-translated-v6`` represents well-known communities value
2133 ``ROUTE_FILTER_TRANSLATED_v6`` ``0xFFFF0004`` ``65535:4``.
2136 ``route-filter-v6`` represents well-known communities value
2137 ``ROUTE_FILTER_v6`` ``0xFFFF0005`` ``65535:5``.
2140 ``llgr-stale`` represents well-known communities value ``LLGR_STALE``
2141 ``0xFFFF0006`` ``65535:6``.
2142 Assigned and intended only for use with routers supporting the
2143 Long-lived Graceful Restart Capability as described in
2144 [Draft-IETF-uttaro-idr-bgp-persistence]_.
2145 Routers receiving routes with this community may (depending on
2146 implementation) choose allow to reject or modify routes on the
2147 presence or absence of this community.
2150 ``no-llgr`` represents well-known communities value ``NO_LLGR``
2151 ``0xFFFF0007`` ``65535:7``.
2152 Assigned and intended only for use with routers supporting the
2153 Long-lived Graceful Restart Capability as described in
2154 [Draft-IETF-uttaro-idr-bgp-persistence]_.
2155 Routers receiving routes with this community may (depending on
2156 implementation) choose allow to reject or modify routes on the
2157 presence or absence of this community.
2159 ``accept-own-nexthop``
2160 ``accept-own-nexthop`` represents well-known communities value
2161 ``accept-own-nexthop`` ``0xFFFF0008`` ``65535:8``.
2162 [Draft-IETF-agrewal-idr-accept-own-nexthop]_ describes
2163 how to tag and label VPN routes to be able to send traffic between VRFs
2164 via an internal layer 2 domain on the same PE device. Refer to
2165 [Draft-IETF-agrewal-idr-accept-own-nexthop]_ for full details.
2168 ``blackhole`` represents well-known communities value ``BLACKHOLE``
2169 ``0xFFFF029A`` ``65535:666``. :rfc:`7999` documents sending prefixes to
2170 EBGP peers and upstream for the purpose of blackholing traffic.
2171 Prefixes tagged with the this community should normally not be
2172 re-advertised from neighbors of the originating network. Upon receiving
2173 ``BLACKHOLE`` community from a BGP speaker, ``NO_ADVERTISE`` community
2174 is added automatically.
2177 ``no-export`` represents well-known communities value ``NO_EXPORT``
2178 ``0xFFFFFF01``. All routes carry this value must not be advertised to
2179 outside a BGP confederation boundary. If neighboring BGP peer is part of BGP
2180 confederation, the peer is considered as inside a BGP confederation
2181 boundary, so the route will be announced to the peer.
2184 ``no-advertise`` represents well-known communities value ``NO_ADVERTISE``
2185 ``0xFFFFFF02``. All routes carry this value must not be advertise to other
2189 ``local-AS`` represents well-known communities value ``NO_EXPORT_SUBCONFED``
2190 ``0xFFFFFF03``. All routes carry this value must not be advertised to
2191 external BGP peers. Even if the neighboring router is part of confederation,
2192 it is considered as external BGP peer, so the route will not be announced to
2196 ``no-peer`` represents well-known communities value ``NOPEER``
2197 ``0xFFFFFF04`` ``65535:65284``. :rfc:`3765` is used to communicate to
2198 another network how the originating network want the prefix propagated.
2200 When the communities attribute is received duplicate community values in the
2201 attribute are ignored and value is sorted in numerical order.
2203 .. [Draft-IETF-uttaro-idr-bgp-persistence] <https://tools.ietf.org/id/draft-uttaro-idr-bgp-persistence-04.txt>
2204 .. [Draft-IETF-agrewal-idr-accept-own-nexthop] <https://tools.ietf.org/id/draft-agrewal-idr-accept-own-nexthop-00.txt>
2206 .. _bgp-community-lists:
2210 Community lists are user defined lists of community attribute values. These
2211 lists can be used for matching or manipulating the communities attribute in
2214 There are two types of community list:
2217 This type accepts an explicit value for the attribute.
2220 This type accepts a regular expression. Because the regex must be
2221 interpreted on each use expanded community lists are slower than standard
2224 .. clicmd:: bgp community-list standard NAME permit|deny COMMUNITY
2226 This command defines a new standard community list. ``COMMUNITY`` is
2227 communities value. The ``COMMUNITY`` is compiled into community structure.
2228 We can define multiple community list under same name. In that case match
2229 will happen user defined order. Once the community list matches to
2230 communities attribute in BGP updates it return permit or deny by the
2231 community list definition. When there is no matched entry, deny will be
2232 returned. When ``COMMUNITY`` is empty it matches to any routes.
2234 .. clicmd:: bgp community-list expanded NAME permit|deny COMMUNITY
2236 This command defines a new expanded community list. ``COMMUNITY`` is a
2237 string expression of communities attribute. ``COMMUNITY`` can be a regular
2238 expression (:ref:`bgp-regular-expressions`) to match the communities
2239 attribute in BGP updates. The expanded community is only used to filter,
2243 It is recommended to use the more explicit versions of this command.
2245 .. clicmd:: bgp community-list NAME permit|deny COMMUNITY
2247 When the community list type is not specified, the community list type is
2248 automatically detected. If ``COMMUNITY`` can be compiled into communities
2249 attribute, the community list is defined as a standard community list.
2250 Otherwise it is defined as an expanded community list. This feature is left
2251 for backward compatibility. Use of this feature is not recommended.
2253 Note that all community lists share the same namespace, so it's not
2254 necessary to specify ``standard`` or ``expanded``; these modifiers are
2257 .. clicmd:: show bgp community-list [NAME detail]
2259 Displays community list information. When ``NAME`` is specified the
2260 specified community list's information is shown.
2264 # show bgp community-list
2265 Named Community standard list CLIST
2266 permit 7675:80 7675:100 no-export
2268 Named Community expanded list EXPAND
2271 # show bgp community-list CLIST detail
2272 Named Community standard list CLIST
2273 permit 7675:80 7675:100 no-export
2277 .. _bgp-numbered-community-lists:
2279 Numbered Community Lists
2280 ^^^^^^^^^^^^^^^^^^^^^^^^
2282 When number is used for BGP community list name, the number has
2283 special meanings. Community list number in the range from 1 and 99 is
2284 standard community list. Community list number in the range from 100
2285 to 500 is expanded community list. These community lists are called
2286 as numbered community lists. On the other hand normal community lists
2287 is called as named community lists.
2289 .. clicmd:: bgp community-list (1-99) permit|deny COMMUNITY
2291 This command defines a new community list. The argument to (1-99) defines
2292 the list identifier.
2294 .. clicmd:: bgp community-list (100-500) permit|deny COMMUNITY
2296 This command defines a new expanded community list. The argument to
2297 (100-500) defines the list identifier.
2299 .. _bgp-community-alias:
2304 BGP community aliases are useful to quickly identify what communities are set
2305 for a specific prefix in a human-readable format. Especially handy for a huge
2306 amount of communities. Accurately defined aliases can help you faster spot
2309 .. clicmd:: bgp community alias NAME ALIAS
2311 This command creates an alias name for a community that will be used
2312 later in various CLI outputs in a human-readable format.
2316 ~# vtysh -c 'show run' | grep 'bgp community alias'
2317 bgp community alias 65001:14 community-1
2318 bgp community alias 65001:123:1 lcommunity-1
2320 ~# vtysh -c 'show ip bgp 172.16.16.1/32'
2321 BGP routing table entry for 172.16.16.1/32, version 21
2322 Paths: (2 available, best #2, table default)
2323 Advertised to non peer-group peers:
2325 192.168.0.2 from 192.168.0.2 (172.16.16.1)
2326 Origin incomplete, metric 0, valid, external, best (Neighbor IP)
2327 Community: 65001:12 65001:13 community-1 65001:65534
2328 Large Community: lcommunity-1 65001:123:2
2329 Last update: Fri Apr 16 12:51:27 2021
2331 .. clicmd:: show bgp [afi] [safi] [all] alias WORD [wide|json]
2333 Display prefixes with matching BGP community alias.
2335 .. _bgp-using-communities-in-route-map:
2337 Using Communities in Route Maps
2338 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2340 In :ref:`route-map` we can match on or set the BGP communities attribute. Using
2341 this feature network operator can implement their network policy based on BGP
2342 communities attribute.
2344 The following commands can be used in route maps:
2346 .. clicmd:: match alias WORD
2348 This command performs match to BGP updates using community alias WORD. When
2349 the one of BGP communities value match to the one of community alias value in
2350 community alias, it is match.
2352 .. clicmd:: match community WORD exact-match [exact-match]
2354 This command perform match to BGP updates using community list WORD. When
2355 the one of BGP communities value match to the one of communities value in
2356 community list, it is match. When `exact-match` keyword is specified, match
2357 happen only when BGP updates have completely same communities value
2358 specified in the community list.
2360 .. clicmd:: set community <none|COMMUNITY> additive
2362 This command sets the community value in BGP updates. If the attribute is
2363 already configured, the newly provided value replaces the old one unless the
2364 ``additive`` keyword is specified, in which case the new value is appended
2365 to the existing value.
2367 If ``none`` is specified as the community value, the communities attribute
2370 It is not possible to set an expanded community list.
2372 .. clicmd:: set comm-list WORD delete
2374 This command remove communities value from BGP communities attribute. The
2375 ``word`` is community list name. When BGP route's communities value matches
2376 to the community list ``word``, the communities value is removed. When all
2377 of communities value is removed eventually, the BGP update's communities
2378 attribute is completely removed.
2380 .. _bgp-communities-example:
2382 Example Configuration
2383 ^^^^^^^^^^^^^^^^^^^^^
2385 The following configuration is exemplary of the most typical usage of BGP
2386 communities attribute. In the example, AS 7675 provides an upstream Internet
2387 connection to AS 100. When the following configuration exists in AS 7675, the
2388 network operator of AS 100 can set local preference in AS 7675 network by
2389 setting BGP communities attribute to the updates.
2394 neighbor 192.168.0.1 remote-as 100
2395 address-family ipv4 unicast
2396 neighbor 192.168.0.1 route-map RMAP in
2399 bgp community-list 70 permit 7675:70
2400 bgp community-list 80 permit 7675:80
2401 bgp community-list 90 permit 7675:90
2403 route-map RMAP permit 10
2405 set local-preference 70
2407 route-map RMAP permit 20
2409 set local-preference 80
2411 route-map RMAP permit 30
2413 set local-preference 90
2416 The following configuration announces ``10.0.0.0/8`` from AS 100 to AS 7675.
2417 The route has communities value ``7675:80`` so when above configuration exists
2418 in AS 7675, the announced routes' local preference value will be set to 80.
2424 neighbor 192.168.0.2 remote-as 7675
2425 address-family ipv4 unicast
2426 neighbor 192.168.0.2 route-map RMAP out
2429 ip prefix-list PLIST permit 10.0.0.0/8
2431 route-map RMAP permit 10
2432 match ip address prefix-list PLIST
2433 set community 7675:80
2436 The following configuration is an example of BGP route filtering using
2437 communities attribute. This configuration only permit BGP routes which has BGP
2438 communities value (``0:80`` and ``0:90``) or ``0:100``. The network operator can
2439 set special internal communities value at BGP border router, then limit the
2440 BGP route announcements into the internal network.
2445 neighbor 192.168.0.1 remote-as 100
2446 address-family ipv4 unicast
2447 neighbor 192.168.0.1 route-map RMAP in
2450 bgp community-list 1 permit 0:80 0:90
2451 bgp community-list 1 permit 0:100
2453 route-map RMAP permit in
2457 The following example filters BGP routes which have a community value of
2458 ``1:1``. When there is no match community-list returns ``deny``. To avoid
2459 filtering all routes, a ``permit`` line is set at the end of the
2465 neighbor 192.168.0.1 remote-as 100
2466 address-family ipv4 unicast
2467 neighbor 192.168.0.1 route-map RMAP in
2470 bgp community-list standard FILTER deny 1:1
2471 bgp community-list standard FILTER permit
2473 route-map RMAP permit 10
2474 match community FILTER
2477 The communities value keyword ``internet`` has special meanings in standard
2478 community lists. In the below example ``internet`` matches all BGP routes even
2479 if the route does not have communities attribute at all. So community list
2480 ``INTERNET`` is the same as ``FILTER`` in the previous example.
2484 bgp community-list standard INTERNET deny 1:1
2485 bgp community-list standard INTERNET permit internet
2488 The following configuration is an example of communities value deletion. With
2489 this configuration the community values ``100:1`` and ``100:2`` are removed
2490 from BGP updates. For communities value deletion, only ``permit``
2491 community-list is used. ``deny`` community-list is ignored.
2496 neighbor 192.168.0.1 remote-as 100
2497 address-family ipv4 unicast
2498 neighbor 192.168.0.1 route-map RMAP in
2501 bgp community-list standard DEL permit 100:1 100:2
2503 route-map RMAP permit 10
2504 set comm-list DEL delete
2507 .. _bgp-extended-communities-attribute:
2509 Extended Communities Attribute
2510 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2512 BGP extended communities attribute is introduced with MPLS VPN/BGP technology.
2513 MPLS VPN/BGP expands capability of network infrastructure to provide VPN
2514 functionality. At the same time it requires a new framework for policy routing.
