.\" $OpenBSD: bgpd.conf.5,v 1.144 2016/06/28 16:59:14 jca Exp $ .\" .\" Copyright (c) 2004 Claudio Jeker .\" Copyright (c) 2003, 2004 Henning Brauer .\" Copyright (c) 2002 Daniel Hartmeier .\" .\" Permission to use, copy, modify, and distribute this software for any .\" purpose with or without fee is hereby granted, provided that the above .\" copyright notice and this permission notice appear in all copies. .\" .\" THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES .\" WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF .\" MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR .\" ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES .\" WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN .\" ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF .\" OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. .\" .Dd $Mdocdate: June 28 2016 $ .Dt BGPD.CONF 5 .Os .Sh NAME .Nm bgpd.conf .Nd Border Gateway Protocol daemon configuration file .Sh DESCRIPTION The .Xr bgpd 8 daemon implements the Border Gateway Protocol version 4 as described in RFC 4271. .Sh SECTIONS The .Nm config file is divided into five main sections. .Bl -tag -width xxxx .It Sy Macros User-defined variables may be defined and used later, simplifying the configuration file. .It Sy Global Configuration Global settings for .Xr bgpd 8 . .It Sy Routing Domain Configuration The definition and properties for BGP MPLS VPNs are set in this section. .It Sy Neighbors and Groups .Xr bgpd 8 establishes sessions with .Em neighbors . The neighbor definition and properties are set in this section, as well as grouping neighbors for the ease of configuration. .It Sy Filter Filter rules for incoming and outgoing .Em UPDATES . .El .Pp With the exception of macros, the sections should be grouped and appear in .Nm in the order shown above. .Pp The current line can be extended over multiple lines using a backslash .Pq Sq \e . Comments can be put anywhere in the file using a hash mark .Pq Sq # , and extend to the end of the current line. Care should be taken when commenting out multi-line text: the comment is effective until the end of the entire block. .Pp Argument names not beginning with a letter, digit, or underscore must be quoted. .Pp Additional configuration files can be included with the .Ic include keyword, for example: .Bd -literal -offset indent include "/etc/bgpd/bgpd-10.0.0.1.filter" .Ed .Sh MACROS Macros can be defined that will later be expanded in context. Macro names must start with a letter, digit, or underscore, and may contain any of those characters. Macro names may not be reserved words (for example, .Ic AS , .Ic neighbor , or .Ic group ) . Macros are not expanded inside quotes. .Pp For example: .Bd -literal -offset indent peer1="1.2.3.4" neighbor $peer1 { remote-as 65001 } .Ed .Sh GLOBAL CONFIGURATION There are quite a few settings that affect the operation of the .Xr bgpd 8 daemon globally. .Pp .Bl -tag -width Ds -compact .It Ic AS Ar as-number Op Ar as-number Set the local .Em autonomous system number to .Ar as-number . If the first AS number is a 4-byte AS it is possible to specify a secondary 2-byte AS number which is used for neighbors which do not support 4-byte AS numbers. The default for the secondary AS is 23456. .Pp The AS numbers are assigned by local RIRs, such as: .Pp .Bl -tag -width xxxxxxxx -compact .It AfriNIC for Africa .It APNIC for Asia Pacific .It ARIN for North America and parts of the Caribbean .It LACNIC for Latin America and the Caribbean .It RIPE NCC for Europe, the Middle East, and parts of Asia .El .Pp For example: .Bd -literal -offset indent AS 65001 .Ed .Pp sets the local AS to 65001. .Pp The AS numbers 64512 \(en 65534 are designated for private use. The AS number 23456 is a specially designated Autonomous System Number and should not be used. 4-byte AS numbers are specified as two numbers separated by a dot (ASDOT format), for example: .Bd -literal -offset indent AS 3.10 .Ed .Pp or as a large number (ASPLAIN format), for example: .Bd -literal -offset indent AS 196618 .Ed .Pp .It Ic connect-retry Ar seconds Set the number of seconds before retrying to open a connection. This timer should be sufficiently large in EBGP configurations. The default is 120 seconds. .Pp .It Xo .Ic dump .Op Ic rib Ar name .Pq Ic table Ns | Ns Ic table-mp Ns | Ns Ic