bird/doc/bird.sgml

<!doctype birddoc system>

<!--
	BIRD 3.0 documentation

This documentation can have 4 forms: sgml (this is master copy), html, ASCII
text and dvi/postscript (generated from sgml using sgmltools). You should always
edit master copy.

This is a slightly modified linuxdoc dtd. Anything in <descrip> tags is
considered definition of configuration primitives, <cf> is fragment of
configuration within normal text, <m> is "meta" information within fragment of
configuration - something in config which is not keyword.

    (set-fill-column 80)

    Copyright 1999 - 2022 CZ.NIC, z.s.p.o , distribute under GPL version 2 or later.

 -->

<book>

<title>BIRD 3.0 User's Guide
<author>
Ondrej Filip <it/&lt;feela@network.cz&gt;/,
Martin Mares <it/&lt;mj@ucw.cz&gt;/,
Maria Matejka <it/&lt;mq@jmq.cz&gt;/,
Ondrej Zajicek <it/&lt;santiago@crfreenet.org&gt;/
</author>

<abstract>
This document contains user documentation for the BIRD Internet Routing Daemon project.
</abstract>

<!-- Table of contents -->
<toc>

<!-- Begin the document -->


<chapt>Introduction
<label id="intro">

<sect>What is BIRD
<label id="what-is-bird">

<p>The name `BIRD' is actually an acronym standing for `BIRD Internet Routing
Daemon'. Let's take a closer look at the meaning of the name:

<p><em/BIRD/: Well, we think we have already explained that. It's an acronym
standing for `BIRD Internet Routing Daemon', you remember, don't you? :-)

<p><em/Internet Routing/: It's a program (well, a daemon, as you are going to
discover in a moment) which works as a dynamic router in an Internet type
network (that is, in a network running either the IPv4 or the IPv6 protocol).
Routers are devices which forward packets between interconnected networks in
order to allow hosts not connected directly to the same local area network to
communicate with each other. They also communicate with the other routers in the
Internet to discover the topology of the network which allows them to find
optimal (in terms of some metric) rules for forwarding of packets (which are
called routing tables) and to adapt themselves to the changing conditions such
as outages of network links, building of new connections and so on. Most of
these routers are costly dedicated devices running obscure firmware which is
hard to configure and not open to any changes (on the other hand, their special
hardware design allows them to keep up with lots of high-speed network
interfaces, better than general-purpose computer does). Fortunately, most
operating systems of the UNIX family allow an ordinary computer to act as a
router and forward packets belonging to the other hosts, but only according to a
statically configured table.

<p>A <em/Routing Daemon/ is in UNIX terminology a non-interactive program
running on background which does the dynamic part of Internet routing, that is
it communicates with the other routers, calculates routing tables and sends them
to the OS kernel which does the actual packet forwarding. There already exist
other such routing daemons: routed (RIP only), GateD (non-free),
<HTMLURL URL="http://www.zebra.org" name="Zebra"> and
<HTMLURL URL="http://sourceforge.net/projects/mrt" name="MRTD">,
but their capabilities are limited and they are relatively hard to configure
and maintain.

<p>BIRD is an Internet Routing Daemon designed to avoid all of these shortcomings,
to support all the routing technology used in the today's Internet or planned to
be used in near future and to have a clean extensible architecture allowing new
routing protocols to be incorporated easily. Among other features, BIRD
supports:

<itemize>
	<item>both IPv4 and IPv6 protocols
	<item>multiple routing tables
	<item>the Border Gateway Protocol (BGPv4)
	<item>the Routing Information Protocol (RIPv2, RIPng)
	<item>the Open Shortest Path First protocol (OSPFv2, OSPFv3)
	<item>the Babel Routing Protocol
	<item>the Router Advertisements for IPv6 hosts
	<item>a virtual protocol for exchange of routes between different
		routing tables on a single host
	<item>a command-line interface allowing on-line control and inspection
		of status of the daemon
	<item>soft reconfiguration (no need to use complex online commands to
		change the configuration, just edit the configuration file and
		notify BIRD to re-read it and it will smoothly switch itself to
		the new configuration, not disturbing routing protocols unless
		they are affected by the configuration changes)
	<item>a powerful language for route filtering
</itemize>

<p>BIRD has been developed at the Faculty of Math and Physics, Charles
University, Prague, Czech Republic as a student project. It can be freely
distributed under the terms of the GNU General Public License.

<p>BIRD has been designed to work on all UNIX-like systems. It has been
developed and tested under Linux 2.0 to 2.6, and then ported to FreeBSD, NetBSD
and OpenBSD, porting to other systems (even non-UNIX ones) should be relatively
easy due to its highly modular architecture.

<p>BIRD 1.x supported either IPv4 or IPv6 protocol, but had to be compiled separately
for each one. BIRD~2 supports both of them with a possibility of further extension.
BIRD~2 supports Linux at least 3.16, FreeBSD 10, NetBSD 7.0, and OpenBSD 5.8.
Anyway, it will probably work well also on older systems.

<sect>Installing BIRD
<label id="install">

<p>On a recent UNIX system with GNU development tools (GCC, binutils, m4, make)
and Perl, installing BIRD should be as easy as:

<code>
	./configure
	make
	make install
	vi /usr/local/etc/bird.conf
	bird
</code>

<p>You can use <tt>./configure --help</tt> to get a list of configure
options. The most important ones are: <tt/--with-protocols=/ to produce a slightly smaller
BIRD executable by configuring out routing protocols you don't use, and
<tt/--prefix=/ to install BIRD to a place different from <file>/usr/local</file>.


<sect>Running BIRD
<label id="argv">

<p>You can pass several command-line options to bird:

<descrip>
	<tag><label id="argv-config">-c <m/config name/</tag>
	use given configuration file instead of <it/prefix/<file>/etc/bird.conf</file>.

	<tag><label id="argv-debug">-d</tag>
	enable debug messages to stderr, and run bird in foreground.

	<tag><label id="argv-debug-file">-D <m/filename of debug log/</tag>
	enable debug messages to given file.

	<tag><label id="argv-foreground">-f</tag>
	run bird in foreground.

	<tag><label id="argv-group">-g <m/group/</tag>
	use that group ID, see the next section for details.

	<tag><label id="argv-help">-h, --help</tag>
	display command-line options to bird.

	<tag><label id="argv-local">-l</tag>
	look for a configuration file and a communication socket in the current
	working directory instead of in default system locations. However, paths
	specified by options <cf/-c/, <cf/-s/ have higher priority.

	<tag><label id="argv-parse">-p</tag>
	just parse the config file and exit. Return value is zero if the config
	file is valid, nonzero if there are some errors.

	<tag><label id="argv-pid">-P <m/name of PID file/</tag>
	create a PID file with given filename.

	<tag><label id="argv-recovery">-R</tag>
	apply graceful restart recovery after start.

	<tag><label id="argv-socket">-s <m/name of communication socket/</tag>
	use given filename for a socket for communications with the client,
	default is <it/prefix/<file>/var/run/bird.ctl</file>.

	<tag><label id="argv-user">-u <m/user/</tag>
	drop privileges and use that user ID, see the next section for details.