2515 With BGP Extended Communities Attribute we can use Route Target or Site of
2516 Origin for implementing network policy for MPLS VPN/BGP.
2518 BGP Extended Communities Attribute is similar to BGP Communities Attribute. It
2519 is an optional transitive attribute. BGP Extended Communities Attribute can
2520 carry multiple Extended Community value. Each Extended Community value is
2523 BGP Extended Communities Attribute provides an extended range compared with BGP
2524 Communities Attribute. Adding to that there is a type field in each value to
2525 provides community space structure.
2527 There are two format to define Extended Community value. One is AS based format
2528 the other is IP address based format.
2531 This is a format to define AS based Extended Community value. ``AS`` part
2532 is 2 octets Global Administrator subfield in Extended Community value.
2533 ``VAL`` part is 4 octets Local Administrator subfield. ``7675:100``
2534 represents AS 7675 policy value 100.
2537 This is a format to define IP address based Extended Community value.
2538 ``IP-Address`` part is 4 octets Global Administrator subfield. ``VAL`` part
2539 is 2 octets Local Administrator subfield.
2541 .. _bgp-extended-community-lists:
2543 Extended Community Lists
2544 ^^^^^^^^^^^^^^^^^^^^^^^^
2546 .. clicmd:: bgp extcommunity-list standard NAME permit|deny EXTCOMMUNITY
2548 This command defines a new standard extcommunity-list. `extcommunity` is
2549 extended communities value. The `extcommunity` is compiled into extended
2550 community structure. We can define multiple extcommunity-list under same
2551 name. In that case match will happen user defined order. Once the
2552 extcommunity-list matches to extended communities attribute in BGP updates
2553 it return permit or deny based upon the extcommunity-list definition. When
2554 there is no matched entry, deny will be returned. When `extcommunity` is
2555 empty it matches to any routes.
2557 A special handling for ``internet`` community is applied. It matches
2560 .. clicmd:: bgp extcommunity-list expanded NAME permit|deny LINE
2562 This command defines a new expanded extcommunity-list. `line` is a string
2563 expression of extended communities attribute. `line` can be a regular
2564 expression (:ref:`bgp-regular-expressions`) to match an extended communities
2565 attribute in BGP updates.
2567 Note that all extended community lists shares a single name space, so it's
2568 not necessary to specify their type when creating or destroying them.
2570 .. clicmd:: show bgp extcommunity-list [NAME detail]
2572 This command displays current extcommunity-list information. When `name` is
2573 specified the community list's information is shown.
2576 .. _bgp-extended-communities-in-route-map:
2578 BGP Extended Communities in Route Map
2579 """""""""""""""""""""""""""""""""""""
2581 .. clicmd:: match extcommunity WORD
2583 .. clicmd:: set extcommunity none
2585 This command resets the extended community value in BGP updates. If the attribute is
2586 already configured or received from the peer, the attribute is discarded and set to
2587 none. This is useful if you need to strip incoming extended communities.
2589 .. clicmd:: set extcommunity rt EXTCOMMUNITY
2591 This command set Route Target value.
2593 .. clicmd:: set extcommunity soo EXTCOMMUNITY
2595 This command set Site of Origin value.
2597 .. clicmd:: set extcommunity bandwidth <(1-25600) | cumulative | num-multipaths> [non-transitive]
2599 This command sets the BGP link-bandwidth extended community for the prefix
2600 (best path) for which it is applied. The link-bandwidth can be specified as
2601 an ``explicit value`` (specified in Mbps), or the router can be told to use
2602 the ``cumulative bandwidth`` of all multipaths for the prefix or to compute
2603 it based on the ``number of multipaths``. The link bandwidth extended
2604 community is encoded as ``transitive`` unless the set command explicitly
2605 configures it as ``non-transitive``.
2607 .. seealso:: :ref:`wecmp_linkbw`
2609 Note that the extended expanded community is only used for `match` rule, not for
2612 .. _bgp-large-communities-attribute:
2614 Large Communities Attribute
2615 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
2617 The BGP Large Communities attribute was introduced in Feb 2017 with
2620 The BGP Large Communities Attribute is similar to the BGP Communities Attribute
2621 except that it has 3 components instead of two and each of which are 4 octets
2622 in length. Large Communities bring additional functionality and convenience
2623 over traditional communities, specifically the fact that the ``GLOBAL`` part
2624 below is now 4 octets wide allowing seamless use in networks using 4-byte ASNs.
2626 ``GLOBAL:LOCAL1:LOCAL2``
2627 This is the format to define Large Community values. Referencing :rfc:`8195`
2628 the values are commonly referred to as follows:
2630 - The ``GLOBAL`` part is a 4 octet Global Administrator field, commonly used
2631 as the operators AS number.
2632 - The ``LOCAL1`` part is a 4 octet Local Data Part 1 subfield referred to as
2634 - The ``LOCAL2`` part is a 4 octet Local Data Part 2 field and referred to
2635 as the parameter subfield.
2637 As an example, ``65551:1:10`` represents AS 65551 function 1 and parameter
2638 10. The referenced RFC above gives some guidelines on recommended usage.
2640 .. _bgp-large-community-lists:
2642 Large Community Lists
2643 """""""""""""""""""""
2645 Two types of large community lists are supported, namely `standard` and
2648 .. clicmd:: bgp large-community-list standard NAME permit|deny LARGE-COMMUNITY
2650 This command defines a new standard large-community-list. `large-community`
2651 is the Large Community value. We can add multiple large communities under
2652 same name. In that case the match will happen in the user defined order.
2653 Once the large-community-list matches the Large Communities attribute in BGP
2654 updates it will return permit or deny based upon the large-community-list
2655 definition. When there is no matched entry, a deny will be returned. When
2656 `large-community` is empty it matches any routes.
2658 .. clicmd:: bgp large-community-list expanded NAME permit|deny LINE
2660 This command defines a new expanded large-community-list. Where `line` is a
2661 string matching expression, it will be compared to the entire Large
2662 Communities attribute as a string, with each large-community in order from
2663 lowest to highest. `line` can also be a regular expression which matches
2664 this Large Community attribute.
2666 Note that all community lists share the same namespace, so it's not
2667 necessary to specify ``standard`` or ``expanded``; these modifiers are
2670 .. clicmd:: show bgp large-community-list
2672 .. clicmd:: show bgp large-community-list NAME detail
2674 This command display current large-community-list information. When
2675 `name` is specified the community list information is shown.
2677 .. clicmd:: show ip bgp large-community-info
2679 This command displays the current large communities in use.
2681 .. _bgp-large-communities-in-route-map:
2683 Large Communities in Route Map
2684 """"""""""""""""""""""""""""""
2686 .. clicmd:: match large-community LINE [exact-match]
2688 Where `line` can be a simple string to match, or a regular expression. It
2689 is very important to note that this match occurs on the entire
2690 large-community string as a whole, where each large-community is ordered
2691 from lowest to highest. When `exact-match` keyword is specified, match
2692 happen only when BGP updates have completely same large communities value
2693 specified in the large community list.
2695 .. clicmd:: set large-community LARGE-COMMUNITY
2697 .. clicmd:: set large-community LARGE-COMMUNITY LARGE-COMMUNITY
2699 .. clicmd:: set large-community LARGE-COMMUNITY additive
2701 These commands are used for setting large-community values. The first
2702 command will overwrite any large-communities currently present.
2703 The second specifies two large-communities, which overwrites the current
2704 large-community list. The third will add a large-community value without
2705 overwriting other values. Multiple large-community values can be specified.
2707 Note that the large expanded community is only used for `match` rule, not for
2710 .. _bgp-roles-and-only-to-customers:
2712 BGP Roles and Only to Customers
2713 -------------------------------
2715 BGP roles are defined in :rfc:`9234` and provide an easy way to route leaks
2716 prevention, detection and mitigation.
2718 To enable its mechanics, you must set your local role to reflect your type of
2719 peering relationship with your neighbor. Possible values of ``LOCAL-ROLE`` are:
2727 The local Role value is negotiated with the new BGP Role capability with a
2728 built-in check of the corresponding value. In case of mismatch the new OPEN
2729 Roles Mismatch Notification <2, 11> would be sent.
2731 The correct Role pairs are:
2733 * Provider - Customer
2735 * RS-Server - RS-Client
2737 .. code-block:: shell
2739 ~# vtysh -c 'show bgp neighbor' | grep 'Role'
2740 Local Role: customer
2741 Neighbor Role: provider
2742 Role: advertised and received
2744 If strict-mode is set BGP session won't become established until BGP neighbor
2745 set local Role on its side. This configuration parameter is defined in
2746 :rfc:`9234` and used to enforce corresponding configuration at your
2747 counter-part side. Default value - disabled.
2749 Routes that sent from provider, rs-server, or peer local-role (or if received
2750 by customer, rs-clinet, or peer local-role) will be marked with a new
2751 Only to Customer (OTC) attribute.
2753 Routes with this attribute can only be sent to your neighbor if your
2754 local-role is provider or rs-server. Routes with this attribute can be
2755 received only if your local-role is customer or rs-client.
2757 In case of peer-peer relationship routes can be received only if
2758 OTC value is equal to your neighbor AS number.
2760 All these rules with OTC help to detect and mitigate route leaks and
2761 happened automatically if local-role is set.
2763 .. clicmd:: neighbor PEER local-role LOCAL-ROLE [strict-mode]
2765 This command set your local-role to ``LOCAL-ROLE``:
2766 <provider|rs-server|rs-client|customer|peer>.
2768 This role helps to detect and prevent route leaks.
2770 If ``strict-mode`` is set, your neighbor must send you Capability with the
2771 value of his role (by setting local-role on his side). Otherwise, a Role
2772 Mismatch Notification will be sent.
2779 *bgpd* supports :abbr:`L3VPN (Layer 3 Virtual Private Networks)` :abbr:`VRFs
2780 (Virtual Routing and Forwarding)` for IPv4 :rfc:`4364` and IPv6 :rfc:`4659`.
2781 L3VPN routes, and their associated VRF MPLS labels, can be distributed to VPN
2782 SAFI neighbors in the *default*, i.e., non VRF, BGP instance. VRF MPLS labels
2783 are reached using *core* MPLS labels which are distributed using LDP or BGP
2784 labeled unicast. *bgpd* also supports inter-VRF route leaking.
2787 L3VPN over GRE interfaces
2788 ^^^^^^^^^^^^^^^^^^^^^^^^^
2790 In MPLS-VPN or SRv6-VPN, an L3VPN next-hop entry requires that the path
2791 chosen respectively contains a labelled path or a valid SID IPv6 address.
2792 Otherwise the L3VPN entry will not be installed. It is possible to ignore
2793 that check when the path chosen by the next-hop uses a GRE interface, and
2794 there is a route-map configured at inbound side of ipv4-vpn or ipv6-vpn
2795 address family with following syntax:
2797 .. clicmd:: set l3vpn next-hop encapsulation gre
2799 The incoming BGP L3VPN entry is accepted, provided that the next hop of the
2800 L3VPN entry uses a path that takes the GRE tunnel as outgoing interface. The
2801 remote endpoint should be configured just behind the GRE tunnel; remote
2802 device configuration may vary depending whether it acts at edge endpoint or
2803 not: in any case, the expectation is that incoming MPLS traffic received at
2804 this endpoint should be considered as a valid path for L3VPN.
2806 .. _bgp-vrf-route-leaking:
2811 BGP routes may be leaked (i.e. copied) between a unicast VRF RIB and the VPN
2812 SAFI RIB of the default VRF for use in MPLS-based L3VPNs. Unicast routes may
2813 also be leaked between any VRFs (including the unicast RIB of the default BGP
2814 instanced). A shortcut syntax is also available for specifying leaking from one
2815 VRF to another VRF using the default instance's VPN RIB as the intermediary. A
2816 common application of the VRF-VRF feature is to connect a customer's private
2817 routing domain to a provider's VPN service. Leaking is configured from the
2818 point of view of an individual VRF: ``import`` refers to routes leaked from VPN
2819 to a unicast VRF, whereas ``export`` refers to routes leaked from a unicast VRF
2825 Routes exported from a unicast VRF to the VPN RIB must be augmented by two
2828 - an :abbr:`RD (Route Distinguisher)`
2829 - an :abbr:`RTLIST (Route-target List)`
2831 Configuration for these exported routes must, at a minimum, specify these two
2834 Routes imported from the VPN RIB to a unicast VRF are selected according to
2835 their RTLISTs. Routes whose RTLIST contains at least one route-target in
2836 common with the configured import RTLIST are leaked. Configuration for these
2837 imported routes must specify an RTLIST to be matched.
2839 The RD, which carries no semantic value, is intended to make the route unique
2840 in the VPN RIB among all routes of its prefix that originate from all the
2841 customers and sites that are attached to the provider's VPN service.
2842 Accordingly, each site of each customer is typically assigned an RD that is
2843 unique across the entire provider network.
2845 The RTLIST is a set of route-target extended community values whose purpose is
2846 to specify route-leaking policy. Typically, a customer is assigned a single
2847 route-target value for import and export to be used at all customer sites. This
2848 configuration specifies a simple topology wherein a customer has a single
2849 routing domain which is shared across all its sites. More complex routing
2850 topologies are possible through use of additional route-targets to augment the
2851 leaking of sets of routes in various ways.