table-v2 .Ar file Op Ar timeout .Xc .It Xo .Ic dump .Pq Ic all Ns | Ns Ic updates .Pq Ic in Ns | Ns Ic out .Ar file Op Ar timeout .Xc Dump the RIB, a.k.a. the .Em routing information base , and all BGP messages in Multi-threaded Routing Toolkit (MRT) format. It is possible to dump alternate RIB with the use of .Ar name . .Pp For example, the following will dump the entire table to the .Xr strftime 3 Ns -expanded filename. Only the .Ic table-v2 format is able to dump a multi-protocol RIB correctly. Both .Ic table and .Ic table-mp formats are more or less limited when handling multi-protocol entries and are only left around to support 3rd party tools not handling the new format. The timeout is optional: .Bd -literal -offset indent dump table "/tmp/rib-dump-%H%M" 300 .Ed .Pp Similar to the table dump, but this time all BGP messages and .Em state transitions will be dumped to the specified file: .Bd -literal -offset indent dump all in "/tmp/all-in-%H%M" 300 .Ed .Pp As before, but only the .Em UPDATE messages will be dumped to the file: .Bd -literal -offset indent dump updates in "/tmp/updates-in-%H%M" 300 .Ed .Pp It is also possible to dump outgoing messages: .Bd -literal -offset indent dump all out "/tmp/all-out-%H%M" 300 # or dump updates out "/tmp/updates-out-%H%M" 300 .Ed .Pp .It Ic fib-priority Ar prio Set the routing priority to .Ar prio . The default is 48. .Pp .It Xo .Ic fib-update .Pq Ic yes Ns | Ns Ic no .Xc If set to .Ic no , do not update the Forwarding Information Base, a.k.a. the kernel routing table. The default is .Ic yes . .Pp .It Ic holdtime Ar seconds Set the holdtime in seconds. The holdtime is reset to its initial value every time either a .Em KEEPALIVE or an .Em UPDATE message is received from the neighbor. If the holdtime expires the session is dropped. The default is 90 seconds. Neighboring systems negotiate the holdtime used when the connection is established in the .Em OPEN messages. Each neighbor announces its configured holdtime; the smaller one is then agreed upon. .Pp .It Ic holdtime min Ar seconds The minimal accepted holdtime in seconds. This value must be greater than or equal to 3. .Pp .It Ic listen on Ar address Specify the local IP address .Xr bgpd 8 should listen on. .Bd -literal -offset indent listen on 127.0.0.1 .Ed .Pp .It Ic log updates Log received and sent updates. .Pp .It Xo .Ic network .Ar address Ns Li / Ns Ar prefix .Op Ic set ...\& .Xc .It Xo .Ic network .Pq Ic inet Ns | Ns Ic inet6 .Ic static Op Ic set ...\& .Xc .It Xo .Ic network .Pq Ic inet Ns | Ns Ic inet6 .Ic connected Op Ic set ...\& .Xc Announce the specified network as belonging to our AS. If set to .Ic connected , routes to directly attached networks will be announced. If set to .Ic static , all static routes will be announced. .Bd -literal -offset indent network 192.168.7.0/24 .Ed .Pp It is possible to set default .Em AS path attributes per .Ic network statement: .Bd -literal -offset indent network 192.168.7.0/24 set localpref 220 .Ed .Pp See also the .Sx ATTRIBUTE SET section. .Pp .It Xo .Ic nexthop .Ic qualify .Ic via .Pq Ic bgp Ns | Ns Ic default .Xc If set to .Ic bgp , .Xr bgpd 8 may use BGP routes to verify nexthops. If set to .Ic default , bgpd may use the default route to verify nexthops. By default bgpd will only use static routes or routes added by other routing daemons like .Xr ospfd 8 . .Pp .It Xo .Ic rde .Ic med .Ic compare .Pq Ic always Ns | Ns Ic strict .Xc If set to .Ic always , the .Em MULTI_EXIT_DISC attributes will always be compared. The default is .Ic strict , where the metric is only compared between peers belonging to the same AS. .Pp .It Xo .Ic rde .Ic rib Ar name .Op Ic no Ic evaluate .Xc .It Xo .Ic rde .Ic rib Ar name .Op Ic rtable Ar number .Xc Create an additional RIB named .Ar name . It is possible to disable the decision process per RIB with the .Ic no Ic evaluate flag. If a .Ic rtable is specified, routes will be exported to the given kernel routing table. Currently the routing table must belong to the default routing domain and nexthop verification happens on table 0. Routes in the specified table will not be considered for nexthop verification. .Ic Adj-RIB-In and .Ic Loc-RIB are created automatically and used as default. .Pp .It Xo .Ic rde .Ic route-age .Pq Ic ignore Ns | Ns Ic evaluate .Xc If set to .Ic evaluate , the best path selection will not only be based on the path attributes but also on the age of the route, giving preference to the older, typically more stable, route. In this case the decision process is no longer deterministic. The default is .Ic ignore . .Pp .It Xo .Ic route-collector .Pq Ic yes Ns | Ns Ic no .Xc If set to .Ic yes , the route selection process is turned off. The default is .Ic no . .Pp .It Ic router-id Ar address Set the router ID to the given IP address, which must be local to the machine. .Bd -literal -offset indent router-id 10.0.0.1 .Ed .Pp If not given, the BGP ID is determined as the biggest IP address assigned to the local machine. .Pp .It Ic rtable Ar number Work with the given kernel routing table instead of the default table, .Ar 0 . Note that table 0 is used for nexthop verification. Routes in the specified table will not be considered for nexthop verification. This is the same as using the following syntax: .Bd -literal -offset indent rde rib Loc-RIB rtable number .Ed .Pp .It Ic socket Qo Ar path Qc Op Ic restricted Create a control socket at .Ar path . If .Ic restricted is specified a restricted control socket will be created. By default .Pa /var/run/bgpd.sock is used and no restricted socket is created. .Pp .It Xo .Ic transparent-as .Pq Ic yes Ns | Ns Ic no .Xc If set to .Ic yes , .Em AS paths to EBGP neighbors are not prepended with the local AS. The default is .Ic no . .El .Sh ROUTING DOMAIN CONFIGURATION .Xr bgpd 8 supports the setup and distribution of Virtual Private Networks. It is possible to import and export prefixes between routing domains. Each routing domain is specified by an .Ic rdomain section, which allows properties to be set specifically for that rdomain: .Bd -literal -offset indent rdomain 1 { descr "a rdomain" rd 65002:1 import-target rt 65002:42 export-target rt 65002:42 network 192.168.1/24 depend on mpe0 } .Ed .Pp There are several routing domain properties: .Pp .Bl -tag -width Ds -compact .It Ic depend on Ar interface Routes added to the rdomain will use this interface as the outgoing interface. Normally this will be an MPLS Provider Edge, .Xr mpe 4 , interface that is part of the rdomain. Local networks will be announced with the MPLS label specified on the interface. .Pp .It Ic descr Ar description Add a description. The description is used when logging but has no further meaning to .Xr bgpd 8 . .Pp .It Ic export-target Ar subtype Ar as-number : Ns Ar local .It Ic export-target Ar subtype Ar IP : Ns Ar local Specify an extended community which will be attached to announced networks. More than one .Ic export-target can be specified. See also the .Sx ATTRIBUTE SET section for further information about the encoding. The .Ar subtype should be set to .Ar rt for best compatibility with other implementations. .Pp .It Xo .Ic fib-update .Pq Ic yes Ns | Ns Ic no .Xc If set to .Ic no , do not update the Forwarding Information Base, a.k.a. the kernel routing table. The default is .Ic yes . .Pp .It Ic import-target Ar subtype Ar as-number : Ns Ar local .It Ic import-target Ar subtype Ar IP : Ns Ar local Only prefixes matching one of the specified .Ic import-targets will be imported into the rdomain. More than one .Ic import-target can be specified. See also the .Sx ATTRIBUTE SET section for further information about the encoding of extended communities. The .Ar subtype should be set to .Ar rt for best compatibility with other implementations. .Pp .It Ic network Ar arguments ... Define which networks should be exported into this VPN. See also the .Ic nexthop section in .Sx GLOBAL CONFIGURATION for further information about the arguments. .Pp .It Ic rd Ar as-number : Ns Ar local .It Ic rd Ar IP : Ns Ar local The sole purpose of the Route Distinguisher .Ic rd is to ensure that possible common prefixes are distinct between VPNs. The .Ic rd is neither used to identify the origin of the prefix nor to control into which VPNs the prefix is distributed to. The .Ar as-number or .Ar IP of a .Ic rd should be set to a number or IP that was assigned by an appropriate authority. Whereas .Ar local can be chosen by the local operator. .El .Sh NEIGHBORS AND GROUPS .Xr bgpd 8 establishes TCP connections to other BGP speakers called .Em neighbors . Each neighbor is specified by a .Ic neighbor section, which allows properties to