	<tag><label id="argv-version">--version</tag>
	display bird version.
</descrip>

<p>BIRD writes messages about its work to log files or syslog (according to config).


<sect>Privileges
<label id="privileges">

<p>BIRD, as a routing daemon, uses several privileged operations (like setting
routing table and using raw sockets). Traditionally, BIRD is executed and runs
with root privileges, which may be prone to security problems. The recommended
way is to use a privilege restriction (options <cf/-u/, <cf/-g/). In that case
BIRD is executed with root privileges, but it changes its user and group ID to
an unprivileged ones, while using Linux capabilities to retain just required
privileges (capabilities CAP_NET_*). Note that the control socket is created
before the privileges are dropped, but the config file is read after that. The
privilege restriction is not implemented in BSD port of BIRD.

<p>An unprivileged user (as an argument to <cf/-u/ options) may be the user
<cf/nobody/, but it is suggested to use a new dedicated user account (like
<cf/bird/). The similar considerations apply for the group option, but there is
one more condition -- the users in the same group can use <file/birdc/ to
control BIRD.

<p>Finally, there is a possibility to use external tools to run BIRD in an
environment with restricted privileges. This may need some configuration, but it
is generally easy -- BIRD needs just the standard library, privileges to read
the config file and create the control socket and the CAP_NET_* capabilities.


<chapt>Architecture
<label id="architecture">

<sect>Routing tables
<label id="routing-tables">

<p>The heart of BIRD is a routing table. BIRD has several independent routing tables;
each of them contains routes of exactly one <m/nettype/ (see below). There are two
default tables -- <cf/master4/ for IPv4 routes and <cf/master6/ for IPv6 routes.
Other tables must be explicitly configured.

<p>
These routing tables are not kernel forwarding tables. No forwarding is done by
BIRD. If you want to forward packets using the routes in BIRD tables, you may
use the Kernel protocol (see below) to synchronize them with kernel FIBs.

<p>
Every nettype defines a (kind of) primary key on routes. Every route source can
supply one route for every possible primary key; new route announcement replaces
the old route from the same source, keeping other routes intact. BIRD always
chooses the best route for each primary key among the known routes and keeps the
others as suboptimal. When the best route is retracted, BIRD re-runs the best
route selection algorithm to find the current best route.

<p>
The global best route selection algorithm is (roughly) as follows:

<itemize>
	<item>Preferences of the routes are compared.
	<item>Source protocol instance preferences are compared.
	<item>If source protocols are the same (e.g. BGP vs. BGP), the protocol's route selection algorithm is invoked.
	<item>If source protocols are different (e.g. BGP vs. OSPF), result of the algorithm is undefined.
</itemize>

<p><label id="dsc-table-sorted">Usually, a routing table just chooses a selected
route from a list of entries for one network. Optionally, these lists of entries
are kept completely sorted (according to preference or some protocol-dependent
metric). See <ref id="rtable-sorted" name="sorted"> table option for details.

<sect>Routes and network types
<label id="routes">

<p>BIRD works with several types of routes. Some of them are typical IP routes,
others are better described as forwarding rules. We call them all routes,
regardless of this difference.

<p>Every route consists of several attributes (read more about them in the
<ref id="route-attributes" name="Route attributes"> section); the common for all
routes are:

<itemize>
	<item>IP address of router which told us about this route
	<item>Source protocol instance
	<item>Route preference
	<item>Optional attributes defined by protocols
</itemize>

<p>Other attributes depend on nettypes. Some of them are part of the primary key, these are marked (PK).

<sect1>IPv4 and IPv6 routes
<label id="ip-routes">

<p>The traditional routes. Configuration keywords are <cf/ipv4/ and <cf/ipv6/.

<itemize>
	<item>(PK) Route destination (IP prefix together with its length)
	<item>Route next hops (see below)
</itemize>

<sect1>IPv6 source-specific routes
<label id="ip-sadr-routes">

<p>The IPv6 routes containing both destination and source prefix. They are used
for source-specific routing (SSR), also called source-address dependent routing
(SADR), see <rfc id="8043">. Currently limited mostly to the Babel protocol.
Configuration keyword is <cf/ipv6 sadr/.

<itemize>
	<item>(PK) Route destination (IP prefix together with its length)
	<item>(PK) Route source (IP prefix together with its length)
	<item>Route next hops (see below)
</itemize>

<sect1>VPN IPv4 and IPv6 routes
<label id="vpn-routes">

<p>Routes for IPv4 and IPv6 with VPN Route Distinguisher (<rfc id="4364">).
Configuration keywords are <cf/vpn4/ and <cf/vpn6/.

<itemize>
	<item>(PK) Route destination (IP prefix together with its length)
	<item>(PK) Route distinguisher (according to <rfc id="4364">)
	<item>Route next hops
</itemize>

<sect1>Route Origin Authorization for IPv4 and IPv6
<label id="roa-routes">

<p>These entries can be used to validate route origination of BGP routes.
A ROA entry specifies prefixes which could be originated by an AS number.
Their keywords are <cf/roa4/ and <cf/roa6/.

<itemize>
	<item>(PK) IP prefix together with its length
	<item>(PK) Matching prefix maximal length
	<item>(PK) AS number
</itemize>

<sect1>Flowspec for IPv4 and IPv6
<label id="flow-routes">

<p>Flowspec rules are a form of firewall and traffic flow control rules
distributed mostly via BGP. These rules may help the operators stop various
network attacks in the beginning before eating up the whole bandwidth.
Configuration keywords are <cf/flow4/ and <cf/flow6/.

<itemize>
	<item>(PK) IP prefix together with its length
	<item>(PK) Flow definition data
	<item>Flow action (encoded internally as BGP communities according to <rfc id="5575">)
</itemize>

<sect1>MPLS switching rules
<label id="mpls-routes">

<p>MPLS routes control MPLS forwarding in the same way as IP routes control IP
forwarding. MPLS-aware routing protocols produce both labeled IP routes and
corresponding MPLS routes. Configuration keyword is <cf/mpls/.

<itemize>
	<item>(PK) MPLS label
	<item>Route next hops
</itemize>

<sect1>Route next hops
<label id="route-next-hop">

<p>This is not a nettype. The route next hop is a complex attribute common for many
nettypes as you can see before. Every next hop has its assigned device
(either assumed from its IP address or set explicitly). It may have also
an IP address and an MPLS stack (one or both independently).
Maximal MPLS stack depth is set (in compile time) to 8 labels.

<p>Every route (when eligible to have a next hop) can have more than one next hop.
In that case, every next hop has also its weight.

<sect>Protocols and channels
<label id="protocols-concept">

<p>BIRD protocol is an abstract class of producers and consumers of the routes.
Each protocol may run in multiple instances and bind on one side to route
tables via channels, on the other side to specified listen sockets (BGP),
interfaces (Babel, OSPF, RIP), APIs (Kernel, Direct), or nothing (Static, Pipe).

<p>There are also two protocols that do not have any channels -- BFD and Device.
Both of them are kind of service for other protocols.