2853 When using the shortcut syntax for vrf-to-vrf leaking, the RD and RT are
2856 General configuration
2857 ^^^^^^^^^^^^^^^^^^^^^
2859 Configuration of route leaking between a unicast VRF RIB and the VPN SAFI RIB
2860 of the default VRF is accomplished via commands in the context of a VRF
2863 .. clicmd:: rd vpn export AS:NN|IP:nn
2865 Specifies the route distinguisher to be added to a route exported from the
2866 current unicast VRF to VPN.
2868 .. clicmd:: rt vpn import|export|both RTLIST...
2870 Specifies the route-target list to be attached to a route (export) or the
2871 route-target list to match against (import) when exporting/importing between
2872 the current unicast VRF and VPN.
2874 The RTLIST is a space-separated list of route-targets, which are BGP
2875 extended community values as described in
2876 :ref:`bgp-extended-communities-attribute`.
2878 .. clicmd:: label vpn export (0..1048575)|auto
2880 Enables an MPLS label to be attached to a route exported from the current
2881 unicast VRF to VPN. If the value specified is ``auto``, the label value is
2882 automatically assigned from a pool maintained by the Zebra daemon. If Zebra
2883 is not running, or if this command is not configured, automatic label
2884 assignment will not complete, which will block corresponding route export.
2886 .. clicmd:: nexthop vpn export A.B.C.D|X:X::X:X
2888 Specifies an optional nexthop value to be assigned to a route exported from
2889 the current unicast VRF to VPN. If left unspecified, the nexthop will be set
2890 to 0.0.0.0 or 0:0::0:0 (self).
2892 .. clicmd:: route-map vpn import|export MAP
2894 Specifies an optional route-map to be applied to routes imported or exported
2895 between the current unicast VRF and VPN.
2897 .. clicmd:: import|export vpn
2899 Enables import or export of routes between the current unicast VRF and VPN.
2901 .. clicmd:: import vrf VRFNAME
2903 Shortcut syntax for specifying automatic leaking from vrf VRFNAME to
2904 the current VRF using the VPN RIB as intermediary. The RD and RT
2905 are auto derived and should not be specified explicitly for either the
2906 source or destination VRF's.
2908 This shortcut syntax mode is not compatible with the explicit
2909 `import vpn` and `export vpn` statements for the two VRF's involved.
2910 The CLI will disallow attempts to configure incompatible leaking
2913 .. clicmd:: bgp retain route-target all
2915 It is possible to retain or not VPN prefixes that are not imported by local
2916 VRF configuration. This can be done via the following command in the context
2917 of the global VPNv4/VPNv6 family. This command defaults to on and is not
2919 The `no bgp retain route-target all` form of the command is displayed.
2921 .. clicmd:: neighbor <A.B.C.D|X:X::X:X|WORD> soo EXTCOMMUNITY
2923 Without this command, SoO extended community attribute is configured using
2924 an inbound route map that sets the SoO value during the update process.
2925 With the introduction of the new BGP per-neighbor Site-of-Origin (SoO) feature,
2926 two new commands configured in sub-modes under router configuration mode
2927 simplify the SoO value configuration.
2929 If we configure SoO per neighbor at PEs, the SoO community is automatically
2930 added for all routes from the CPEs. Routes are validated and prevented from
2931 being sent back to the same CPE (e.g.: multi-site). This is especially needed
2932 when using ``as-override`` or ``allowas-in`` to prevent routing loops.
2934 .. clicmd:: mpls bgp forwarding
2936 It is possible to permit BGP install VPN prefixes without transport labels,
2937 by issuing the following command under the interface configuration context.
2938 This configuration will install VPN prefixes originated from an e-bgp session,
2939 and with the next-hop directly connected.
2946 .. clicmd:: segment-routing srv6
2948 Use SRv6 backend with BGP L3VPN, and go to its configuration node.
2950 .. clicmd:: locator NAME
2952 Specify the SRv6 locator to be used for SRv6 L3VPN. The Locator name must
2953 be set in zebra, but user can set it in any order.
2955 General configuration
2956 ^^^^^^^^^^^^^^^^^^^^^
2958 Configuration of the SRv6 SID used to advertise a L3VPN for both IPv4 and IPv6
2959 is accomplished via the following command in the context of a VRF:
2961 .. clicmd:: sid vpn per-vrf export (1..1048575)|auto
2963 Enables a SRv6 SID to be attached to a route exported from the current
2964 unicast VRF to VPN. A single SID is used for both IPv4 and IPv6 address
2965 families. If you want to set a SID for only IPv4 address family or IPv6
2966 address family, you need to use the command ``sid vpn export (1..1048575)|auto``
2967 in the context of an address-family. If the value specified is ``auto``,
2968 the SID value is automatically assigned from a pool maintained by the Zebra
2969 daemon. If Zebra is not running, or if this command is not configured, automatic
2970 SID assignment will not complete, which will block corresponding route export.
2974 Ethernet Virtual Network - EVPN
2975 -------------------------------
2977 Note: When using EVPN features and if you have a large number of hosts, make
2978 sure to adjust the size of the arp neighbor cache to avoid neighbor table
2979 overflow and/or excessive garbage collection. On Linux, the size of the table
2980 and garbage collection frequency can be controlled via the following
2981 sysctl configurations:
2983 .. code-block:: shell
2985 net.ipv4.neigh.default.gc_thresh1
2986 net.ipv4.neigh.default.gc_thresh2
2987 net.ipv4.neigh.default.gc_thresh3
2989 net.ipv6.neigh.default.gc_thresh1
2990 net.ipv6.neigh.default.gc_thresh2
2991 net.ipv6.neigh.default.gc_thresh3
2993 For more information, see ``man 7 arp``.
2995 .. _bgp-enabling-evpn:
3000 EVPN should be enabled on the BGP instance corresponding to the VRF acting as
3001 the underlay for the VXLAN tunneling. In most circumstances this will be the
3002 default VRF. The command to enable EVPN for a BGP instance is
3003 ``advertise-all-vni`` which lives under ``address-family l2vpn evpn``:
3009 address-family l2vpn evpn
3012 A more comprehensive configuration example can be found in the :ref:`evpn` page.
3014 .. _bgp-evpn-l3-route-targets:
3016 EVPN L3 Route-Targets
3017 ^^^^^^^^^^^^^^^^^^^^^
3019 .. clicmd:: route-target <import|export|both> <RTLIST|auto>
3021 Modify the route-target set for EVPN advertised type-2/type-5 routes.
3022 RTLIST is a list of any of matching
3023 ``(A.B.C.D:MN|EF:OPQR|GHJK:MN|*:OPQR|*:MN)`` where ``*`` indicates wildcard
3024 matching for the AS number. It will be set to match any AS number. This is
3025 useful in datacenter deployments with Downstream VNI. ``auto`` is used to
3026 retain the autoconfigure that is default behavior for L3 RTs.
3028 .. _bgp-evpn-advertise-pip:
3033 In a EVPN symmetric routing MLAG deployment, all EVPN routes advertised
3034 with anycast-IP as next-hop IP and anycast MAC as the Router MAC (RMAC - in
3035 BGP EVPN Extended-Community).
3036 EVPN picks up the next-hop IP from the VxLAN interface's local tunnel IP and
3037 the RMAC is obtained from the MAC of the L3VNI's SVI interface.
3038 Note: Next-hop IP is used for EVPN routes whether symmetric routing is
3039 deployed or not but the RMAC is only relevant for symmetric routing scenario.
3041 Current behavior is not ideal for Prefix (type-5) and self (type-2)
3042 routes. This is because the traffic from remote VTEPs routed sub optimally
3043 if they land on the system where the route does not belong.
3045 The advertise-pip feature advertises Prefix (type-5) and self (type-2)
3046 routes with system's individual (primary) IP as the next-hop and individual
3047 (system) MAC as Router-MAC (RMAC), while leaving the behavior unchanged for
3050 To support this feature there needs to have ability to co-exist a
3051 (system-MAC, system-IP) pair with a (anycast-MAC, anycast-IP) pair with the
3052 ability to terminate VxLAN-encapsulated packets received for either pair on
3053 the same L3VNI (i.e associated VLAN). This capability is needed per tenant
3056 To derive the system-MAC and the anycast MAC, there must be a
3057 separate/additional MAC-VLAN interface corresponding to L3VNI’s SVI.
3058 The SVI interface’s MAC address can be interpreted as system-MAC
3059 and MAC-VLAN interface's MAC as anycast MAC.
3061 To derive system-IP and anycast-IP, the default BGP instance's router-id is used
3062 as system-IP and the VxLAN interface’s local tunnel IP as the anycast-IP.
3064 User has an option to configure the system-IP and/or system-MAC value if the
3065 auto derived value is not preferred.
3067 Note: By default, advertise-pip feature is enabled and user has an option to
3068 disable the feature via configuration CLI. Once the feature is disabled under
3069 bgp vrf instance or MAC-VLAN interface is not configured, all the routes follow
3070 the same behavior of using same next-hop and RMAC values.
3072 .. clicmd:: advertise-pip [ip <addr> [mac <addr>]]
3074 Enables or disables advertise-pip feature, specify system-IP and/or system-MAC
3077 EVPN advertise-svi-ip
3078 ^^^^^^^^^^^^^^^^^^^^^
3079 Typically, the SVI IP address is reused on VTEPs across multiple racks. However,
3080 if you have unique SVI IP addresses that you want to be reachable you can use the
3081 advertise-svi-ip option. This option advertises the SVI IP/MAC address as a type-2
3082 route and eliminates the need for any flooding over VXLAN to reach the IP from a
3085 .. clicmd:: advertise-svi-ip
3087 Note that you should not enable both the advertise-svi-ip and the advertise-default-gw
3090 .. _bgp-evpn-overlay-index-gateway-ip:
3092 EVPN Overlay Index Gateway IP
3093 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
3094 RFC https://datatracker.ietf.org/doc/html/rfc9136 explains the use of overlay
3095 indexes for recursive route resolution for EVPN type-5 route.
3097 We support gateway IP overlay index.
3098 A gateway IP, advertised with EVPN prefix route, is used to find an EVPN MAC/IP
3099 route with its IP field same as the gateway IP. This MAC/IP entry provides the
3100 nexthop VTEP and the tunnel information required for the VxLAN encapsulation.
3106 . +--------+ BGP +--------+ BGP +--------+ +--------+
3107 SN1 | | IPv4 | | EVPN | | | |
3108 ======+ Host1 +------+ PE1 +------+ PE2 +------+ Host2 +
3110 +--------+ +--------+ +--------+ +--------+
3112 Consider above topology where prefix SN1 is connected behind host1. Host1
3113 advertises SN1 to PE1 over BGP IPv4 session. PE1 advertises SN1 to PE2 using
3114 EVPN type-5 route with host1 IP as the gateway IP. PE1 also advertises
3115 Host1 MAC/IP as type-2 route which is used to resolve host1 gateway IP.
3117 PE2 receives this type-5 route and imports it into the vrf based on route
3118 targets. BGP prefix imported into the vrf uses gateway IP as its BGP nexthop.
3119 This route is installed into zebra if following conditions are satisfied:
3121 1. Gateway IP nexthop is L3 reachable.
3122 2. PE2 has received EVPN type-2 route with IP field set to gateway IP.
3124 Topology requirements:
3126 1. This feature is supported for asymmetric routing model only. While
3127 sending packets to SN1, ingress PE (PE2) performs routing and
3128 egress PE (PE1) performs only bridging.
3129 2. This feature supports only traditional(non vlan-aware) bridge model. Bridge
3130 interface associated with L2VNI is an L3 interface. i.e., this interface is
3131 configured with an address in the L2VNI subnet. Note that the gateway IP
3132 should also have an address in the same subnet.
3133 3. As this feature works in asymmetric routing model, all L2VNIs and corresponding
3134 VxLAN and bridge interfaces should be present at all the PEs.
3135 4. L3VNI configuration is required to generate and import EVPN type-5 routes.
3136 L3VNI VxLAN and bridge interfaces also should be present.
3138 A PE can use one of the following two mechanisms to advertise an EVPN type-5
3139 route with gateway IP.
3141 1. CLI to add gateway IP while generating EVPN type-5 route from a BGP IPv4/IPv6
3144 .. clicmd:: advertise <ipv4|ipv6> unicast [gateway-ip]
3146 When this CLI is configured for a BGP vrf under L2VPN EVPN address family, EVPN
3147 type-5 routes are generated for BGP prefixes in the vrf. Nexthop of the BGP
3148 prefix becomes the gateway IP of the corresponding type-5 route.
3150 If the above command is configured without the "gateway-ip" keyword, type-5
3151 routes are generated without overlay index.
3153 2. Add gateway IP to EVPN type-5 route using a route-map:
3155 .. clicmd:: set evpn gateway-ip <ipv4|ipv6> <addr>
3157 When route-map with above set clause is applied as outbound policy in BGP, it
3158 will set the gateway-ip in EVPN type-5 NLRI.
3160 Example configuration:
3165 neighbor 192.168.0.1 remote-as 101
3167 address-family ipv4 l2vpn evpn
3168 neighbor 192.168.0.1 route-map RMAP out
3171 route-map RMAP permit 10
3172 set evpn gateway-ip 10.0.0.1
3173 set evpn gateway-ip 10::1
3175 A PE that receives a type-5 route with gateway IP overlay index should have
3176 "enable-resolve-overlay-index" configuration enabled to recursively resolve the
3177 overlay index nexthop and install the prefix into zebra.