be set specifically for that neighbor: .Bd -literal -offset indent neighbor 10.0.0.2 { remote-as 65002 descr "a neighbor" } .Ed .Pp Multiple neighbors can be grouped together by a .Ic group section. Each .Ic neighbor section within the .Ic group section inherits all properties from its group: .Bd -literal -offset indent group "peering AS65002" { remote-as 65002 neighbor 10.0.0.2 { descr "AS65002-p1" } neighbor 10.0.0.3 { descr "AS65002-p2" } } .Ed .Pp Instead of the neighbor's IP address, an address/netmask pair may be given: .Bd -literal -offset indent neighbor 10.0.0.0/8 .Ed .Pp In this case, the neighbor specification becomes a .Em template , and if a neighbor connects from an IP address within the given network, the template is .Em cloned , inheriting everything from the template but the remote address, which is replaced by the connecting neighbor's address. With a template specification it is valid to omit .Ic remote-as ; .Xr bgpd 8 will then accept any AS the neighbor presents in the .Em OPEN message. .Pp There are several neighbor properties: .Pp .Bl -tag -width Ds -compact .It Xo .Ic announce .Sm off .Pq Ic all | none | self | default-route .Sm on .Xc If set to .Ic none , no .Em UPDATE messages will be sent to the neighbor. If set to .Ic default-route , only the default route will be announced to the neighbor. If set to .Ic all , all generated .Em UPDATE messages will be sent to the neighbor. This is usually used for .Em transit AS's and .Em IBGP peers. The default value for .Em EBGP peers is .Ic self , which limits the sent .Em UPDATE messages to announcements of the local AS. The default for IBGP peers is .Ic all . .Pp .It Xo .Ic announce .Pq Ic IPv4 Ns | Ns Ic IPv6 .Pq Ic none Ns | Ns Ic unicast Ns | Ns Ic vpn .Xc For the given address family, control which subsequent address families (at the moment, only .Em none , which disables the announcement of that address family, .Em unicast , and .Em vpn , which allows the distribution of BGP MPLS VPNs, are supported) are announced during the capabilities negotiation. Only routes for that address family and subsequent address family will be announced and processed. .Pp .It Xo .Ic announce as-4byte .Pq Ic yes Ns | Ns Ic no .Xc If set to .Ic no , the 4-byte AS capability is not announced and so native 4-byte AS support is disabled. The default is .Ic yes . .Pp .It Xo .Ic announce capabilities .Pq Ic yes Ns | Ns Ic no .Xc If set to .Ic no , capability negotiation is disabled during the establishment of the session. This can be helpful to connect to old or broken BGP implementations. The default is .Ic yes . .Pp .It Xo .Ic announce refresh .Pq Ic yes Ns | Ns Ic no .Xc If set to .Ic no , the route refresh capability is not announced. The default is .Ic yes . .Pp .It Xo .Ic announce restart .Pq Ic yes Ns | Ns Ic no .Xc If set to .Ic no , the graceful restart capability is not announced. Currently only the End-of-RIB marker is supported and announced by the .Ic restart capability. The default is .Ic yes . .Pp .It Ic demote Ar group Increase the .Xr carp 4 demotion counter on the given interface group, usually .Ar carp , when the session is not in state .Em ESTABLISHED . The demotion counter will be increased as soon as .Xr bgpd 8 starts and decreased 60 seconds after the session went to state .Em ESTABLISHED . For neighbors added at runtime, the demotion counter is only increased after the session has been .Em ESTABLISHED at least once before dropping. .Pp For more information on interface groups, see the .Ic group keyword in .Xr ifconfig 8 . .Pp .It Ic depend on Ar interface The neighbor session will be kept in state .Em IDLE as long as .Ar interface reports no link. For .Xr carp 4 interfaces, no link means that the interface is currently .Em backup . This is primarily intended to be used with .Xr carp 4 to reduce failover times. .Pp The state of the network interfaces on the system can be viewed using the .Cm show interfaces command to .Xr bgpctl 8 . .Pp .It Ic descr Ar description Add a description. The description is used when logging neighbor events, in status reports, for specifying neighbors, etc., but has no further meaning to .Xr bgpd 8 . .Pp .It Ic down Do not start the session when bgpd comes up but stay in .Em IDLE . .Pp .It Xo .Ic dump .Pq Ic all Ns | Ns Ic updates .Pq Ic in Ns | Ns Ic out .Ar file Op Ar timeout .Xc Do a peer specific MRT dump. Peer specific dumps are limited to .Ic all and .Ic updates . See also the .Ic dump section in .Sx GLOBAL CONFIGURATION . .Pp .It Xo .Ic enforce neighbor-as .Pq Ic yes Ns | Ns Ic no .Xc If set to .Ic yes , .Em AS paths whose .Em leftmost AS is not equal to the .Em remote AS of the neighbor are rejected and a .Em NOTIFICATION is sent back. The default value for IBGP peers is .Ic no otherwise the default is .Ic yes . .Pp .It Ic holdtime Ar seconds Set the holdtime in seconds. Inherited from the global configuration if not given. .Pp .It Ic holdtime min Ar seconds Set the minimal acceptable holdtime. Inherited from the global configuration if not given. .Pp .It Xo .Ic ipsec .Pq Ic ah Ns | Ns Ic esp .Pq Ic in Ns | Ns Ic out .Ic spi Ar spi-number authspec Op Ar encspec .Xc Enable IPsec with static keying. There must be at least two .Ic ipsec statements per peer with manual keying, one per direction. .Ar authspec specifies the authentication algorithm and key. It can be .Bd -literal -offset indent sha1 md5 .Ed .Pp .Ar encspec specifies the encryption algorithm and key. .Ic ah does not support encryption. With .Ic esp , encryption is optional. .Ar encspec can be .Bd -literal -offset indent 3des 3des-cbc aes aes-128-cbc .Ed .Pp Keys must be given in hexadecimal format. .Pp .It Xo .Ic ipsec .Pq Ic ah Ns | Ns Ic esp .Ic ike .Xc Enable IPsec with dynamic keying. In this mode, .Xr bgpd 8 sets up the flows, and a key management daemon such as .Xr isakmpd 8 is responsible for managing the session keys. With .Xr isakmpd 8 , it is sufficient to copy the peer's public key, found in .Pa /etc/isakmpd/local.pub , to the local machine. It must be stored in a file named after the peer's IP address and must be stored in .Pa /etc/isakmpd/pubkeys/ipv4/ . The local public key must be copied to the peer in the same way. As .Xr bgpd 8 manages the flows on its own, it is sufficient to restrict .Xr isakmpd 8 to only take care of keying by specifying the flags .Fl Ka . This can be done in .Xr rc.conf.local 8 . After starting the .Xr isakmpd 8 and .Xr bgpd 8 daemons on both sides, the session should be established. .Pp .It Ic local-address Ar address When .Xr bgpd 8 initiates the TCP connection to the neighbor system, it normally does not bind to a specific IP address. If a .Ic local-address is given, .Xr bgpd 8 binds to this address first. .Pp .It Ic log no Disable neighbor specific logging. .Pp .It Ic log updates Log received and sent updates for this neighbor. .Pp .It Xo .Ic max-prefix Ar number .Op Ic restart Ar number .Xc Terminate the session after .Ar number prefixes have been received (no such limit is imposed by default). If .Ic restart is specified, the session will be restarted after .Ar number minutes. .Pp .It Ic multihop Ar hops Neighbors not in the same AS as the local .Xr bgpd 8 normally have to be directly connected to the local machine. If this is not the case, the .Ic multihop statement defines the maximum hops the neighbor may be away. .Pp .It Ic passive Do not attempt to actively open a TCP connection to the neighbor system. .Pp .It Ic remote-as Ar as-number Set the AS number of the remote system. .Pp .It Ic rib Ar name Bind the neighbor to the specified RIB. .Pp .It Ic route-reflector Op Ar address Act as an RFC 4456 .Em route-reflector for this neighbor. An optional cluster ID can be specified; otherwise the BGP ID will be used. .Pp .It Ic set Ar attribute ... Set the .Em AS path attributes to some default per .Ic neighbor or .Ic group block: .Bd -literal -offset indent set localpref 300 .Ed .Pp See also the .Sx ATTRIBUTE SET section. Set parameters are applied to the received prefixes; the only exceptions are .Ic prepend-self , .Ic nexthop no-modify and .Ic nexthop self . These sets are rewritten into filter rules and can be viewed with .Dq bgpd -nv . .Pp .It Xo .Ic softreconfig .Pq Ic in Ns | Ns Ic out .Pq Ic yes Ns | Ns Ic no .Xc Turn soft reconfiguration on or off for the specified direction. If soft reconfiguration is turned on, filter changes will be applied on configuration reloads. If turned off, a BGP session needs to be cleared to apply the filter changes. Enabling .Ic softreconfig in will raise the memory requirements of .Xr bgpd 8 because the unmodified .Em AS path attributes need to be stored as