<p>Each protocol is connected to a routing table through a channel. Some protocols
support only one channel (OSPF, RIP), some protocols support more channels (BGP, Direct).
Each channel has two filters which can accept, reject and modify the routes.
An <it/export/ filter is applied to routes passed from the routing table to the protocol,
an <it/import/ filter is applied to routes in the opposite direction.

<sect>Graceful restart
<label id="graceful-restart">

<p>When BIRD is started after restart or crash, it repopulates routing tables in
an uncoordinated manner, like after clean start. This may be impractical in some
cases, because if the forwarding plane (i.e. kernel routing tables) remains
intact, then its synchronization with BIRD would temporarily disrupt packet
forwarding until protocols converge. Graceful restart is a mechanism that could
help with this issue. Generally, it works by starting protocols and letting them
repopulate routing tables while deferring route propagation until protocols
acknowledge their convergence. Note that graceful restart behavior have to be
configured for all relevant protocols and requires protocol-specific support
(currently implemented for Kernel and BGP protocols), it is activated for
particular boot by option <cf/-R/.

<p>Some protocols (e.g. BGP) could be restarted gracefully after both
intentional outage and crash, while others (e.g. OSPF) after intentional outage
only. For planned graceful restart, BIRD must be shut down by
<ref id="cli-graceful-restart" name="graceful restart"> command instead of
regular <ref id="cli-down" name="down"> command. In this way routing neighbors
are notified about planned graceful restart and routes are kept in kernel table
after shutdown.

<sect>MPLS
<label id="mpls">

<p>Multiprotocol Label Switching (MPLS) is a networking technology which works
below IP routing but above the link (e.g. ethernet) layer. It is described in
<rfc id="3031">.

In regular IP forwarding, the destination address of a packet is independently
examined in each hop, a route with longest prefix match is selected from the
routing table, and packet is processed accordingly. In general, there is no
difference between border routers and internal routers w.r.t. IP forwarding.

In MPLS forwarding, when a packet enters the network, it is classified (based on
destination address, ingress interface and other factors) into one of forwarding
equivalence classes (FECs), then a header with a MPLS label identifying the FEC
is attached to it, and the packet is forwarded. In internal routers, only the
MPLS label is examined, the matching MPLS route is selected from the MPLS
routing table, and the packet is processed accordingly. The specific value of
MPLS label has local meaning only and may change between hops (that is why it is
called label switching). When the packet leaves the network, the MPLS header is
removed.

The advantage of the MPLS approach is that other factors than the destination
address can be considered and used consistently in the whole network, for
example IP traffic with multiple overlapping private address ranges could be
mixed together, or particular paths for specific flows could be defined. Another
advantage is that MPLS forwarding by internal routers can be much simpler than
IP forwarding, as instead of the longest prefix match algorithm it uses simpler
exact match for MPLS route selection. The disadvantage is additional complexity
in signaling. For further details, see <rfc id="3031">.

MPLS-aware routing protocols not only distribute IP routing information, but
they also distribute labels. Therefore, they produce labeled routes - routes
representing label switched paths (LSPs) through the MPLS domain. Such routes
have IP prefix and next hop address like regular (non-labeled) routes, but they
also have local MPLS label (in route attribute <ref id="rta-mpls-label"
name="mpls_label">) and outgoing MPLS label (as a part of the next hop). They
are stored in regular IP routing tables.

Labeled routes are used for exchange of routing information between routing
protocols and for ingress (IP -&gt; MPLS) forwarding, but they are not directly
used for MPLS forwarding. For that purpose <ref id="mpls-routes" name="MPLS
routes"> are used. These are routes that have local MPLS label as a primary key
and they are stored in the MPLS routing table.

In BIRD, the whole process generally works this way: A MPLS-aware routing
protocol (say BGP) receives routing information including remote label. It
produces a route with attribute <ref id="rta-mpls-policy" name="mpls_policy">
specifying desired <ref id="mpls-channel-label-policy" name="MPLS label policy">.
Such route then passes the import filter (which could modify the MPLS label
policy or perhaps assign a static label) and when it is accepted, a local MPLS
label is selected (according to the label policy) and attached to the route,
producing labeled route. When a new MPLS label is allocated, the MPLS-aware
protocol automatically produces corresponding MPLS route. When all labeled
routes that use specific local MPLS label are retracted, the corresponding MPLS
route is retracted too.

There are three important concepts for MPLS in BIRD: MPLS domains, MPLS tables
and MPLS channels. MPLS domain represents an independent label space, all
MPLS-aware protocols are associated with some MPLS domain. It is responsible for
label management, handling label allocation requests from MPLS-aware protocols.
MPLS table is just a routing table for MPLS routes. Routers usually have one
MPLS domain and one MPLS table, with Kernel protocol to export MPLS routes into
kernel FIB.

MPLS channels make protocols MPLS-aware, they are responsible for keeping track
of active FECs (and corresponding allocated labels), selecting FECs / local
labels for labeled routes, and maintaining correspondence between labeled routes
and MPLS routes.

Note that local labels are allocated to individual MPLS-aware protocols and
therefore it is not possible to share local labels between different protocols.


<chapt>Configuration
<label id="config">

<sect>Introduction
<label id="config-intro">

<p>BIRD is configured using a text configuration file. Upon startup, BIRD reads
<it/prefix/<file>/etc/bird.conf</file> (unless the <tt/-c/ command line option
is given). Configuration may be changed at user's request: if you modify the
config file and then signal BIRD with <tt/SIGHUP/, it will adjust to the new
config. Then there's the client which allows you to talk with BIRD in an
extensive way.

<p>In the config, everything on a line after <cf/#/ or inside <cf>/* */</cf> is
a comment, whitespace characters are treated as a single space. If there's a
variable number of options, they are grouped using the <cf/{ }/ brackets. Each
option is terminated by a <cf/;/. Configuration is case sensitive. There are two
ways how to name symbols (like protocol names, filter names, constants etc.).
You can either use a simple string starting with a letter (or underscore)
followed by any combination of letters, numbers and underscores (e.g. <cf/R123/,
<cf/my_filter/, <cf/bgp5/) or you can enclose the name into apostrophes (<cf/'/)
and than you can use any combination of numbers, letters, underscores, hyphens,
dots and colons (e.g.  <cf/'1:strange-name'/, <cf/'-NAME-'/, <cf/'cool::name'/).

<p>Here is an example of a simple config file. It enables synchronization of
routing tables with OS kernel, learns network interfaces and runs RIP on all
network interfaces found.

<code>
protocol kernel {
	ipv4 {
		export all;	# Default is export none
	};
	persist;		# Don't remove routes on BIRD shutdown
}

protocol device {
}

protocol rip {
	ipv4 {
		import all;
		export all;
	};
	interface "*";
}
</code>


<sect>Global options
<label id="global-opts">

<p><descrip>
	<tag><label id="opt-include">include "<m/filename/";</tag>
	This statement causes inclusion of a new file. The <m/filename/ could
	also be a wildcard, in that case matching files are included in
	alphabetic order. The maximal depth is 8. Note that this statement can
	be used anywhere in the config file, even inside other options, but
	always on the beginning of line. In the following example, the first
	semicolon belongs to the <cf/include/, the second to <cf/ipv6 table/.
	If the <file/tablename.conf/ contains exactly one token (the name of the
	table), this construction is correct:
<code>
ipv6 table
include "tablename.conf";;
</code>

	<tag><label id="opt-log">log "<m/filename/" [<m/limit/ "<m/backup/"] | fixed "<m/filename/" <m/size/ | syslog [name <m/name/] | stderr all|{ <m/list of classes/ }</tag>
	Set logging of messages having the given class (either <cf/all/ or <cf>{
	error|trace [, <m/.../] }</cf> etc.) into selected destination - a file
	specified as a filename string (with optional log rotation information),
	syslog (with optional name argument), or the stderr output.