3179 .. clicmd:: enable-resolve-overlay-index
3181 Example configuration:
3186 bgp router-id 192.168.100.1
3187 no bgp ebgp-requires-policy
3188 neighbor 10.0.1.2 remote-as 65002
3190 address-family l2vpn evpn
3191 neighbor 10.0.1.2 activate
3193 enable-resolve-overlay-index
3202 All-Active Multihoming is used for redundancy and load sharing. Servers
3203 are attached to two or more PEs and the links are bonded (link-aggregation).
3204 This group of server links is referred to as an Ethernet Segment.
3208 An Ethernet Segment can be configured by specifying a system-MAC and a
3209 local discriminator or a complete ESINAME against the bond interface on the
3212 .. clicmd:: evpn mh es-id <(1-16777215)|ESINAME>
3214 .. clicmd:: evpn mh es-sys-mac X:X:X:X:X:X
3216 The sys-mac and local discriminator are used for generating a 10-byte,
3217 Type-3 Ethernet Segment ID. ESINAME is a 10-byte, Type-0 Ethernet Segment ID -
3218 "00:AA:BB:CC:DD:EE:FF:GG:HH:II".
3220 Type-1 (EAD-per-ES and EAD-per-EVI) routes are used to advertise the locally
3221 attached ESs and to learn off remote ESs in the network. Local Type-2/MAC-IP
3222 routes are also advertised with a destination ESI allowing for MAC-IP syncing
3223 between Ethernet Segment peers.
3224 Reference: RFC 7432, RFC 8365
3226 EVPN-MH is intended as a replacement for MLAG or Anycast VTEPs. In
3227 multihoming each PE has an unique VTEP address which requires the introduction
3228 of a new dataplane construct, MAC-ECMP. Here a MAC/FDB entry can point to a
3229 list of remote PEs/VTEPs.
3233 Type-4 (ESR) routes are used for Designated Forwarder (DF) election. DFs
3234 forward BUM traffic received via the overlay network. This implementation
3235 uses a preference based DF election specified by draft-ietf-bess-evpn-pref-df.
3236 The DF preference is configurable per-ES (via zebra) -
3238 .. clicmd:: evpn mh es-df-pref (1-16777215)
3240 BUM traffic is rxed via the overlay by all PEs attached to a server but
3241 only the DF can forward the de-capsulated traffic to the access port. To
3242 accommodate that non-DF filters are installed in the dataplane to drop
3245 Similarly traffic received from ES peers via the overlay cannot be forwarded
3246 to the server. This is split-horizon-filtering with local bias.
3250 Some vendors do not send EAD-per-EVI routes. To interop with them we
3251 need to relax the dependency on EAD-per-EVI routes and activate a remote
3252 ES-PE based on just the EAD-per-ES route.
3254 Note that by default we advertise and expect EAD-per-EVI routes.
3256 .. clicmd:: disable-ead-evi-rx
3258 .. clicmd:: disable-ead-evi-tx
3262 As the primary purpose of EVPN-MH is redundancy keeping the failover efficient
3263 is a recurring theme in the implementation. Following sub-features have
3264 been introduced for the express purpose of efficient ES failovers.
3266 - Layer-2 Nexthop Groups and MAC-ECMP via L2NHG.
3268 - Host routes (for symmetric IRB) via L3NHG.
3269 On dataplanes that support layer3 nexthop groups the feature can be turned
3270 on via the following BGP config -
3272 .. clicmd:: use-es-l3nhg
3274 - Local ES (MAC/Neigh) failover via ES-redirect.
3275 On dataplanes that do not have support for ES-redirect the feature can be
3276 turned off via the following zebra config -
3278 .. clicmd:: evpn mh redirect-off
3280 Uplink/Core tracking
3281 """"""""""""""""""""
3282 When all the underlay links go down the PE no longer has access to the VxLAN
3283 +overlay. To prevent blackholing of traffic the server/ES links are
3284 protodowned on the PE. A link can be setup for uplink tracking via the
3285 following zebra configuration -
3287 .. clicmd:: evpn mh uplink
3289 Proxy advertisements
3290 """"""""""""""""""""
3291 To handle hitless upgrades support for proxy advertisement has been added
3292 as specified by draft-rbickhart-evpn-ip-mac-proxy-adv. This allows a PE
3293 (say PE1) to proxy advertise a MAC-IP rxed from an ES peer (say PE2). When
3294 the ES peer (PE2) goes down PE1 continues to advertise hosts learnt from PE2
3295 for a holdtime during which it attempts to establish local reachability of
3296 the host. This holdtime is configurable via the following zebra commands -
3298 .. clicmd:: evpn mh neigh-holdtime (0-86400)
3300 .. clicmd:: evpn mh mac-holdtime (0-86400)
3304 When a switch is rebooted we wait for a brief period to allow the underlay
3305 and EVPN network to converge before enabling the ESs. For this duration the
3306 ES bonds are held protodown. The startup delay is configurable via the
3307 following zebra command -
3309 .. clicmd:: evpn mh startup-delay (0-3600)
3311 EAD-per-ES fragmentation
3312 """"""""""""""""""""""""
3313 The EAD-per-ES route carries the EVI route targets for all the broadcast
3314 domains associated with the ES. Depending on the EVI scale the EAD-per-ES
3315 route maybe fragmented.
3317 The number of EVIs per-EAD route can be configured via the following
3320 .. clicmd:: [no] ead-es-frag evi-limit (1-1000)
3322 Sample Configuration
3323 ^^^^^^^^^^^^^^^^^^^^^
3329 address-family l2vpn evpn
3330 ead-es-frag evi-limit 200
3335 EAD-per-ES route-target
3336 """""""""""""""""""""""
3337 The EAD-per-ES route by default carries all the EVI route targets. Depending
3338 on EVI scale that can result in route fragmentation. In some cases it maybe
3339 necessary to avoid this fragmentation and that can be done via the following
3341 1. Configure a single supplementary BD per-tenant VRF. This SBD needs to
3342 be provisioned on all EVPN PEs associated with the tenant-VRF.
3343 2. Config the SBD's RT as the EAD-per-ES route's export RT.
3345 Sample Configuration
3346 ^^^^^^^^^^^^^^^^^^^^^
3352 address-family l2vpn evpn
3353 ead-es-route-target export 5556:1001
3354 ead-es-route-target export 5556:1004
3355 ead-es-route-target export 5556:1008
3359 Support with VRF network namespace backend
3360 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
3361 It is possible to separate overlay networks contained in VXLAN interfaces from
3362 underlay networks by using VRFs. VRF-lite and VRF-netns backends can be used for
3363 that. In the latter case, it is necessary to set both bridge and vxlan interface
3364 in the same network namespace, as below example illustrates:
3366 .. code-block:: shell
3370 ip link add name vxlan101 type vxlan id 101 dstport 4789 dev eth0 local 10.1.1.1
3371 ip link set dev vxlan101 netns vrf1
3372 ip netns exec vrf1 ip link set dev lo up
3373 ip netns exec vrf1 brctl addbr bridge101
3374 ip netns exec vrf1 brctl addif bridge101 vxlan101
3376 This makes it possible to separate not only layer 3 networks like VRF-lite networks.
3377 Also, VRF netns based make possible to separate layer 2 networks on separate VRF
3380 .. _bgp-conditional-advertisement:
3382 BGP Conditional Advertisement
3383 -----------------------------
3384 The BGP conditional advertisement feature uses the ``non-exist-map`` or the
3385 ``exist-map`` and the ``advertise-map`` keywords of the neighbor advertise-map
3386 command in order to track routes by the route prefix.
3389 1. If a route prefix is not present in the output of non-exist-map command,
3390 then advertise the route specified by the advertise-map command.
3392 2. If a route prefix is present in the output of non-exist-map command,
3393 then do not advertise the route specified by the addvertise-map command.
3396 1. If a route prefix is present in the output of exist-map command,
3397 then advertise the route specified by the advertise-map command.
3399 2. If a route prefix is not present in the output of exist-map command,
3400 then do not advertise the route specified by the advertise-map command.
3402 This feature is useful when some prefixes are advertised to one of its peers
3403 only if the information from the other peer is not present (due to failure in
3404 peering session or partial reachability etc).
3406 The conditional BGP announcements are sent in addition to the normal
3407 announcements that a BGP router sends to its peer.
3409 The conditional advertisement process is triggered by the BGP scanner process,
3410 which runs every 60 by default. This means that the maximum time for the
3411 conditional advertisement to take effect is the value of the process timer.
3413 As an optimization, while the process always runs on each timer expiry, it
3414 determines whether or not the conditional advertisement policy or the routing
3415 table has changed; if neither have changed, no processing is necessary and the
3416 scanner exits early.
3418 .. clicmd:: neighbor A.B.C.D advertise-map NAME [exist-map|non-exist-map] NAME
3420 This command enables BGP scanner process to monitor routes specified by
3421 exist-map or non-exist-map command in BGP table and conditionally advertises
3422 the routes specified by advertise-map command.
3424 .. clicmd:: bgp conditional-advertisement timer (5-240)
3426 Set the period to rerun the conditional advertisement scanner process. The
3427 default is 60 seconds.
3429 Sample Configuration
3430 ^^^^^^^^^^^^^^^^^^^^^
3434 ip address 10.10.10.2/24
3437 ip address 10.10.20.2/24
3440 ip address 203.0.113.1/32
3443 bgp log-neighbor-changes
3444 no bgp ebgp-requires-policy
3445 neighbor 10.10.10.1 remote-as 1
3446 neighbor 10.10.20.3 remote-as 3
3448 address-family ipv4 unicast
3449 neighbor 10.10.10.1 soft-reconfiguration inbound
3450 neighbor 10.10.20.3 soft-reconfiguration inbound
3451 neighbor 10.10.20.3 advertise-map ADV-MAP non-exist-map EXIST-MAP
3454 ip prefix-list DEFAULT seq 5 permit 192.0.2.5/32
3455 ip prefix-list DEFAULT seq 10 permit 192.0.2.1/32
3456 ip prefix-list EXIST seq 5 permit 10.10.10.10/32
3457 ip prefix-list DEFAULT-ROUTE seq 5 permit 0.0.0.0/0
3458 ip prefix-list IP1 seq 5 permit 10.139.224.0/20
3460 bgp community-list standard DC-ROUTES seq 5 permit 64952:3008
3461 bgp community-list standard DC-ROUTES seq 10 permit 64671:501
3462 bgp community-list standard DC-ROUTES seq 15 permit 64950:3009
3463 bgp community-list standard DEFAULT-ROUTE seq 5 permit 65013:200
3465 route-map ADV-MAP permit 10
3466 match ip address prefix-list IP1
3468 route-map ADV-MAP permit 20
3469 match community DC-ROUTES
3471 route-map EXIST-MAP permit 10
3472 match community DEFAULT-ROUTE
3473 match ip address prefix-list DEFAULT-ROUTE
3479 When default route is present in R2'2 BGP table, 10.139.224.0/20 and 192.0.2.1/32 are not advertised to R3.
3483 Router2# show ip bgp
3484 BGP table version is 20, local router ID is 203.0.113.1, vrf id 0
3485 Default local pref 100, local AS 2
3486 Status codes: s suppressed, d damped, h history, * valid, > best, = multipath,
3487 i internal, r RIB-failure, S Stale, R Removed
3488 Nexthop codes: @NNN nexthop's vrf id, < announce-nh-self
3489 Origin codes: i - IGP, e - EGP, ? - incomplete
3490 RPKI validation codes: V valid, I invalid, N Not found
3492 Network Next Hop Metric LocPrf Weight Path
3493 *> 0.0.0.0/0 10.10.10.1 0 0 1 i
3494 *> 10.139.224.0/20 10.10.10.1 0 0 1 ?
3495 *> 192.0.2.1/32 10.10.10.1 0 0 1 i
3496 *> 192.0.2.5/32 10.10.10.1 0 0 1 i
3498 Displayed 4 routes and 4 total paths
3499 Router2# show ip bgp neighbors 10.10.20.3
3501 !--- Output suppressed.
3503 For address family: IPv4 Unicast
3504 Update group 7, subgroup 7
3505 Packet Queue length 0
3506 Inbound soft reconfiguration allowed
3507 Community attribute sent to this neighbor(all)
3508 Condition NON_EXIST, Condition-map *EXIST-MAP, Advertise-map *ADV-MAP, status: Withdraw
3511 !--- Output suppressed.
3513 Router2# show ip bgp neighbors 10.10.20.3 advertised-routes
3514 BGP table version is 20, local router ID is 203.0.113.1, vrf id 0
3515 Default local pref 100, local AS 2
3516 Status codes: s suppressed, d damped, h history, * valid, > best, = multipath,
3517 i internal, r RIB-failure, S Stale, R Removed
3518 Nexthop codes: @NNN nexthop's vrf id, < announce-nh-self
3519 Origin codes: i - IGP, e - EGP, ? - incomplete
3520 RPKI validation codes: V valid, I invalid, N Not found
3522 Network Next Hop Metric LocPrf Weight Path
3523 *> 0.0.0.0/0 0.0.0.0 0 1 i
3524 *> 192.0.2.5/32 0.0.0.0 0 1 i
3526 Total number of prefixes 2
3528 When default route is not present in R2'2 BGP table, 10.139.224.0/20 and 192.0.2.1/32 are advertised to R3.
3532 Router2# show ip bgp
3533 BGP table version is 21, local router ID is 203.0.113.1, vrf id 0
3534 Default local pref 100, local AS 2
3535 Status codes: s suppressed, d damped, h history, * valid, > best, = multipath,
3536 i internal, r RIB-failure, S Stale, R Removed
3537 Nexthop codes: @NNN nexthop's vrf id, < announce-nh-self
3538 Origin codes: i - IGP, e - EGP, ? - incomplete
3539 RPKI validation codes: V valid, I invalid, N Not found
3541 Network Next Hop Metric LocPrf Weight Path
3542 *> 10.139.224.0/20 10.10.10.1 0 0 1 ?