well. The default is .Ic yes . .Pp .It Ic tcp md5sig password Ar secret .It Ic tcp md5sig key Ar secret Enable TCP MD5 signatures per RFC 2385. The shared secret can either be given as a password or hexadecimal key. .Bd -literal -offset indent tcp md5sig password mekmitasdigoat tcp md5sig key deadbeef .Ed .Pp .It Xo .Ic transparent-as .Pq Ic yes Ns | Ns Ic no .Xc If set to .Ic yes , .Em AS paths to EBGP neighbors are not prepended with the local AS. The default is inherited from the global .Ic transparent-as setting. .Pp .It Xo .Ic ttl-security .Pq Ic yes Ns | Ns Ic no .Xc Enable or disable ttl-security. When enabled, outgoing packets are sent using a TTL of 255 and a check is made against an incoming packet's TTL. For directly connected peers, incoming packets are required to have a TTL of 255, ensuring they have not been routed. For multihop peers, incoming packets are required to have a TTL of 256 minus multihop distance, ensuring they have not passed through more than the expected number of hops. The default is .Ic no . .El .Sh FILTER .Xr bgpd 8 has the ability to .Ic allow and .Ic deny .Em UPDATES based on .Em prefix or .Em AS path attributes . In addition, .Em UPDATES may also be modified by filter rules. .Pp For each .Em UPDATE processed by the filter, the filter rules are evaluated in sequential order, from first to last. The last matching .Ic allow or .Ic deny rule decides what action is taken. .Pp The following actions can be used in the filter: .Bl -tag -width xxxxxxxx .It Ic allow The .Em UPDATE is passed. .It Ic deny The .Em UPDATE is blocked. .It Ic match Apply the filter attribute set without influencing the filter decision. .El .Sh PARAMETERS The rule parameters specify the .Em UPDATES to which a rule applies. An .Em UPDATE always comes from, or goes to, one neighbor. Most parameters are optional, but each can appear at most once per rule. If a parameter is specified, the rule only applies to packets with matching attributes. .Pp .Bl -tag -width Ds -compact .It Xo .Ar as-type Op Ar operator .Ar as-number .Xc This rule applies only to .Em UPDATES where the .Em AS path matches. The .Ar as-number is matched against a part of the .Em AS path specified by the .Ar as-type : .Pp .Bl -tag -width transmit-as -compact .It Ic AS (any part) .It Ic peer-as (leftmost AS number) .It Ic source-as (rightmost AS number) .It Ic transit-as (all but the rightmost AS number) .El .Pp .Ar as-number is an AS number as explained above under .Sx GLOBAL CONFIGURATION . It may be set to .Ic neighbor-as , which is expanded to the current neighbor remote AS number. .Pp The .Ar operator can be unspecified (this case is identical to the equality operator), or one of the numerical operators .Bd -literal -offset indent = (equal) != (unequal) - (range including boundaries) >< (except range) .Ed .Pp >< and - are binary operators (they take two arguments); with these, .Ar as-number cannot be set to .Ic neighbor-as . .Pp Multiple .Ar as-number entries for a given type or .Ar as-type as-number entries may also be specified, separated by commas or whitespace, if enclosed in curly brackets: .Bd -literal -offset indent deny from any AS { 1, 2, 3 } deny from any { AS 1, source-as 2, transit-as 3 } deny from any { AS { 1, 2, 3 }, source-as 4, transit-as 5 } .Ed .Pp .It Xo .Ic community .Ar as-number Ns Li \&: Ns Ar local .Xc .It Ic community Ar name This rule applies only to .Em UPDATES where the .Ic community path attribute is present and matches. Communities are specified as .Ar as-number : Ns Ar local , where .Ar as-number is an AS number and .Ar local is a locally significant number between zero and .Li 65535 . Both .Ar as-number and .Ar local may be set to .Sq * to do wildcard matching. Alternatively, well-known communities may be given by name instead and include .Ic BLACKHOLE , .Ic NO_EXPORT , .Ic NO_ADVERTISE , .Ic NO_EXPORT_SUBCONFED , and .Ic NO_PEER . Both .Ar as-number and .Ar local may be set to .Ic neighbor-as , which is expanded to the current neighbor remote AS number. .Pp .It Xo .Ic ext-community .Ar subtype Ar as-number : Ns Ar local .Xc .It Xo .Ic ext-community .Ar subtype Ar IP : Ns Ar local .Xc .It Xo .Ic ext-community .Ar subtype Ar numvalue .Xc This rule applies only to .Em UPDATES where the .Em extended community path attribute is present and matches. Extended