	Classes are:
	<cf/info/, <cf/warning/, <cf/error/ and <cf/fatal/ for messages about local problems,
	<cf/debug/ for debugging messages,
	<cf/trace/ when you want to know what happens in the network,
	<cf/remote/ for messages about misbehavior of remote machines,
	<cf/auth/ about authentication failures,
	<cf/bug/ for internal BIRD bugs.

	Logging directly to file supports basic log rotation -- there is an
	optional log file limit and a backup filename, when log file reaches the
	limit, the current log file is renamed to the backup filename and a new
	log file is created. It's also possible to log to a single file behaving
	as a ring buffer with a fixed size.

	You may specify more than one <cf/log/ line to establish logging to
	multiple destinations. Default: log everything to the system log, or
	to the debug output if debugging is enabled by <cf/-d//<cf/-D/
	command-line option.

	<tag><label id="opt-debug-protocols">debug protocols all|off|{ states|routes|filters|interfaces|events|packets [, <m/.../] }</tag>
	Set global defaults of protocol debugging options.
	See <ref id="proto-debug" name="debug"> in the following section.
	Default: off.

	<tag><label id="opt-debug-channels">debug channels all|off|{ states|routes|filters|events [, <m/.../] }</tag>
	Set global defaults of channel debugging options.
	See <ref id="channel-debug" name="debug"> in the channel section.
	Default: off.

	<tag><label id="opt-debug-tables">debug tables all|off|{ states|routes|filters|events [, <m/.../] }</tag>
	Set global defaults of table debugging options.
	See <ref id="rtable-debug" name="debug"> in the table section.
	Default: off.

	<tag><label id="opt-debug-commands">debug commands <m/number/</tag>
	Control logging of client connections (0 for no logging, 1 for logging
	of connects and disconnects, 2 and higher for logging of all client
	commands). Default: 0.

	<tag><label id="opt-debug-latency">debug latency <m/switch/</tag>
	Activate tracking of elapsed time for internal events. Recent events
	could be examined using <cf/dump events/ command. Default: off.

	<tag><label id="opt-debug-latency-limit">debug latency limit <m/time/</tag>
	If <cf/debug latency/ is enabled, this option allows to specify a limit
	for elapsed time. Events exceeding the limit are logged. Default: 1 s.

	<tag><label id="opt-watchdog-warn">watchdog warning <m/time/</tag>
	Set time limit for I/O loop cycle. If one iteration took more time to
	complete, a warning is logged. Default: 5 s.

	<tag><label id="opt-watchdog-timeout">watchdog timeout <m/time/</tag>
	Set time limit for I/O loop cycle. If the limit is breached, BIRD is
	killed by abort signal. The timeout has effective granularity of
	seconds, zero means disabled. Default: disabled (0).

	<tag><label id="opt-threads">threads <m/number/</tag>
	Set how many worker threads should BIRD spawn. Tests show that every
        thread can utilize one complete CPU core, therefore you probably want to
        keep at least one free core. The maximum feasible thread count heavily
        depends on the actual workload and must be determined by testing or estimation. Default: 1

	<tag><label id="opt-mrtdump">mrtdump "<m/filename/"</tag>
	Set MRTdump file name. This option must be specified to allow MRTdump
	feature. Default: no dump file.

	<tag><label id="opt-mrtdump-protocols">mrtdump protocols all|off|{ states|messages [, <m/.../] }</tag>
	Set global defaults of MRTdump options. See <cf/mrtdump/ in the
	following section. Default: off.

	<tag><label id="opt-filter">filter <m/name local variables/{ <m/commands/ }</tag>
	Define a filter. You can learn more about filters in the following
	chapter.

	<tag><label id="opt-function">function <m/name/ (<m/parameters/) [ -&gt; <m/return type/ ] <m/local variables/ { <m/commands/ }</tag>
	Define a function. You can learn more about functions in the following chapter.

	<tag><label id="opt-protocol">protocol rip|ospf|bgp|<m/.../ [<m/name/ [from <m/name2/]] { <m>protocol options</m> }</tag>
	Define a protocol instance called <cf><m/name/</cf> (or with a name like
	"rip5" generated automatically if you don't specify any
	<cf><m/name/</cf>). You can learn more about configuring protocols in
	their own chapters. When <cf>from <m/name2/</cf> expression is used,
	initial protocol options are taken from protocol or template
	<cf><m/name2/</cf> You can run more than one instance of most protocols
	(like RIP or BGP). By default, no instances are configured.

	<tag><label id="opt-template">template rip|ospf|bgp|<m/.../ [<m/name/ [from <m/name2/]] { <m>protocol options</m> }</tag>
	Define a protocol template instance called <m/name/ (or with a name like
	"bgp1" generated automatically if you don't specify any	<m/name/).
	Protocol templates can be used to group common options when many
	similarly configured protocol instances are to be defined. Protocol
	instances (and other templates) can use templates by using <cf/from/
	expression and the name of the template. At the moment templates (and
	<cf/from/ expression) are not implemented for OSPF protocol.

	<tag><label id="opt-define">define <m/constant/ = <m/expression/</tag>
	Define a constant. You can use it later in every place you could use a
	value of the same type. Besides, there are some predefined numeric
	constants based on /etc/iproute2/rt_* files. A list of defined constants
	can be seen (together with other symbols) using 'show symbols' command.

	<tag><label id="opt-attribute">attribute <m/type/ <m/name/</tag>
	Declare a custom route attribute. You can set and get it in filters like
	any other route attribute. This feature is intended for marking routes
	in import filters for export filtering purposes instead of locally
	assigned BGP communities which have to be deleted in export filters.

	<tag><label id="opt-router-id">router id <m/IPv4 address/</tag>
	Set BIRD's router ID. It's a world-wide unique identification of your
	router, usually one of router's IPv4 addresses. Default: the lowest
	IPv4 address of a non-loopback interface.

	<tag><label id="opt-router-id-from">router id from [-] [ "<m/mask/" ] [ <m/prefix/ ] [, <m/.../]</tag>
	Set BIRD's router ID based on an IPv4 address of an interface specified by
	an interface pattern.
	See <ref id="proto-iface" name="interface"> section for detailed
	description of interface patterns with extended clauses.

	<tag><label id="opt-hostname">hostname "<m/name/"</tag>
	Set hostname. Default: node name as returned by `uname -n'.

	<tag><label id="opt-graceful-restart">graceful restart wait <m/number/</tag>
	During graceful restart recovery, BIRD waits for convergence of routing
	protocols. This option allows to specify a timeout for the recovery to
	prevent waiting indefinitely if some protocols cannot converge. Default:
	240 seconds.