3543 *> 192.0.2.1/32 10.10.10.1 0 0 1 i
3544 *> 192.0.2.5/32 10.10.10.1 0 0 1 i
3546 Displayed 3 routes and 3 total paths
3548 Router2# show ip bgp neighbors 10.10.20.3
3550 !--- Output suppressed.
3552 For address family: IPv4 Unicast
3553 Update group 7, subgroup 7
3554 Packet Queue length 0
3555 Inbound soft reconfiguration allowed
3556 Community attribute sent to this neighbor(all)
3557 Condition NON_EXIST, Condition-map *EXIST-MAP, Advertise-map *ADV-MAP, status: Advertise
3560 !--- Output suppressed.
3562 Router2# show ip bgp neighbors 10.10.20.3 advertised-routes
3563 BGP table version is 21, local router ID is 203.0.113.1, vrf id 0
3564 Default local pref 100, local AS 2
3565 Status codes: s suppressed, d damped, h history, * valid, > best, = multipath,
3566 i internal, r RIB-failure, S Stale, R Removed
3567 Nexthop codes: @NNN nexthop's vrf id, < announce-nh-self
3568 Origin codes: i - IGP, e - EGP, ? - incomplete
3569 RPKI validation codes: V valid, I invalid, N Not found
3571 Network Next Hop Metric LocPrf Weight Path
3572 *> 10.139.224.0/20 0.0.0.0 0 1 ?
3573 *> 192.0.2.1/32 0.0.0.0 0 1 i
3574 *> 192.0.2.5/32 0.0.0.0 0 1 i
3576 Total number of prefixes 3
3584 .. clicmd:: show debug
3586 Show all enabled debugs.
3588 .. clicmd:: show bgp listeners
3590 Display Listen sockets and the vrf that created them. Useful for debugging of when
3591 listen is not working and this is considered a developer debug statement.
3593 .. clicmd:: debug bgp allow-martian
3595 Enable or disable BGP accepting martian nexthops from a peer. Please note
3596 this is not an actual debug command and this command is also being deprecated
3597 and will be removed soon. The new command is :clicmd:`bgp allow-martian-nexthop`
3599 .. clicmd:: debug bgp bfd
3601 Enable or disable debugging for BFD events. This will show BFD integration
3602 library messages and BGP BFD integration messages that are mostly state
3603 transitions and validation problems.
3605 .. clicmd:: debug bgp conditional-advertisement
3607 Enable or disable debugging of BGP conditional advertisement.
3609 .. clicmd:: debug bgp neighbor-events
3611 Enable or disable debugging for neighbor events. This provides general
3612 information on BGP events such as peer connection / disconnection, session
3613 establishment / teardown, and capability negotiation.
3615 .. clicmd:: debug bgp updates
3617 Enable or disable debugging for BGP updates. This provides information on
3618 BGP UPDATE messages transmitted and received between local and remote
3621 .. clicmd:: debug bgp keepalives
3623 Enable or disable debugging for BGP keepalives. This provides information on
3624 BGP KEEPALIVE messages transmitted and received between local and remote
3627 .. clicmd:: debug bgp bestpath <A.B.C.D/M|X:X::X:X/M>
3629 Enable or disable debugging for bestpath selection on the specified prefix.
3631 .. clicmd:: debug bgp nht
3633 Enable or disable debugging of BGP nexthop tracking.
3635 .. clicmd:: debug bgp update-groups
3637 Enable or disable debugging of dynamic update groups. This provides general
3638 information on group creation, deletion, join and prune events.
3640 .. clicmd:: debug bgp zebra
3642 Enable or disable debugging of communications between *bgpd* and *zebra*.
3644 Dumping Messages and Routing Tables
3645 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
3647 .. clicmd:: dump bgp all PATH [INTERVAL]
3649 .. clicmd:: dump bgp all-et PATH [INTERVAL]
3652 Dump all BGP packet and events to `path` file.
3653 If `interval` is set, a new file will be created for echo `interval` of
3654 seconds. The path `path` can be set with date and time formatting
3655 (strftime). The type ‘all-et’ enables support for Extended Timestamp Header
3656 (:ref:`packet-binary-dump-format`).
3658 .. clicmd:: dump bgp updates PATH [INTERVAL]
3660 .. clicmd:: dump bgp updates-et PATH [INTERVAL]
3663 Dump only BGP updates messages to `path` file.
3664 If `interval` is set, a new file will be created for echo `interval` of
3665 seconds. The path `path` can be set with date and time formatting
3666 (strftime). The type ‘updates-et’ enables support for Extended Timestamp
3667 Header (:ref:`packet-binary-dump-format`).
3669 .. clicmd:: dump bgp routes-mrt PATH
3671 .. clicmd:: dump bgp routes-mrt PATH INTERVAL
3674 Dump whole BGP routing table to `path`. This is heavy process. The path
3675 `path` can be set with date and time formatting (strftime). If `interval` is
3676 set, a new file will be created for echo `interval` of seconds.
3678 Note: the interval variable can also be set using hours and minutes: 04h20m00.
3681 .. _bgp-other-commands:
3686 The following are available in the top level *enable* mode:
3688 .. clicmd:: clear bgp \*
3692 .. clicmd:: clear bgp ipv4|ipv6 \*
3694 Clear all peers with this address-family activated.
3696 .. clicmd:: clear bgp ipv4|ipv6 unicast \*
3698 Clear all peers with this address-family and sub-address-family activated.
3700 .. clicmd:: clear bgp ipv4|ipv6 PEER
3702 Clear peers with address of X.X.X.X and this address-family activated.
3704 .. clicmd:: clear bgp ipv4|ipv6 unicast PEER
3706 Clear peer with address of X.X.X.X and this address-family and sub-address-family activated.
3708 .. clicmd:: clear bgp ipv4|ipv6 PEER soft|in|out
3710 Clear peer using soft reconfiguration in this address-family.
3712 .. clicmd:: clear bgp ipv4|ipv6 unicast PEER soft|in|out
3714 Clear peer using soft reconfiguration in this address-family and sub-address-family.
3716 .. clicmd:: clear bgp [ipv4|ipv6] [unicast] PEER|\* message-stats
3718 Clear BGP message statistics for a specified peer or for all peers,
3719 optionally filtered by activated address-family and sub-address-family.
3721 The following are available in the ``router bgp`` mode:
3723 .. clicmd:: write-quanta (1-64)
3725 BGP message Tx I/O is vectored. This means that multiple packets are written
3726 to the peer socket at the same time each I/O cycle, in order to minimize
3727 system call overhead. This value controls how many are written at a time.
3728 Under certain load conditions, reducing this value could make peer traffic
3729 less 'bursty'. In practice, leave this settings on the default (64) unless
3730 you truly know what you are doing.
3732 .. clicmd:: read-quanta (1-10)
3734 Unlike Tx, BGP Rx traffic is not vectored. Packets are read off the wire one
3735 at a time in a loop. This setting controls how many iterations the loop runs
3736 for. As with write-quanta, it is best to leave this setting on the default.
3738 The following command is available in ``config`` mode as well as in the
3739 ``router bgp`` mode:
3741 .. clicmd:: bgp graceful-shutdown
3743 The purpose of this command is to initiate BGP Graceful Shutdown which
3744 is described in :rfc:`8326`. The use case for this is to minimize or
3745 eliminate the amount of traffic loss in a network when a planned
3746 maintenance activity such as software upgrade or hardware replacement
3747 is to be performed on a router. The feature works by re-announcing
3748 routes to eBGP peers with the GRACEFUL_SHUTDOWN community included.
3749 Peers are then expected to treat such paths with the lowest preference.
3750 This happens automatically on a receiver running FRR; with other
3751 routing protocol stacks, an inbound policy may have to be configured.
3752 In FRR, triggering graceful shutdown also results in announcing a
3753 LOCAL_PREF of 0 to iBGP peers.
3755 Graceful shutdown can be configured per BGP instance or globally for
3756 all of BGP. These two options are mutually exclusive. The no form of
3757 the command causes graceful shutdown to be stopped, and routes will
3758 be re-announced without the GRACEFUL_SHUTDOWN community and/or with
3759 the usual LOCAL_PREF value. Note that if this option is saved to
3760 the startup configuration, graceful shutdown will remain in effect
3761 across restarts of *bgpd* and will need to be explicitly disabled.
3763 .. clicmd:: bgp input-queue-limit (1-4294967295)
3765 Set the BGP Input Queue limit for all peers when messaging parsing. Increase
3766 this only if you have the memory to handle large queues of messages at once.
3768 .. clicmd:: bgp output-queue-limit (1-4294967295)
3770 Set the BGP Output Queue limit for all peers when messaging parsing. Increase
3771 this only if you have the memory to handle large queues of messages at once.
3773 .. _bgp-displaying-bgp-information:
3775 Displaying BGP Information
3776 ==========================
3778 The following four commands display the IPv6 and IPv4 routing tables, depending
3779 on whether or not the ``ip`` keyword is used.
3780 Actually, :clicmd:`show ip bgp` command was used on older `Quagga` routing
3781 daemon project, while :clicmd:`show bgp` command is the new format. The choice
3782 has been done to keep old format with IPv4 routing table, while new format
3783 displays IPv6 routing table.
3785 .. clicmd:: show ip bgp [all] [wide|json [detail]]
3787 .. clicmd:: show ip bgp A.B.C.D [json]
3789 .. clicmd:: show bgp [all] [wide|json [detail]]
3791 .. clicmd:: show bgp X:X::X:X [json]
3793 These commands display BGP routes. When no route is specified, the default
3794 is to display all BGP routes.
3798 BGP table version is 0, local router ID is 10.1.1.1
3799 Status codes: s suppressed, d damped, h history, * valid, > best, i - internal
3800 Origin codes: i - IGP, e - EGP, ? - incomplete
3802 Network Next Hop Metric LocPrf Weight Path
3803 \*> 1.1.1.1/32 0.0.0.0 0 32768 i
3805 Total number of prefixes 1
3807 If ``wide`` option is specified, then the prefix table's width is increased
3808 to fully display the prefix and the nexthop.
3810 This is especially handy dealing with IPv6 prefixes and
3811 if :clicmd:`[no] bgp default show-nexthop-hostname` is enabled.
3813 If ``all`` option is specified, ``ip`` keyword is ignored, show bgp all and
3814 show ip bgp all commands display routes for all AFIs and SAFIs.
3816 If ``json`` option is specified, output is displayed in JSON format.
3818 If ``detail`` option is specified after ``json``, more verbose JSON output
3821 Some other commands provide additional options for filtering the output.
3823 .. clicmd:: show [ip] bgp regexp LINE
3825 This command displays BGP routes using AS path regular expression
3826 (:ref:`bgp-regular-expressions`).
3828 .. clicmd:: show [ip] bgp [all] summary [wide] [json]
3830 Show a bgp peer summary for the specified address family.
3832 The old command structure :clicmd:`show ip bgp` may be removed in the future
3833 and should no longer be used. In order to reach the other BGP routing tables
3834 other than the IPv6 routing table given by :clicmd:`show bgp`, the new command
3835 structure is extended with :clicmd:`show bgp [afi] [safi]`.
3837 ``wide`` option gives more output like ``LocalAS`` and extended ``Desc`` to
3842 exit1# show ip bgp summary wide
3844 IPv4 Unicast Summary (VRF default):
3845 BGP router identifier 192.168.100.1, local AS number 65534 vrf-id 0
3847 RIB entries 5, using 920 bytes of memory
3848 Peers 1, using 27 KiB of memory
3850 Neighbor V AS LocalAS MsgRcvd MsgSent TblVer InQ OutQ Up/Down State/PfxRcd PfxSnt Desc
3851 192.168.0.2 4 65030 123 15 22 0 0 0 00:07:00 0 1 us-east1-rs1.frrouting.org
3853 Total number of neighbors 1
3856 .. clicmd:: show bgp [afi] [safi] [all] [wide|json]
3858 .. clicmd:: show bgp vrfs [<VRFNAME$vrf_name>] [json]
3860 The command displays all bgp vrf instances basic info like router-id,
3861 configured and established neighbors,
3862 evpn related basic info like l3vni, router-mac, vxlan-interface.
3863 User can get that information as JSON format when ``json`` keyword
3864 at the end of cli is presented.