Communities are specified by a .Ar subtype and normally two values, a globally unique part (e.g. the AS number) and a local part. See also the .Sx ATTRIBUTE SET section for further information about the encoding. .Pp .It Xo .Pq Ic from Ns | Ns Ic to .Ar peer .Xc This rule applies only to .Em UPDATES coming from, or going to, this particular neighbor. This parameter must be specified. .Ar peer is one of the following: .Pp .Bl -tag -width "group descr" -compact .It Ic any Any neighbor will be matched. .It Ar address Neighbors with this address will be matched. .It Ic group Ar descr Neighbors in this group will be matched. .It Ic AS Ar as-number Neighbors with this AS will be matched. .El .Pp Multiple .Ar peer entries may also be specified, separated by commas or whitespace, if enclosed in curly brackets: .Bd -literal -offset indent deny from { 128.251.16.1, 251.128.16.2, group hojo } .Ed .Pp .It Pq Ic inet Ns | Ns Ic inet6 Match only routes in the IPv4 or IPv6 address families, respectively. .Ic inet is an alias for .Qq prefix 0.0.0.0/0 prefixlen >= 0 ; .Ic inet6 is an alias for .Qq prefix ::/0 prefixlen >= 0 . .Pp .It Ic max-as-len Ar len This rule applies only to .Em UPDATES where the .Em AS path has more than .Ar len elements. .Pp .It Ic max-as-seq Ar len This rule applies only to .Em UPDATES where a single .Em AS number is repeated more than .Ar len times. .Pp .It Ic nexthop Ar address This rule applies only to .Em UPDATES where the nexthop is equal to .Ar address . The .Ar address can be set to .Em neighbor in which case the nexthop is compared against the address of the neighbor. Nexthop filtering is not supported on locally announced networks and one must take into consideration previous rules overwriting nexthops. .Pp .It Ic prefix Ar address Ns Li / Ns Ar len .It Ic prefix Ar address Ns Li / Ns Ar len Ic prefixlen Ar range .It Ic prefix Ar address Ns Li / Ns Ar len Ic or-longer This rule applies only to .Em UPDATES for the specified prefix. .Pp Multiple entries may be specified, separated by commas or whitespace, if enclosed in curly brackets: .Bd -literal -offset indent deny from any prefix { 192.168.0.0/16, 10.0.0.0/8 or-longer } .Ed .Pp Multiple lists can also be specified, which is useful for macro expansion: .Bd -literal -offset indent good="{ 192.168.0.0/16, 172.16.0.0/12, 10.0.0.0/8 }" bad="{ 224.0.0.0/4 prefixlen >= 4, 240.0.0.0/4 prefixlen >= 4 }" ugly="{ 127.0.0.1/8, 169.254.0.0/16 }" deny from any prefix { $good $bad $ugly } .Ed .Pp Prefix length ranges are specified by using these operators: .Bd -literal -offset indent = (equal) != (unequal) < (less than) <= (less than or equal) > (greater than) >= (greater than or equal) - (range including boundaries) >< (except range) .Ed .Pp >< and - are binary operators (they take two arguments). For instance, to match all prefix lengths >= 8 and <= 12, and hence the CIDR netmasks 8, 9, 10, 11 and 12: .Bd -literal -offset indent prefixlen 8-12 .Ed .Pp Or, to match all prefix lengths < 8 or > 12, and hence the CIDR netmasks 0\(en7 and 13\(en32: .Bd -literal -offset indent prefixlen 8><12 .Ed .Pp This will match all prefixes in the 10.0.0.0/8 netblock with netmasks longer than 16: .Bd -literal -offset indent prefix 10.0.0.0/8 prefixlen > 16 .Ed .Pp .Ic or-longer is a shorthand for: .Bd -literal -offset indent .Ic prefix Ar address Ns Li / Ns Ar len Ic prexiflen Ic >= Ar len .Ed .Pp .It Ic quick If an .Em UPDATE matches a rule which has the .Ic quick option set, this rule is considered the last matching rule, and evaluation of subsequent rules is skipped. .Pp .It Ic rib Ar name Apply rule only to the specified RIB. This only applies for received updates, so not for rules using the .Ar to peer parameter. .Pp .It Ic set Ar attribute ... All matching rules can set the .Em AS path attributes to some default. The set of every matching rule is applied, not only the last matching one. See also the following section. .El .Sh ATTRIBUTE SET .Em AS path attributes can be modified with .Ic set . .Pp .Ic set can be used on .Ic network statements, in .Ic neighbor or .Ic group blocks, and on filter rules. Attribute sets can be expressed as lists. .Pp The following attributes can be modified: .Pp .Bl -tag -width Ds -compact .It Xo .Ic community Op Ar delete .Ar as-number : Ns Ar local .Xc .It Xo .Ic community