	<tag><label id="opt-timeformat">timeformat route|protocol|base|log "<m/format1/" [<m/limit/ "<m/format2/"]</tag>
	This option allows to specify a format of date/time used by BIRD. The
	first argument specifies for which purpose such format is used.
	<cf/route/ is a format used in 'show route' command output,
	<cf/protocol/ is used in 'show protocols' command output, <cf/base/ is
	used for other commands and <cf/log/ is used in a log file.

	"<m/format1/" is a format string using <it/strftime(3)/ notation (see
	<it/man strftime/ for details). It is extended to support sub-second
	time part with variable precision (up to microseconds) using "%f"
	conversion code (e.g., "%T.%3f" is hh:mm:ss.sss time). <m/limit/ and
	"<m/format2/" allow to specify the second format string for times in
	past deeper than <m/limit/ seconds.

	There are several shorthands: <cf/iso long/ is a ISO 8601 date/time
	format (YYYY-MM-DD hh:mm:ss) that can be also specified using <cf/"%F
	%T"/. Similarly, <cf/iso long ms/ and <cf/iso long us/ are ISO 8601
	date/time formats with millisecond or microsecond precision.
	<cf/iso short/ is a variant of ISO 8601 that uses just the time format
	(hh:mm:ss) for near times (up to 20 hours in the past) and the date
	format (YYYY-MM-DD) for far times. This is a shorthand for <cf/"%T"
	72000 "%F"/. And there are also <cf/iso short ms/ and <cf/iso short us/
	high-precision variants of that.

	By default, BIRD uses the <cf/iso short ms/ format for <cf/route/ and
	<cf/protocol/ times, and the <cf/iso long ms/ format for <cf/base/ and
	<cf/log/ times.

	<tag><label id="opt-table"><m/nettype/ table <m/name/ [ { <m/option/; [<m/.../] } ]</tag>
	Define a new routing table. The default routing tables <cf/master4/ and
	<cf/master6/ are defined implicitly, other routing tables have to be
	defined by this option. See the <ref id="rtable-opts"
	name="routing table configuration section"> for routing table options.

	<tag><label id="opt-mpls-domain">mpls domain <m/name/ [ { <m/option/; [<m/.../] } ]</tag>
	Define a new MPLS domain. MPLS domains represent independent label
	spaces and are responsible for MPLS label management. All MPLS-aware
	protocols are associated with some MPLS domain. See the <ref id="mpls-opts"
	name="MPLS configuration section"> for MPLS domain options.

	<tag><label id="opt-eval">eval <m/expr/</tag>
	Evaluates given filter expression. It is used by the developers for testing of filters.
</descrip>


<sect>Routing table options
<label id="rtable-opts">

<p>Most routing tables do not need any options and are defined without an option
block, but there are still some options to tweak routing table behavior. Note
that implicit tables (<cf/master4/ and <cf/master6/) can be redefined in order
to set options.

<descrip>
	<tag><label id="rtable-sorted">sorted <m/switch/</tag>
	Usually, a routing table just chooses the selected (best) route from a
	list of routes for each network, while keeping remaining routes unsorted.
	If enabled, these lists of routes are kept completely sorted (according
	to preference or some protocol-dependent metric).

	This is needed for some protocol features (e.g. <cf/secondary/ option of
	BGP protocol, which allows to accept not just a selected route, but the
	first route (in the sorted list) that is accepted by filters), but it is
	incompatible with some other features (e.g. <cf/deterministic med/
	option of BGP protocol, which activates a way of choosing selected route
	that cannot be described using comparison and ordering). Minor advantage
	is that routes are shown sorted in <cf/show route/, minor disadvantage
	is that it is slightly more computationally expensive. Default: off.

	<tag><label id="rtable-trie">trie <m/switch/</tag>
	BIRD routing tables are implemented with hash tables, which is efficient
	for exact-match lookups, but inconvenient for longest-match lookups or
	interval lookups (finding superprefix or subprefixes). This option
	activates additional trie structure that is used to accelerate these
	lookups, while using the hash table for exact-match lookups.

	This has advantage for <ref id="rpki" name="RPKI"> (on ROA tables),
	for <ref id="bgp-gateway" name="recursive next-hops"> (on IGP tables),
	and is required for <ref id="bgp-validate" name="flowspec validation">
	(on base IP tables). Another advantage is that interval results (like
	from <cf/show route in .../ command) are lexicographically sorted. The
	disadvantage is that trie-enabled routing tables require more memory,
	which may be an issue especially in multi-table setups. Default: off.

	<tag><label id="rtable-gc-threshold">gc threshold <m/number/</tag>
	Specify a minimum amount of removed networks that triggers a garbage
	collection (GC) cycle. Default: 1000.

	<tag><label id="rtable-gc-period">gc period <m/time/</tag>
	Specify a period of time between consecutive GC cycles. When there is a
	significant amount of route withdraws, GC cycles are executed repeatedly
	with given period time (with some random factor). When there is just
	small amount of changes, GC cycles are not executed. In extensive route
	server setups, running GC on hundreds of full BGP routing tables can
	take significant amount of time, therefore they should use higher GC
	periods. Default: adaptive, based on number of routing tables in the
	configuration. From 10 s (with <= 25 routing tables) up to 600 s (with
	>= 1500 routing tables).

	<tag><label id="rtable-cork-threshold">cork threshold <m/number/ <m/number/</tag>
	Too many pending exports may lead to memory bloating. In such cases,
        BIRD tries to relieve the memory pressure by pausing some routines until
        the queue sizes get low enough. This option allows the user to set the
        thresholds; first value is the low threshold (when to resume), the
        second one is the high threshold (when to pause). The higher is the
        threshold, the more memory can get used. In most cases, the defaults
	should work for you. Default: 1024 8192.

	<tag><label id="rtable-export-settle-time">export settle time <m/time/ <m/time/</tag>
	Minimum and maximum settle times, respectively, for export announcements.
	When multiple routes are changing, this mechanism waits for the changes
	to settle before waking up sleeping export threads but if the changes are coming
	steadily, BIRD isn't waiting forever; at most the maximum time.
	Default values: <cf/1 ms 100 ms/. You have to always provide both values.

	<tag><label id="rtable-route-refresh-export-settle-time">route refresh export settle time <m/time/ <m/time/</tag>
	Minimum and maximum settle times, respectively, for export announcements
	(the same as above), valid when any channel is currently doing a route refresh.
	This serves a purpose of even more aggresive change bundling, knowing that there
	is some active process generating changes in a fast pace. If you don't want
	this feature, set this to the same values as <ref id="rtable-export-settle-time" name="export settle time">.
	Default values: <cf/100 ms 3 s/.

	<tag><label id="rtable-debug">debug all|off|{ states|routes|events [, <m/.../] }</tag>
	Set table debugging options. Each table can write some trace messages
	into log with category <cf/trace/. You can request <cf/all/ trace messages
	or select some types: <cf/states/ for table state changes and auxiliary
	processes, <cf/routes/ for auxiliary route notifications (next hop update,
        flowspec revalidation) and <cf/events/ for more detailed auxiliary routine
	debug. See also <ref id="channel-debug" name="channel debugging option">.
	Default: off.