3868 torc-11# show bgp vrfs
3869 Type Id routerId #PeersCfg #PeersEstb Name
3870 L3-VNI RouterMAC Interface
3871 DFLT 0 17.0.0.6 3 3 default
3872 0 00:00:00:00:00:00 unknown
3873 VRF 21 17.0.0.6 0 0 sym_1
3874 8888 34:11:12:22:22:01 vlan4034_l3
3875 VRF 32 17.0.0.6 0 0 sym_2
3876 8889 34:11:12:22:22:01 vlan4035_l3
3878 Total number of VRFs (including default): 3
3880 .. clicmd:: show bgp [<ipv4|ipv6> <unicast|multicast|vpn|labeled-unicast|flowspec> | l2vpn evpn]
3882 These commands display BGP routes for the specific routing table indicated by
3883 the selected afi and the selected safi. If no afi and no safi value is given,
3884 the command falls back to the default IPv6 routing table.
3886 .. clicmd:: show bgp l2vpn evpn route [type <macip|2|multicast|3|es|4|prefix|5>]
3888 EVPN prefixes can also be filtered by EVPN route type.
3890 .. clicmd:: show bgp vni <all|VNI> [vtep VTEP] [type <ead|1|macip|2|multicast|3>] [<detail|json>]
3892 Display per-VNI EVPN routing table in bgp. Filter route-type, vtep, or VNI.
3894 .. clicmd:: show bgp [afi] [safi] [all] summary [json]
3896 Show a bgp peer summary for the specified address family, and subsequent
3899 .. clicmd:: show bgp [afi] [safi] [all] summary failed [json]
3901 Show a bgp peer summary for peers that are not successfully exchanging routes
3902 for the specified address family, and subsequent address-family.
3904 .. clicmd:: show bgp [afi] [safi] [all] summary established [json]
3906 Show a bgp peer summary for peers that are successfully exchanging routes
3907 for the specified address family, and subsequent address-family.
3909 .. clicmd:: show bgp [afi] [safi] [all] summary neighbor [PEER] [json]
3911 Show a bgp summary for the specified peer, address family, and
3912 subsequent address-family. The neighbor filter can be used in combination
3913 with the failed, established filters.
3915 .. clicmd:: show bgp [afi] [safi] [all] summary remote-as <internal|external|ASN> [json]
3917 Show a bgp peer summary for the specified remote-as ASN or type (``internal``
3918 for iBGP and ``external`` for eBGP sessions), address family, and subsequent
3919 address-family. The remote-as filter can be used in combination with the
3920 failed, established filters.
3922 .. clicmd:: show bgp [afi] [safi] [all] summary terse [json]
3924 Shorten the output. Do not show the following information about the BGP
3925 instances: the number of RIB entries, the table version and the used memory.
3926 The ``terse`` option can be used in combination with the remote-as, neighbor,
3927 failed and established filters, and with the ``wide`` option as well.
3929 .. clicmd:: show bgp [afi] [safi] [neighbor [PEER] [routes|advertised-routes|received-routes] [<A.B.C.D/M|X:X::X:X/M> | detail] [json]
3931 This command shows information on a specific BGP peer of the relevant
3932 afi and safi selected.
3934 The ``routes`` keyword displays only routes in this address-family's BGP
3935 table that were received by this peer and accepted by inbound policy.
3937 The ``advertised-routes`` keyword displays only the routes in this
3938 address-family's BGP table that were permitted by outbound policy and
3939 advertised to to this peer.
3941 The ``received-routes`` keyword displays all routes belonging to this
3942 address-family (prior to inbound policy) that were received by this peer.
3944 If a specific prefix is specified, the detailed version of that prefix will
3947 If ``detail`` option is specified, the detailed version of all routes
3948 will be displayed. The same format as ``show [ip] bgp [afi] [safi] PREFIX``
3949 will be used, but for the whole table of received, advertised or filtered
3952 If ``json`` option is specified, output is displayed in JSON format.
3954 .. clicmd:: show bgp [<view|vrf> VIEWVRFNAME] [afi] [safi] neighbors PEER received prefix-filter [json]
3956 Display Address Prefix ORFs received from this peer.
3958 .. clicmd:: show bgp [afi] [safi] [all] dampening dampened-paths [wide|json]
3960 Display paths suppressed due to dampening of the selected afi and safi
3963 .. clicmd:: show bgp [afi] [safi] [all] dampening flap-statistics [wide|json]
3965 Display flap statistics of routes of the selected afi and safi selected.
3967 .. clicmd:: show bgp [afi] [safi] [all] dampening parameters [json]
3969 Display details of configured dampening parameters of the selected afi and
3972 If the ``json`` option is specified, output is displayed in JSON format.
3974 .. clicmd:: show bgp [afi] [safi] [all] version (1-4294967295) [wide|json]
3976 Display prefixes with matching version numbers. The version number and
3977 above having prefixes will be listed here.
3979 It helps to identify which prefixes were installed at some point.
3981 Here is an example of how to check what prefixes were installed starting
3982 with an arbitrary version:
3984 .. code-block:: shell
3986 # vtysh -c 'show bgp ipv4 unicast json' | jq '.tableVersion'
3988 # vtysh -c 'show ip bgp version 9 json' | jq -r '.routes | keys[]'
3990 # vtysh -c 'show ip bgp version 8 json' | jq -r '.routes | keys[]'
3994 .. clicmd:: show bgp [afi] [safi] statistics
3996 Display statistics of routes of the selected afi and safi.
3998 .. clicmd:: show bgp statistics-all
4000 Display statistics of routes of all the afi and safi.
4002 .. clicmd:: show [ip] bgp [afi] [safi] [all] cidr-only [wide|json]
4004 Display routes with non-natural netmasks.
4006 .. clicmd:: show [ip] bgp [afi] [safi] [all] prefix-list WORD [wide|json]
4008 Display routes that match the specified prefix-list.
4010 If ``wide`` option is specified, then the prefix table's width is increased
4011 to fully display the prefix and the nexthop.
4013 If the ``json`` option is specified, output is displayed in JSON format.
4015 .. clicmd:: show [ip] bgp [afi] [safi] [all] access-list WORD [wide|json]
4017 Display routes that match the specified access-list.
4019 .. clicmd:: show [ip] bgp [afi] [safi] [all] filter-list WORD [wide|json]
4021 Display routes that match the specified AS-Path filter-list.
4023 If ``wide`` option is specified, then the prefix table's width is increased
4024 to fully display the prefix and the nexthop.
4026 If the ``json`` option is specified, output is displayed in JSON format.
4028 .. clicmd:: show [ip] bgp [afi] [safi] [all] route-map WORD [wide|json]
4030 Display routes that match the specified route-map.
4032 If ``wide`` option is specified, then the prefix table's width is increased
4033 to fully display the prefix and the nexthop.
4035 If the ``json`` option is specified, output is displayed in JSON format.
4037 .. clicmd:: show [ip] bgp [afi] [safi] [all] <A.B.C.D/M|X:X::X:X/M> longer-prefixes [wide|json]
4039 Displays the specified route and all more specific routes.
4041 If ``wide`` option is specified, then the prefix table's width is increased
4042 to fully display the prefix and the nexthop.
4044 If the ``json`` option is specified, output is displayed in JSON format.
4046 .. clicmd:: show [ip] bgp [afi] [safi] [all] neighbors A.B.C.D [advertised-routes|received-routes|filtered-routes] [<A.B.C.D/M|X:X::X:X/M> | detail] [json|wide]
4048 Display the routes advertised to a BGP neighbor or received routes
4049 from neighbor or filtered routes received from neighbor based on the
4052 If ``wide`` option is specified, then the prefix table's width is increased
4053 to fully display the prefix and the nexthop.
4055 This is especially handy dealing with IPv6 prefixes and
4056 if :clicmd:`[no] bgp default show-nexthop-hostname` is enabled.
4058 If ``all`` option is specified, ``ip`` keyword is ignored and,
4059 routes displayed for all AFIs and SAFIs.
4060 if afi is specified, with ``all`` option, routes will be displayed for
4061 each SAFI in the selcted AFI
4063 If a specific prefix is specified, the detailed version of that prefix will
4066 If ``detail`` option is specified, the detailed version of all routes
4067 will be displayed. The same format as ``show [ip] bgp [afi] [safi] PREFIX``
4068 will be used, but for the whole table of received, advertised or filtered
4071 If ``json`` option is specified, output is displayed in JSON format.
4073 .. clicmd:: show [ip] bgp [afi] [safi] [all] detail-routes
4075 Display the detailed version of all routes. The same format as using
4076 ``show [ip] bgp [afi] [safi] PREFIX``, but for the whole BGP table.
4078 If ``all`` option is specified, ``ip`` keyword is ignored and,
4079 routes displayed for all AFIs and SAFIs.
4081 If ``afi`` is specified, with ``all`` option, routes will be displayed for
4082 each SAFI in the selected AFI.
4084 .. _bgp-display-routes-by-community:
4086 Displaying Routes by Community Attribute
4087 ----------------------------------------
4089 The following commands allow displaying routes based on their community
4092 .. clicmd:: show [ip] bgp <ipv4|ipv6> [all] community [wide|json]
4094 .. clicmd:: show [ip] bgp <ipv4|ipv6> [all] community COMMUNITY [wide|json]
4096 .. clicmd:: show [ip] bgp <ipv4|ipv6> [all] community COMMUNITY exact-match [wide|json]
4098 These commands display BGP routes which have the community attribute.
4099 attribute. When ``COMMUNITY`` is specified, BGP routes that match that
4100 community are displayed. When `exact-match` is specified, it display only
4101 routes that have an exact match.
4103 .. clicmd:: show [ip] bgp <ipv4|ipv6> community-list WORD [json]
4105 .. clicmd:: show [ip] bgp <ipv4|ipv6> community-list WORD exact-match [json]
4107 These commands display BGP routes for the address family specified that
4108 match the specified community list. When `exact-match` is specified, it
4109 displays only routes that have an exact match.
4111 If ``wide`` option is specified, then the prefix table's width is increased
4112 to fully display the prefix and the nexthop.
4114 This is especially handy dealing with IPv6 prefixes and
4115 if :clicmd:`[no] bgp default show-nexthop-hostname` is enabled.
4117 If ``all`` option is specified, ``ip`` keyword is ignored and,
4118 routes displayed for all AFIs and SAFIs.
4119 if afi is specified, with ``all`` option, routes will be displayed for
4120 each SAFI in the selcted AFI
4122 If ``json`` option is specified, output is displayed in JSON format.
4124 .. clicmd:: show bgp labelpool <chunks|inuse|ledger|requests|summary> [json]
4126 These commands display information about the BGP labelpool used for
4127 the association of MPLS labels with routes for L3VPN and Labeled Unicast
4129 If ``chunks`` option is specified, output shows the current list of label
4130 chunks granted to BGP by Zebra, indicating the start and end label in
4133 If ``inuse`` option is specified, output shows the current inuse list of
4134 label to prefix mappings
4136 If ``ledger`` option is specified, output shows ledger list of all
4137 label requests made per prefix
4139 If ``requests`` option is specified, output shows current list of label
4140 requests which have not yet been fulfilled by the labelpool
4142 If ``summary`` option is specified, output is a summary of the counts for
4143 the chunks, inuse, ledger and requests list along with the count of
4144 outstanding chunk requests to Zebra and the number of zebra reconnects
4147 If ``json`` option is specified, output is displayed in JSON format.
4149 .. _bgp-display-routes-by-lcommunity:
4151 Displaying Routes by Large Community Attribute
4152 ----------------------------------------------
4154 The following commands allow displaying routes based on their
4155 large community attribute.
4157 .. clicmd:: show [ip] bgp <ipv4|ipv6> large-community
4159 .. clicmd:: show [ip] bgp <ipv4|ipv6> large-community LARGE-COMMUNITY
4161 .. clicmd:: show [ip] bgp <ipv4|ipv6> large-community LARGE-COMMUNITY exact-match
4163 .. clicmd:: show [ip] bgp <ipv4|ipv6> large-community LARGE-COMMUNITY json
4165 These commands display BGP routes which have the large community attribute.
4166 attribute. When ``LARGE-COMMUNITY`` is specified, BGP routes that match that
4167 large community are displayed. When `exact-match` is specified, it display
4168 only routes that have an exact match. When `json` is specified, it display
4169 routes in json format.
4171 .. clicmd:: show [ip] bgp <ipv4|ipv6> large-community-list WORD
4173 .. clicmd:: show [ip] bgp <ipv4|ipv6> large-community-list WORD exact-match
4175 .. clicmd:: show [ip] bgp <ipv4|ipv6> large-community-list WORD json
4177 These commands display BGP routes for the address family specified that
4178 match the specified large community list. When `exact-match` is specified,
4179 it displays only routes that have an exact match. When `json` is specified,
4180 it display routes in json format.
4182 .. _bgp-display-routes-by-as-path:
4185 Displaying Routes by AS Path
4186 ----------------------------
4188 .. clicmd:: show bgp ipv4|ipv6 regexp LINE
4190 This commands displays BGP routes that matches a regular
4191 expression `line` (:ref:`bgp-regular-expressions`).
4193 .. clicmd:: show [ip] bgp ipv4 vpn
4195 .. clicmd:: show [ip] bgp ipv6 vpn
4197 Print active IPV4 or IPV6 routes advertised via the VPN SAFI.
4199 .. clicmd:: show bgp ipv4 vpn summary
4201 .. clicmd:: show bgp ipv6 vpn summary
4203 Print a summary of neighbor connections for the specified AFI/SAFI combination.
4205 Displaying Routes by Route Distinguisher
4206 ----------------------------------------
4208 .. clicmd:: show bgp [<ipv4|ipv6> vpn | l2vpn evpn [route]] rd <all|RD>
4210 For L3VPN and EVPN address-families, routes can be displayed on a per-RD
4211 (Route Distinguisher) basis or for all RD's.