Op Ar delete .Ar name .Xc Set or delete the .Em COMMUNITIES AS path attribute. Communities are specified as .Ar as-number : Ns Ar local , where .Ar as-number is an AS number and .Ar local is a locally-significant number between zero and .Li 65535 . Alternately, well-known communities may be specified by name: .Ic NO_EXPORT , .Ic NO_ADVERTISE , .Ic NO_EXPORT_SUBCONFED , or .Ic NO_PEER . For .Cm delete , both .Ar as-number and .Ar local may be set to .Sq * to do wildcard matching. .Pp .It Xo .Ic ext-community Op Ar delete .Ar subtype Ar as-number : Ns Ar local .Xc .It Xo .Ic ext-community Op Ar delete .Ar subtype Ar IP : Ns Ar local .Xc .It Xo .Ic ext-community Op Ar delete .Ar subtype Ar numvalue .Xc Set or delete the .Em Extended Community AS path attribute. Extended Communities are specified by a .Ar subtype and normally two values, a globally unique part (e.g. the AS number) and a local part. The type is selected depending on the encoding of the global part. Two-octet AS Specific Extended Communities and Four-octet AS Specific Extended Communities are encoded as .Ar as-number : Ns Ar local . Four-octet encoding is used if the .Ar as-number is bigger than 65535 or if the AS_DOT encoding is used. IPv4 Address Specific Extended Communities are encoded as .Ar IP : Ns Ar local . Opaque Extended Communities are encoded with a single numeric value. Currently the following subtypes are supported: .Bd -literal -offset indent rt Route Target soo Source of Origin odi OSPF Domain Identifier ort OSPF Route Type ori OSPF Router ID bdc BGP Data Collection .Ed .Pp Not all type and subtype value pairs are allowed by IANA and the parser will ensure that no invalid combination is created. .Pp .It Ic localpref Ar number Set the .Em LOCAL_PREF AS path attribute. If .Ar number starts with a plus or minus sign, .Em LOCAL_PREF will be adjusted by adding or subtracting .Ar number ; otherwise it will be set to .Ar number . The default is 100. .Pp .It Ic med Ar number .It Ic metric Ar number Set the .Em MULTI_EXIT_DISC AS path attribute. If .Ar number starts with a plus or minus sign, .Em MULTI_EXIT_DISC will be adjusted by adding or subtracting .Ar number ; otherwise it will be set to .Ar number . .Pp .It Xo .Ic origin .Sm off .Pq Ic igp | egp | incomplete .Sm on .Xc Set the .Em ORIGIN AS path attribute to mark the source of this route as being injected from an igp protocol, an egp protocol or being an aggregated route. .Pp .It Xo .Ic nexthop .Sm off .Pq Ar address | Ic blackhole | reject | self | no-modify .Sm on .Xc Set the .Em NEXTHOP AS path attribute to a different nexthop address or use blackhole or reject routes. If set to .Em no-modify , the nexthop attribute is not modified. Unless set to .Em self , the nexthop is left unmodified for IBGP sessions. .Em self forces the nexthop to be set to the local interface address. .Bd -literal -offset indent set nexthop 192.168.0.1 set nexthop blackhole set nexthop reject set nexthop no-modify set nexthop self .Ed .Pp .It Ic pftable Ar table Add the prefix in the update to the specified .Xr pf 4 table, regardless of whether or not the path was selected for routing. This option may be useful in building realtime blacklists. .Pp .It Ic prepend-neighbor Ar number Prepend the neighbor's AS .Ar number times to the .Em AS path . .Pp .It Ic prepend-self Ar number Prepend the local AS .Ar number times to the .Em AS path . .Pp .It Ic rtlabel Ar label Add the prefix to the kernel routing table with the specified .Ar label . .Pp .It Ic weight Ar number The .Em weight is used to tip prefixes with equally long AS paths in one or the other direction. A prefix is weighed at a very late stage in the decision process. If .Ar number starts with a plus or minus sign, the .Em weight will be adjusted by adding or subtracting .Ar number ; otherwise it will be set to .Ar number . .Em Weight is a local non-transitive attribute and a bgpd-specific extension. For prefixes with equally long paths, the prefix with the larger weight is selected. .El .Sh FILES .Bl -tag -width "/etc/bgpd.conf" -compact .It Pa /etc/bgpd.conf .Xr bgpd 8 configuration file .El .Sh SEE ALSO .Xr strftime 3 , .Xr ipsec 4 , .Xr pf 4 , .Xr tcp 4 , .Xr bgpctl 8 , .Xr bgpd 8 , .Xr ipsecctl 8 , .Xr isakmpd 8 , .Xr rc.conf.local 8 .Sh HISTORY The .Nm file format first appeared in .Ox 3.5 .