</descrip>


<sect>Protocol options
<label id="protocol-opts">

<p>For each protocol instance, you can configure a bunch of options. Some of
them (those described in this section) are generic, some are specific to the
protocol (see sections talking about the protocols).

<p>Several options use a <m/switch/ argument. It can be either <cf/on/,
<cf/yes/ or a numeric expression with a non-zero value for the option to be
enabled or <cf/off/, <cf/no/ or a numeric expression evaluating to zero to
disable it. An empty <m/switch/ is equivalent to <cf/on/ ("silence means
agreement").

<descrip>
	<tag><label id="proto-disabled">disabled <m/switch/</tag>
	Disables the protocol. You can change the disable/enable status from the
	command line interface without needing to touch the configuration.
	Disabled protocols are not activated. Default: protocol is enabled.

	<tag><label id="proto-debug">debug all|off|{ states|routes|filters|interfaces|events|packets [, <m/.../] }</tag>
	Set protocol debugging options. If asked, each protocol is capable of
	writing trace messages about its work to the log (with category
	<cf/trace/). You can either request printing of <cf/all/ trace messages
	or only of the selected types: <cf/states/ for protocol state changes
	(protocol going up, down, starting, stopping etc.), <cf/routes/ for
	routes exchanged with the routing table, <cf/filters/ for details on
	route filtering, <cf/interfaces/ for interface change events sent to
	the protocol, <cf/events/ for events internal to the protocol and
	<cf/packets/ for packets sent and received by the protocol. Classes
	<cf/routes/ and <cf/filters/ can be also set per-channel using
	<ref id="channel-debug" name="channel debugging option">) Default: off.

	<tag><label id="proto-mrtdump">mrtdump all|off|{ states|messages [, <m/.../] }</tag>
	Set protocol MRTdump flags. MRTdump is a standard binary format for
	logging information from routing protocols and daemons. These flags
	control what kind of information is logged from the protocol to the
	MRTdump file (which must be specified by global <cf/mrtdump/ option, see
	the previous section). Although these flags are similar to flags of
	<cf/debug/ option, their meaning is different and protocol-specific. For
	BGP protocol, <cf/states/ logs BGP state changes and <cf/messages/ logs
	received BGP messages. Other protocols does not support MRTdump yet.

	<tag><label id="proto-router-id">router id <m/IPv4 address/</tag>
	This option can be used to override global router id for a given
	protocol. Default: uses global router id.

	<tag><label id="proto-description">description "<m/text/"</tag>
	This is an optional description of the protocol. It is displayed as a
	part of the output of 'show protocols all' command.

	<tag><label id="proto-vrf">vrf "<m/text/"|default</tag>
	Associate the protocol with specific VRF. The protocol will be
	restricted to interfaces assigned to the VRF and will use sockets bound
	to the VRF. A corresponding VRF interface must exist on OS level. For
	kernel protocol, an appropriate table still must be explicitly selected
	by <cf/table/ option.

	By selecting <cf/default/, the protocol is associated with the default
	VRF; i.e., it will be restricted to interfaces not assigned to any
	regular VRF. That is different from not specifying <cf/vrf/ at all, in
	which case the protocol may use any interface regardless of its VRF
	status.

	Note that for proper VRF support it is necessary to use Linux kernel
	version at least 4.14, older versions have limited VRF implementation.
	Before Linux kernel 5.0, a socket bound to a port in default VRF collide
	with others in regular VRFs. In BGP, this can be avoided by using
	<ref id="bgp-strict-bind" name="strict bind"> option.

	<tag><label id="proto-channel"><m/channel name/ [{<m/channel config/}]</tag>
	Every channel must be explicitly stated. See the protocol-specific
	configuration for the list of supported channel names. See the
	<ref id="channel-opts" name="channel configuration section"> for channel
	definition.
</descrip>

<p>There are several options that give sense only with certain protocols:

<descrip>
	<tag><label id="proto-iface">interface [-] [ "<m/mask/" ] [ <m/prefix/ ] [, <m/.../] [ { <m/option/; [<m/.../] } ]</tag>
	Specifies a set of interfaces on which the protocol is activated with
	given interface-specific options. A set of interfaces specified by one
	interface option is described using an interface pattern. The interface
	pattern consists of a sequence of clauses (separated by commas), each
	clause is a mask specified as a shell-like pattern. Interfaces are
	matched by their name.

	An interface matches the pattern if it matches any of its clauses. If
	the clause begins with <cf/-/, matching interfaces are excluded. Patterns
	are processed left-to-right, thus <cf/interface "eth0", -"eth*", "*";/
	means eth0 and all non-ethernets.

	Some protocols (namely OSPFv2 and Direct) support extended clauses that
	may contain a mask, a prefix, or both of them. An interface matches such
	clause if its name matches the mask (if specified) and its address
	matches the prefix (if specified). Extended clauses are used when the
	protocol handles multiple addresses on an interface independently.

	An interface option can be used more times with different interface-specific
	options, in that case for given interface the first matching interface
	option is used.

	This option is allowed in Babel, BFD, Device, Direct, OSPF, RAdv and RIP
	protocols. In OSPF protocol it is used in the <cf/area/ subsection.

	Default: none.

	Examples:

	<cf>interface "*" { type broadcast; };</cf> - start the protocol on all
	interfaces with <cf>type broadcast</cf> option.

	<cf>interface "eth1", "eth4", "eth5" { type ptp; };</cf> - start the
	protocol on enumerated interfaces with <cf>type ptp</cf> option.

	<cf>interface -192.168.1.0/24, 192.168.0.0/16;</cf> - start the protocol
	on all interfaces that have address from 192.168.0.0/16, but not from
	192.168.1.0/24.

	<cf>interface "eth*" 192.168.1.0/24;</cf> - start the protocol on all
	ethernet interfaces that have address from 192.168.1.0/24.

	<tag><label id="proto-tx-class">tx class|dscp <m/num/</tag>
	This option specifies the value of ToS/DS/Class field in IP headers of
	the outgoing protocol packets. This may affect how the protocol packets
	are processed by the network relative to the other network traffic. With
	<cf/class/ keyword, the value (0-255) is used for the whole ToS/Class
	octet (but two bits reserved for ECN are ignored). With	<cf/dscp/
	keyword, the value (0-63) is used just for the DS field in the octet.
	Default value is 0xc0 (DSCP 0x30 - CS6).

	<tag><label id="proto-tx-priority">tx priority <m/num/</tag>
	This option specifies the local packet priority. This may affect how the
	protocol packets are processed in the local TX queues. This option is
	Linux specific. Default value is 7 (highest priority, privileged traffic).

	<tag><label id="proto-pass">password "<m/password/" | <m/bytestring/ [ { <m>password options</m> } ] </tag>
	Specifies a password that can be used by the protocol as a shared secret
	key. Password option can be used more times to specify more passwords.
	If more passwords are specified, it is a protocol-dependent decision
	which one is really used. Specifying passwords does not mean that
	authentication is enabled, authentication can be enabled by separate,
	protocol-dependent <cf/authentication/ option.