4213 .. clicmd:: show bgp l2vpn evpn rd <all|RD> [overlay | tags]
4215 Use the ``overlay`` or ``tags`` keywords to display the overlay/tag
4216 information about the EVPN prefixes in the selected Route Distinguisher.
4218 .. clicmd:: show bgp l2vpn evpn route rd <all|RD> mac <MAC> [ip <MAC>] [json]
4220 For EVPN Type 2 (macip) routes, a MAC address (and optionally an IP address)
4221 can be supplied to the command to only display matching prefixes in the
4224 Displaying Update Group Information
4225 -----------------------------------
4227 .. clicmd:: show bgp update-groups [advertise-queue|advertised-routes|packet-queue]
4229 Display Information about each individual update-group being used.
4230 If SUBGROUP-ID is specified only display about that particular group. If
4231 advertise-queue is specified the list of routes that need to be sent
4232 to the peers in the update-group is displayed, advertised-routes means
4233 the list of routes we have sent to the peers in the update-group and
4234 packet-queue specifies the list of packets in the queue to be sent.
4236 .. clicmd:: show bgp update-groups statistics
4238 Display Information about update-group events in FRR.
4240 Displaying Nexthop Information
4241 ------------------------------
4242 .. clicmd:: show [ip] bgp [<view|vrf> VIEWVRFNAME] nexthop ipv4 [A.B.C.D] [detail] [json]
4244 .. clicmd:: show [ip] bgp [<view|vrf> VIEWVRFNAME] nexthop ipv6 [X:X::X:X] [detail] [json]
4246 .. clicmd:: show [ip] bgp [<view|vrf> VIEWVRFNAME] nexthop [<A.B.C.D|X:X::X:X>] [detail] [json]
4248 .. clicmd:: show [ip] bgp <view|vrf> all nexthop [json]
4250 Display information about nexthops to bgp neighbors. If a certain nexthop is
4251 specified, also provides information about paths associated with the nexthop.
4252 With detail option provides information about gates of each nexthop.
4254 .. clicmd:: show [ip] bgp [<view|vrf> VIEWVRFNAME] import-check-table [detail] [json]
4256 Display information about nexthops from table that is used to check network's
4257 existence in the rib for network statements.
4259 Segment-Routing IPv6
4260 --------------------
4262 .. clicmd:: show bgp segment-routing srv6
4264 This command displays information about SRv6 L3VPN in bgpd. Specifically,
4265 what kind of Locator is being used, and its Locator chunk information.
4266 And the SID of the SRv6 Function that is actually managed on bgpd.
4267 In the following example, bgpd is using a Locator named loc1, and two SRv6
4268 Functions are managed to perform VPNv6 VRF redirect for vrf10 and vrf20.
4272 router# show bgp segment-routing srv6
4277 - sid: 2001:db8:1:1::100
4279 - sid: 2001:db8:1:1::200
4283 vpn_policy[AFI_IP].tovpn_sid: none
4284 vpn_policy[AFI_IP6].tovpn_sid: none
4286 vpn_policy[AFI_IP].tovpn_sid: none
4287 vpn_policy[AFI_IP6].tovpn_sid: 2001:db8:1:1::100
4289 vpn_policy[AFI_IP].tovpn_sid: none
4290 vpn_policy[AFI_IP6].tovpn_sid: 2001:db8:1:1::200
4293 .. _bgp-route-reflector:
4298 BGP routers connected inside the same AS through BGP belong to an internal
4299 BGP session, or IBGP. In order to prevent routing table loops, IBGP does not
4300 advertise IBGP-learned routes to other routers in the same session. As such,
4301 IBGP requires a full mesh of all peers. For large networks, this quickly becomes
4302 unscalable. Introducing route reflectors removes the need for the full-mesh.
4304 When route reflectors are configured, these will reflect the routes announced
4305 by the peers configured as clients. A route reflector client is configured
4308 .. clicmd:: neighbor PEER route-reflector-client
4311 To avoid single points of failure, multiple route reflectors can be configured.
4313 A cluster is a collection of route reflectors and their clients, and is used
4314 by route reflectors to avoid looping.
4316 .. clicmd:: bgp cluster-id A.B.C.D
4318 .. clicmd:: bgp no-rib
4320 To set and unset the BGP daemon ``-n`` / ``--no_kernel`` options during runtime
4321 to disable BGP route installation to the RIB (Zebra), the ``[no] bgp no-rib``
4322 commands can be used;
4324 Please note that setting the option during runtime will withdraw all routes in
4325 the daemons RIB from Zebra and unsetting it will announce all routes in the
4326 daemons RIB to Zebra. If the option is passed as a command line argument when
4327 starting the daemon and the configuration gets saved, the option will persist
4328 unless removed from the configuration with the negating command prior to the
4329 configuration write operation. At this point in time non SAFI_UNICAST BGP
4330 data is not properly withdrawn from zebra when this command is issued.
4332 .. clicmd:: bgp allow-martian-nexthop
4334 When a peer receives a martian nexthop as part of the NLRI for a route
4335 permit the nexthop to be used as such, instead of rejecting and resetting
4338 .. clicmd:: bgp send-extra-data zebra
4340 This command turns on the ability of BGP to send extra data to zebra. Currently,
4341 it's the AS-Path, communities, and the path selection reason. The default
4342 behavior in BGP is not to send this data. If the routes were sent to zebra and
4343 the option is changed, bgpd doesn't reinstall the routes to comply with the new
4346 .. clicmd:: bgp session-dscp (0-63)
4348 This command allows bgp to control, at a global level, the TCP dscp values
4351 .. _bgp-suppress-fib:
4353 Suppressing routes not installed in FIB
4354 =======================================
4356 The FRR implementation of BGP advertises prefixes learnt from a peer to other
4357 peers even if the routes do not get installed in the FIB. There can be
4358 scenarios where the hardware tables in some of the routers (along the path from
4359 the source to destination) is full which will result in all routes not getting
4360 installed in the FIB. If these routes are advertised to the downstream routers
4361 then traffic will start flowing and will be dropped at the intermediate router.
4363 The solution is to provide a configurable option to check for the FIB install
4364 status of the prefixes and advertise to peers if the prefixes are successfully
4365 installed in the FIB. The advertisement of the prefixes are suppressed if it is
4366 not installed in FIB.
4368 The following conditions apply will apply when checking for route installation
4371 1. The advertisement or suppression of routes based on FIB install status
4372 applies only for newly learnt routes from peer (routes which are not in
4374 2. If the route received from peer already exists in BGP local RIB and route
4375 attributes have changed (best path changed), the old path is deleted and
4376 new path is installed in FIB. The FIB install status will not have any
4377 effect. Therefore only when the route is received first time the checks
4379 3. The feature will not apply for routes learnt through other means like
4380 redistribution to bgp from other protocols. This is applicable only to
4382 4. If a route is installed in FIB and then gets deleted from the dataplane,
4383 then routes will not be withdrawn from peers. This will be considered as
4385 5. The feature will slightly increase the time required to advertise the routes
4386 to peers since the route install status needs to be received from the FIB
4387 6. If routes are received by the peer before the configuration is applied, then
4388 the bgp sessions need to be reset for the configuration to take effect.
4389 7. If the route which is already installed in dataplane is removed for some
4390 reason, sending withdraw message to peers is not currently supported.
4392 .. clicmd:: bgp suppress-fib-pending
4394 This command is applicable at the global level and at an individual
4395 bgp level. If applied at the global level all bgp instances will
4396 wait for fib installation before announcing routes and there is no
4397 way to turn it off for a particular bgp vrf.
4404 You can set different routing policy for a peer. For example, you can set
4405 different filter for a peer.
4411 neighbor 10.0.0.1 remote-as 2
4412 address-family ipv4 unicast
4413 neighbor 10.0.0.1 distribute-list 1 in
4417 neighbor 10.0.0.1 remote-as 2
4418 address-family ipv4 unicast
4419 neighbor 10.0.0.1 distribute-list 2 in
4422 This means BGP update from a peer 10.0.0.1 goes to both BGP view 1 and view 2.
4423 When the update is inserted into view 1, distribute-list 1 is applied. On the
4424 other hand, when the update is inserted into view 2, distribute-list 2 is
4428 .. _bgp-regular-expressions:
4430 BGP Regular Expressions
4431 =======================
4433 BGP regular expressions are based on :t:`POSIX 1003.2` regular expressions. The
4434 following description is just a quick subset of the POSIX regular expressions.
4438 Matches any single character.
4441 Matches 0 or more occurrences of pattern.
4444 Matches 1 or more occurrences of pattern.
4447 Match 0 or 1 occurrences of pattern.
4450 Matches the beginning of the line.
4453 Matches the end of the line.
4456 The ``_`` character has special meanings in BGP regular expressions. It
4457 matches to space and comma , and AS set delimiter ``{`` and ``}`` and AS
4458 confederation delimiter ``(`` and ``)``. And it also matches to the
4459 beginning of the line and the end of the line. So ``_`` can be used for AS
4460 value boundaries match. This character technically evaluates to
4464 .. _bgp-configuration-examples:
4466 Miscellaneous Configuration Examples
4467 ====================================
4469 Example of a session to an upstream, advertising only one prefix to it.
4474 bgp router-id 10.236.87.1
4475 neighbor upstream peer-group
4476 neighbor upstream remote-as 64515
4477 neighbor upstream capability dynamic
4478 neighbor 10.1.1.1 peer-group upstream
4479 neighbor 10.1.1.1 description ACME ISP
4481 address-family ipv4 unicast
4482 network 10.236.87.0/24
4483 neighbor upstream prefix-list pl-allowed-adv out
4486 ip prefix-list pl-allowed-adv seq 5 permit 82.195.133.0/25
4487 ip prefix-list pl-allowed-adv seq 10 deny any
4489 A more complex example including upstream, peer and customer sessions
4490 advertising global prefixes and NO_EXPORT prefixes and providing actions for
4491 customer routes based on community values. Extensive use is made of route-maps
4492 and the 'call' feature to support selective advertising of prefixes. This
4493 example is intended as guidance only, it has NOT been tested and almost
4494 certainly contains silly mistakes, if not serious flaws.
4499 bgp router-id 10.236.87.1
4500 neighbor upstream capability dynamic
4501 neighbor cust capability dynamic
4502 neighbor peer capability dynamic
4503 neighbor 10.1.1.1 remote-as 64515
4504 neighbor 10.1.1.1 peer-group upstream
4505 neighbor 10.2.1.1 remote-as 64516
4506 neighbor 10.2.1.1 peer-group upstream
4507 neighbor 10.3.1.1 remote-as 64517
4508 neighbor 10.3.1.1 peer-group cust-default
4509 neighbor 10.3.1.1 description customer1
4510 neighbor 10.4.1.1 remote-as 64518
4511 neighbor 10.4.1.1 peer-group cust
4512 neighbor 10.4.1.1 description customer2
4513 neighbor 10.5.1.1 remote-as 64519
4514 neighbor 10.5.1.1 peer-group peer
4515 neighbor 10.5.1.1 description peer AS 1
4516 neighbor 10.6.1.1 remote-as 64520
4517 neighbor 10.6.1.1 peer-group peer
4518 neighbor 10.6.1.1 description peer AS 2
4520 address-family ipv4 unicast
4521 network 10.123.456.0/24
4522 network 10.123.456.128/25 route-map rm-no-export
4523 neighbor upstream route-map rm-upstream-out out
4524 neighbor cust route-map rm-cust-in in
4525 neighbor cust route-map rm-cust-out out
4526 neighbor cust send-community both
4527 neighbor peer route-map rm-peer-in in
4528 neighbor peer route-map rm-peer-out out
4529 neighbor peer send-community both
4530 neighbor 10.3.1.1 prefix-list pl-cust1-network in
4531 neighbor 10.4.1.1 prefix-list pl-cust2-network in
4532 neighbor 10.5.1.1 prefix-list pl-peer1-network in
4533 neighbor 10.6.1.1 prefix-list pl-peer2-network in
4536 ip prefix-list pl-default permit 0.0.0.0/0
4538 ip prefix-list pl-upstream-peers permit 10.1.1.1/32
4539 ip prefix-list pl-upstream-peers permit 10.2.1.1/32
4541 ip prefix-list pl-cust1-network permit 10.3.1.0/24
4542 ip prefix-list pl-cust1-network permit 10.3.2.0/24
4544 ip prefix-list pl-cust2-network permit 10.4.1.0/24
4546 ip prefix-list pl-peer1-network permit 10.5.1.0/24
4547 ip prefix-list pl-peer1-network permit 10.5.2.0/24
4548 ip prefix-list pl-peer1-network permit 192.168.0.0/24
4550 ip prefix-list pl-peer2-network permit 10.6.1.0/24
4551 ip prefix-list pl-peer2-network permit 10.6.2.0/24
4552 ip prefix-list pl-peer2-network permit 192.168.1.0/24
4553 ip prefix-list pl-peer2-network permit 192.168.2.0/24
4554 ip prefix-list pl-peer2-network permit 172.16.1/24
4556 bgp as-path access-list seq 5 asp-own-as permit ^$
4557 bgp as-path access-list seq 10 asp-own-as permit _64512_
4559 ! #################################################################
4560 ! Match communities we provide actions for, on routes receives from
4561 ! customers. Communities values of <our-ASN>:X, with X, have actions:
4563 ! 100 - blackhole the prefix
4564 ! 200 - set no_export
4565 ! 300 - advertise only to other customers
4566 ! 400 - advertise only to upstreams
4567 ! 500 - set no_export when advertising to upstreams
4568 ! 2X00 - set local_preference to X00
4570 ! blackhole the prefix of the route
4571 bgp community-list standard cm-blackhole permit 64512:100
4573 ! set no-export community before advertising
4574 bgp community-list standard cm-set-no-export permit 64512:200
4576 ! advertise only to other customers
4577 bgp community-list standard cm-cust-only permit 64512:300
4579 ! advertise only to upstreams
4580 bgp community-list standard cm-upstream-only permit 64512:400
4582 ! advertise to upstreams with no-export
4583 bgp community-list standard cm-upstream-noexport permit 64512:500
4585 ! set local-pref to least significant 3 digits of the community
4586 bgp community-list standard cm-prefmod-100 permit 64512:2100
4587 bgp community-list standard cm-prefmod-200 permit 64512:2200
4588 bgp community-list standard cm-prefmod-300 permit 64512:2300
4589 bgp community-list standard cm-prefmod-400 permit 64512:2400
4590 bgp community-list expanded cme-prefmod-range permit 64512:2...