	A password can be specified as a string or as a sequence of hexadecimal
	digit pairs (<ref id="type-bytestring" name="bytestring">).

	This option is allowed in BFD, OSPF, RIP, and Babel protocols. BGP has
	also <cf/password/ option, but it is slightly different and described
	separately. Default: none.
</descrip>

<p>Password option can contain section with some (not necessary all) password sub-options:

<descrip>
	<tag><label id="proto-pass-id">id <M>num</M></tag>
	ID of the password, (0-255). If it is not specified, BIRD will choose ID
	based on an order of the password item in the interface, starting from
	1. For example, second password item in one interface will have default
	ID 2. ID 0 is allowed by BIRD, but some other implementations may not
	allow it. ID is used by some routing protocols to identify which
	password was used to authenticate protocol packets.

	<tag><label id="proto-pass-gen-from">generate from "<m/time/"</tag>
	The start time of the usage of the password for packet signing.
	The format of <cf><m/time/</cf> is <tt>dd-mm-yyyy HH:MM:SS</tt>.

	<tag><label id="proto-pass-gen-to">generate to "<m/time/"</tag>
	The last time of the usage of the password for packet signing.

	<tag><label id="proto-pass-accept-from">accept from "<m/time/"</tag>
	The start time of the usage of the password for packet verification.

	<tag><label id="proto-pass-accept-to">accept to "<m/time/"</tag>
	The last time of the usage of the password for packet verification.

	<tag><label id="proto-pass-from">from "<m/time/"</tag>
	Shorthand for setting both <cf/generate from/ and <cf/accept from/.

	<tag><label id="proto-pass-to">to "<m/time/"</tag>
	Shorthand for setting both <cf/generate to/ and <cf/accept to/.

	<tag><label id="proto-pass-algorithm">algorithm ( keyed md5 | keyed sha1 | hmac sha1 | hmac sha256 | hmac sha384 | hmac sha512 | blake2s128 | blake2s256 | blake2b256 | blake2b512 )</tag>
	The message authentication algorithm for the password when cryptographic
	authentication is enabled. The default value depends on the protocol.
	For RIP and OSPFv2 it is Keyed-MD5 (for compatibility), for OSPFv3 and
	Babel it is HMAC-SHA-256.

</descrip>


<sect>Channel options
<label id="channel-opts">

<p>Every channel belongs to a protocol and is configured inside its block. The
minimal channel config is empty, then it uses default values. The name of the
channel implies its nettype. Channel definitions can be inherited from protocol
templates. Multiple definitions of the same channel are forbidden, but channels
inherited from templates can be updated by new definitions.

<descrip>
	<tag><label id="channel-debug">debug all|off|{ states|routes|filters [, <m/.../] }</tag>
	Set channel debugging options. Like in <ref id="proto-debug"
	name="protocol debugging">, channels are capable of writing trace
	messages about its work to the log (with category <cf/trace/). You can
	either request printing of <cf/all/ trace messages or only of the
	selected types: <cf/states/ for channel state changes (channel going up,
	down, feeding, reloading etc.), <cf/routes/ for routes propagated
	through the channel, <cf/filters/ for details on route filtering,
	remaining debug flags are not used in channel debug. Default: off.

	<tag><label id="proto-table">table <m/name/</tag>
	Specify a table to which the channel is connected. Default: the first
	table of given nettype.

	<tag><label id="proto-preference">preference <m/expr/</tag>
	Sets the preference of routes generated by the protocol and imported
	through this channel. Default: protocol dependent.

	<tag><label id="proto-import">import all | none | filter <m/name/ | filter { <m/filter commands/ } | where <m/boolean filter expression/</tag>
	Specify a filter to be used for filtering routes coming from the
	protocol to the routing table. <cf/all/ is for keeping all routes,
	<cf/none/ is for dropping all routes. Default: <cf/all/ (except for
	EBGP).

	<tag><label id="proto-export">export [ in <m/prefix/ ] <m/filter/</tag>
	This is similar to the <cf>import</cf> keyword, except that it works in
	the direction from the routing table to the protocol. If <cf/in/ keyword is used,
	only routes inside the given prefix are exported. Other routes are completely
	ignored (e.g. no logging and no statistics).
	Default: <cf/none/ (except for EBGP and L3VPN).

	<tag><label id="proto-import-keep-filtered">import keep filtered <m/switch/</tag>
	Usually, if an import filter rejects a route, the route is forgotten.
	When this option is active, these routes are kept in the routing table,
	but they are hidden and not propagated to other protocols. But it is
	possible to show them using <cf/show route filtered/. Note that this
	option does not work for the pipe protocol. Default: off.

	<tag><label id="proto-rpki-reload">rpki reload <m/switch/</tag>
	Import or export filters may depend on route RPKI status (using
	<cf/roa_check()/ operator). In contrast to to other filter operators,
	this status for the same route may change as the content of ROA tables
	changes. When this option is active, BIRD activates automatic reload of
	the appropriate subset of prefixes imported or exported by the channels
	whenever ROA tables are updated (after a short settle
	time). When disabled, route reloads have to be requested manually. The
	option is ignored if <cf/roa_check()/ is not used in channel filters.
	Note that for BGP channels, automatic reload requires
	<ref id="bgp-import-table" name="import table"> or
	<ref id="bgp-export-table" name="export table"> (for respective
	direction). Default: on.

	<tag><label id="rtable-min-settle-time">roa settle time <m/time/ <m/time/</tag>
	Minimum and maximum settle times, respectively, for ROA table changes.
        The automatic reload is triggered after the minimum time after the last
        ROA table change has been received but not later than the maximum time after
        first unprocessed ROA table change. Therefore with default values, the
        automatic reload happens 1 second after the ROA table stops updating, yet if it
	were to be later than 20 seconds after the ROA table starts updating,
	the automatic reload is triggered anyway. Default values: <cf/1 s 20 s/.
	You have to always provide both values.

	<tag><label id="proto-import-limit">import limit [<m/number/ | off ] [action warn | block | restart | disable]</tag>
	Specify an import route limit (a maximum number of routes imported from
	the protocol) and optionally the action to be taken when the limit is
	hit. Warn action just prints warning log message. Block action discards
	new routes coming from the protocol. Restart and disable actions shut
	the protocol down like appropriate commands. Disable is the default
	action if an action is not explicitly specified. Note that limits are
	reset during protocol reconfigure, reload or restart. Default: <cf/off/.

	<tag><label id="proto-receive-limit">receive limit [<m/number/ | off ] [action warn | block | restart | disable]</tag>
	Specify an receive route limit (a maximum number of routes received from
	the protocol and remembered). It works almost identically to <cf>import
	limit</cf> option, the only difference is that if <cf/import keep
	filtered/ option is active, filtered routes are counted towards the
	limit and blocked routes are forgotten, as the main purpose of the
	receive limit is to protect routing tables from overflow. Import limit,
	on the contrary, counts accepted routes only and routes blocked by the
	limit are handled like filtered routes. Default: <cf/off/.