4592 ! Informational communities
4594 ! 3000 - learned from upstream
4595 ! 3100 - learned from customer
4596 ! 3200 - learned from peer
4598 bgp community-list standard cm-learnt-upstream permit 64512:3000
4599 bgp community-list standard cm-learnt-cust permit 64512:3100
4600 bgp community-list standard cm-learnt-peer permit 64512:3200
4602 ! ###################################################################
4603 ! Utility route-maps
4605 ! These utility route-maps generally should not used to permit/deny
4606 ! routes, i.e. they do not have meaning as filters, and hence probably
4607 ! should be used with 'on-match next'. These all finish with an empty
4608 ! permit entry so as not interfere with processing in the caller.
4610 route-map rm-no-export permit 10
4611 set community additive no-export
4612 route-map rm-no-export permit 20
4614 route-map rm-blackhole permit 10
4615 description blackhole, up-pref and ensure it cannot escape this AS
4616 set ip next-hop 127.0.0.1
4617 set local-preference 10
4618 set community additive no-export
4619 route-map rm-blackhole permit 20
4621 ! Set local-pref as requested
4622 route-map rm-prefmod permit 10
4623 match community cm-prefmod-100
4624 set local-preference 100
4625 route-map rm-prefmod permit 20
4626 match community cm-prefmod-200
4627 set local-preference 200
4628 route-map rm-prefmod permit 30
4629 match community cm-prefmod-300
4630 set local-preference 300
4631 route-map rm-prefmod permit 40
4632 match community cm-prefmod-400
4633 set local-preference 400
4634 route-map rm-prefmod permit 50
4636 ! Community actions to take on receipt of route.
4637 route-map rm-community-in permit 10
4638 description check for blackholing, no point continuing if it matches.
4639 match community cm-blackhole
4641 route-map rm-community-in permit 20
4642 match community cm-set-no-export
4645 route-map rm-community-in permit 30
4646 match community cme-prefmod-range
4648 route-map rm-community-in permit 40
4650 ! #####################################################################
4651 ! Community actions to take when advertising a route.
4652 ! These are filtering route-maps,
4654 ! Deny customer routes to upstream with cust-only set.
4655 route-map rm-community-filt-to-upstream deny 10
4656 match community cm-learnt-cust
4657 match community cm-cust-only
4658 route-map rm-community-filt-to-upstream permit 20
4660 ! Deny customer routes to other customers with upstream-only set.
4661 route-map rm-community-filt-to-cust deny 10
4662 match community cm-learnt-cust
4663 match community cm-upstream-only
4664 route-map rm-community-filt-to-cust permit 20
4666 ! ###################################################################
4667 ! The top-level route-maps applied to sessions. Further entries could
4668 ! be added obviously..
4671 route-map rm-cust-in permit 10
4672 call rm-community-in
4674 route-map rm-cust-in permit 20
4675 set community additive 64512:3100
4676 route-map rm-cust-in permit 30
4678 route-map rm-cust-out permit 10
4679 call rm-community-filt-to-cust
4681 route-map rm-cust-out permit 20
4683 ! Upstream transit ASes
4684 route-map rm-upstream-out permit 10
4685 description filter customer prefixes which are marked cust-only
4686 call rm-community-filt-to-upstream
4688 route-map rm-upstream-out permit 20
4689 description only customer routes are provided to upstreams/peers
4690 match community cm-learnt-cust
4693 ! outbound policy is same as for upstream
4694 route-map rm-peer-out permit 10
4695 call rm-upstream-out
4697 route-map rm-peer-in permit 10
4698 set community additive 64512:3200
4701 Example of how to set up a 6-Bone connection.
4705 ! bgpd configuration
4706 ! ==================
4708 ! MP-BGP configuration
4711 bgp router-id 10.0.0.1
4712 neighbor 3ffe:1cfa:0:2:2a0:c9ff:fe9e:f56 remote-as `as-number`
4715 network 3ffe:506::/32
4716 neighbor 3ffe:1cfa:0:2:2a0:c9ff:fe9e:f56 activate
4717 neighbor 3ffe:1cfa:0:2:2a0:c9ff:fe9e:f56 route-map set-nexthop out
4718 neighbor 3ffe:1cfa:0:2:2c0:4fff:fe68:a231 remote-as `as-number`
4719 neighbor 3ffe:1cfa:0:2:2c0:4fff:fe68:a231 route-map set-nexthop out
4722 ipv6 access-list all permit any
4724 ! Set output nexthop address.
4726 route-map set-nexthop permit 10
4727 match ipv6 address all
4728 set ipv6 nexthop global 3ffe:1cfa:0:2:2c0:4fff:fe68:a225
4729 set ipv6 nexthop local fe80::2c0:4fff:fe68:a225
4738 TCP provides a mechanism for the user to specify the max segment size.
4739 setsockopt API is used to set the max segment size for TCP session. We
4740 can configure this as part of BGP neighbor configuration.
4742 This document explains how to avoid ICMP vulnerability issues by limiting
4743 TCP max segment size when you are using MTU discovery. Using MTU discovery
4744 on TCP paths is one method of avoiding BGP packet fragmentation.
4746 TCP negotiates a maximum segment size (MSS) value during session connection
4747 establishment between two peers. The MSS value negotiated is primarily based
4748 on the maximum transmission unit (MTU) of the interfaces to which the
4749 communicating peers are directly connected. However, due to variations in
4750 link MTU on the path taken by the TCP packets, some packets in the network
4751 that are well within the MSS value might be fragmented when the packet size
4752 exceeds the link's MTU.
4754 This feature is supported with TCP over IPv4 and TCP over IPv6.
4758 Below configuration can be done in router bgp mode and allows the user to
4759 configure the tcp-mss value per neighbor. The configuration gets applied
4760 only after hard reset is performed on that neighbor. If we configure tcp-mss
4761 on both the neighbors then both neighbors need to be reset.
4763 The configuration takes effect based on below rules, so there is a configured
4764 tcp-mss and a synced tcp-mss value per TCP session.
4766 By default if the configuration is not done then the TCP max segment size is
4767 set to the Maximum Transmission unit (MTU) – (IP/IP6 header size + TCP header
4768 size + ethernet header). For IPv4 its MTU – (20 bytes IP header + 20 bytes TCP
4769 header + 12 bytes ethernet header) and for IPv6 its MTU – (40 bytes IPv6 header
4770 + 20 bytes TCP header + 12 bytes ethernet header).
4772 If the config is done then it reduces 12-14 bytes for the ether header and
4773 uses it after synchronizing in TCP handshake.
4775 .. clicmd:: neighbor <A.B.C.D|X:X::X:X|WORD> tcp-mss (1-65535)
4777 When tcp-mss is configured kernel reduces 12-14 bytes for ethernet header.
4778 E.g. if tcp-mss is configured as 150 the synced value will be 138.
4780 Note: configured and synced value is different since TCP module will reduce
4781 12 bytes for ethernet header.
4788 frr# show running-config
4789 Building configuration...
4791 Current configuration:
4794 bgp router-id 192.0.2.1
4795 neighbor 198.51.100.2 remote-as 100
4796 neighbor 198.51.100.2 tcp-mss 150 => new entry
4797 neighbor 2001:DB8::2 remote-as 100
4798 neighbor 2001:DB8::2 tcp-mss 400 => new entry
4805 frr# show bgp neighbors 198.51.100.2
4806 BGP neighbor is 198.51.100.2, remote AS 100, local AS 100, internal link
4808 BGP version 4, remote router ID 192.0.2.2, local router ID 192.0.2.1
4809 BGP state = Established, up for 02:15:28
4810 Last read 00:00:28, Last write 00:00:28
4811 Hold time is 180, keepalive interval is 60 seconds
4812 Configured tcp-mss is 150, synced tcp-mss is 138 => new display
4816 frr# show bgp neighbors 2001:DB8::2
4817 BGP neighbor is 2001:DB8::2, remote AS 100, local AS 100, internal link
4819 BGP version 4, remote router ID 192.0.2.2, local router ID 192.0.2.1
4820 BGP state = Established, up for 02:16:34
4821 Last read 00:00:34, Last write 00:00:34
4822 Hold time is 180, keepalive interval is 60 seconds
4823 Configured tcp-mss is 400, synced tcp-mss is 388 => new display
4825 Show command json output:
4826 -------------------------
4830 frr# show bgp neighbors 2001:DB8::2 json
4835 "nbrInternalLink":true,
4838 "remoteRouterId":"192.0.2.2",
4839 "localRouterId":"192.0.2.1",
4840 "bgpState":"Established",
4841 "bgpTimerUpMsec":8349000,
4842 "bgpTimerUpString":"02:19:09",
4843 "bgpTimerUpEstablishedEpoch":1613054251,
4844 "bgpTimerLastRead":9000,
4845 "bgpTimerLastWrite":9000,
4846 "bgpInUpdateElapsedTimeMsecs":8347000,
4847 "bgpTimerHoldTimeMsecs":180000,
4848 "bgpTimerKeepAliveIntervalMsecs":60000,
4849 "bgpTcpMssConfigured":400, => new entry
4850 "bgpTcpMssSynced":388, => new entry
4854 frr# show bgp neighbors 198.51.100.2 json
4859 "nbrInternalLink":true,
4862 "remoteRouterId":"192.0.2.2",
4863 "localRouterId":"192.0.2.1",
4864 "bgpState":"Established",
4865 "bgpTimerUpMsec":8370000,
4866 "bgpTimerUpString":"02:19:30",
4867 "bgpTimerUpEstablishedEpoch":1613054251,
4868 "bgpTimerLastRead":30000,
4869 "bgpTimerLastWrite":30000,
4870 "bgpInUpdateElapsedTimeMsecs":8368000,
4871 "bgpTimerHoldTimeMsecs":180000,
4872 "bgpTimerKeepAliveIntervalMsecs":60000,
4873 "bgpTcpMssConfigured":150, => new entry
4874 "bgpTcpMssSynced":138, => new entry
4876 .. include:: routeserver.rst
4878 .. include:: rpki.rst
4880 .. include:: wecmp_linkbw.rst
4882 .. include:: flowspec.rst
4884 .. [#med-transitivity-rant] For some set of objects to have an order, there *must* be some binary ordering relation that is defined for *every* combination of those objects, and that relation *must* be transitive. I.e.:, if the relation operator is <, and if a < b and b < c then that relation must carry over and it *must* be that a < c for the objects to have an order. The ordering relation may allow for equality, i.e. a < b and b < a may both be true and imply that a and b are equal in the order and not distinguished by it, in which case the set has a partial order. Otherwise, if there is an order, all the objects have a distinct place in the order and the set has a total order)
4885 .. [bgp-route-osci-cond] McPherson, D. and Gill, V. and Walton, D., "Border Gateway Protocol (BGP) Persistent Route Oscillation Condition", IETF RFC3345
4886 .. [stable-flexible-ibgp] Flavel, A. and M. Roughan, "Stable and flexible iBGP", ACM SIGCOMM 2009
4887 .. [ibgp-correctness] Griffin, T. and G. Wilfong, "On the correctness of IBGP configuration", ACM SIGCOMM 2002
4889 .. _bgp-fast-convergence:
4891 BGP fast-convergence support
4892 ============================
4893 Whenever BGP peer address becomes unreachable we must bring down the BGP
4894 session immediately. Currently only single-hop EBGP sessions are brought
4895 down immediately.IBGP and multi-hop EBGP sessions wait for hold-timer
4896 expiry to bring down the sessions.
4898 This new configuration option helps user to teardown BGP sessions immediately
4899 whenever peer becomes unreachable.
4901 .. clicmd:: bgp fast-convergence
4903 This configuration is available at the bgp level. When enabled, configuration
4904 is applied to all the neighbors configured in that bgp instance.
4909 neighbor 10.0.0.2 remote-as 64496
4910 neighbor fd00::2 remote-as 64496
4911 bgp fast-convergence
4913 address-family ipv4 unicast
4917 address-family ipv6 unicast
4918 neighbor fd00::2 activate