	<tag><label id="proto-export-limit">export limit [ <m/number/ | off ] [action warn | block | restart | disable]</tag>
	Specify an export route limit, works similarly to the <cf>import
	limit</cf> option, but for the routes exported to the protocol. This
	option is experimental, there are some problems in details of its
	behavior -- the number of exported routes can temporarily exceed the
	limit without triggering it during protocol reload, exported routes
	counter ignores route blocking and block action also blocks route
	updates of already accepted routes -- and these details will probably
	change in the future. Default: <cf/off/.

	<tag><label id="proto-export-block">export block <m/number/</tag>
        Set the minimum amount of routes exported at once when feeding or
	if `merge paths` or `secondary` is selected. This affects overall latency.
	Basically, when your export filters are very expensive, processing
	the whole block of routes may take too much time. In such cases, you may need to
        shrink this value to improve responsiveness. Default: <cf/16384/.
</descrip>

<p>This is a trivial example of RIP configured for IPv6 on all interfaces:
<code>
protocol rip ng {
	ipv6;
	interface "*";
}
</code>

<p>This is a non-trivial example.
<code>
protocol rip ng {
	ipv6 {
		table mytable6;
		import filter { ... };
		export filter { ... };
		import limit 50;
	};
	interface "*";
}
</code>

<p>And this is even more complicated example using templates.
<code>
template bgp {
	local 198.51.100.14 as 65000;

	ipv4 {
		table mytable4;
		import filter { ... };
		export none;
	};
	ipv6 {
		table mytable6;
		import filter { ... };
		export none;
	};
}

protocol bgp from  {
	neighbor 198.51.100.130 as 64496;

	# IPv4 channel is inherited as-is, while IPv6
	# channel is adjusted by export filter option
	ipv6 {
		export filter { ... };
	};
}
</code>


<sect>MPLS options
<label id="mpls-opts">

<p>The MPLS domain definition is mandatory for a MPLS router. All MPLS channels
and MPLS-aware protocols are associated with some MPLS domain (although usually
implicitly with the sole one). In the MPLS domain definition you can configure
details of MPLS label allocation. Currently, there is just one option:

<descrip>
	<tag><label id="mpls-domain-label-range">label range <m/name/ { start <m/number/; length <m/number/; [<m/.../] }</tag>
	Define a new label range, or redefine implicit label ranges <cf/static/
	and <cf/dynamic/. MPLS channels use configured label ranges for dynamic
	label allocation, while <cf/static/ label range is used for static label
	allocation. The label range definition must specify the extent of the
	range. By default, the range <cf/static/ is 16-1000, while the range
	<cf/dynamic/ is 1000-10000.
</descrip>

<p>MPLS channel should be defined in each MPLS-aware protocol in addition to its
regular channels. It is responsible for label allocation and for announcing MPLS
routes to the MPLS routing table. Besides common <ref id="channel-opts"
name="channel options">, MPLS channels have some specific options:

<descrip>
	<tag><label id="mpls-channel-domain">domain <m/name/</tag>
	Specify a MPLS domain to which this channel and protocol belongs.
	Default: The first defined MPLS domain.

	<tag><label id="mpls-channel-label-range">label range <m/name/</tag>
	Use specific label range for dynamic label allocation. Note that static
	labels always use the range <cf/static/. Default: the range <cf/dynamic/.

	<tag><label id="mpls-channel-label-policy">label policy static|prefix|aggregate|vrf</tag>
	Label policy specifies how routes are grouped to forwarding equivalence
	classes (FECs) and how labels are assigned to them.

	The policy <cf/static/ means no dynamic label allocation is done, and
	static labels must be set in import filters using the route attribute
	<ref id="rta-mpls-label" name="mpls_label">.

	The policy <cf/prefix/ means each prefix uses separate label associated
	with that prefix. When a labeled route is updated, it keeps the label.
	This policy is appropriate for IGPs.

	The policy <cf/aggregate/ means routes are grouped to FECs according to
	their next hops (including next hop labels), and one label is used for
	all routes in the same FEC. When a labeled route is updated, it may
	change next hop, change FEC and therefore change label. This policy is
	appropriate for BGP.

	The policy <cf/vrf/ is only valid in L3VPN protocols. It uses one label
	for all routes from a VRF, while replacing the original next hop with
	lookup in the VRF.

	Default: <cf/prefix/.
</descrip>

<p>This is a trivial example of MPLS setup:
<code>
mpls domain mdom {
	label range bgprange { start 2000; length 1000;	};
}

mpls table mtab;

protocol static {
	ipv6;
	mpls;

	route 2001:db8:1:1/64 mpls 100 via 2001:db8:1:2::1/64 mpls 200;
}

protocol bgp {
	# regular channels
	ipv6 mpls { ...	};
	vpn6 mpls { ... };

	# MPLS channel
	mpls {
		# domain mdom;
		# table mtab;
		label range bgprange;
		label policy aggregate;
	};

	...
}
</code>


<chapt>Remote control
<label id="remote-control">

<p>You can use the command-line client <file>birdc</file> to talk with a running
BIRD. Communication is done using a <file/bird.ctl/ UNIX domain socket (unless
changed with the <tt/-s/ option given to both the server and the client). The
commands can perform simple actions such as enabling/disabling of protocols,
telling BIRD to show various information, telling it to show routing table
filtered by filter, or asking BIRD to reconfigure. Press <tt/?/ at any time to
get online help. Option <tt/-r/ can be used to enable a restricted mode of BIRD
client, which allows just read-only commands (<cf/show .../). Option <tt/-v/ can
be passed to the client, to make it dump numeric return codes along with the
messages. You do not necessarily need to use <file/birdc/ to talk to BIRD, your
own applications could do that, too -- the format of communication between BIRD
and <file/birdc/ is stable (see the programmer's documentation).

<p>There is also lightweight variant of BIRD client called <file/birdcl/, which
does not support command line editing and history and has minimal dependencies.
This is useful for running BIRD in resource constrained environments, where
Readline library (required for regular BIRD client) is not available.

<p>Many commands have the <m/name/ of the protocol instance as an argument.
This argument can be omitted if there exists only a single instance.

<p>Here is a brief list of supported functions:

<descrip>
	<tag><label id="cli-show-status">show status</tag>
	Show router status, that is BIRD version, uptime and time from last
	reconfiguration.

	<tag><label id="cli-show-interfaces">show interfaces [summary]</tag>
	Show the list of interfaces. For each interface, print its type, state,
	MTU and addresses assigned.

	<tag><label id="cli-show-protocols">show protocols [all]</tag>
	Show list of protocol instances along with tables they are connected to
	and protocol status, possibly giving verbose information, if <cf/all/ is
	specified.

	<!-- TODO: Move these protocol-specific remote control commands to the protocol sections -->
	<tag><label id="cli-show-ospf-iface">show ospf interface [<m/name/] ["<m/interface/"]</tag>
	Show detailed information about OSPF interfaces.

	<tag><label id="cli-show-ospf-neighbors">show ospf neighbors [<m/name/] ["<m/interface/"]</tag>
	Show a list of OSPF neighbors and a state of adjacency to them.

	<tag><label id="cli-show-ospf-state">show ospf state [all] [<m/name/]</tag>
	Show detailed information about OSPF areas based on a content of the
	link-state database. It shows network topology, stub networks,
	aggregated networks and routers from other areas and external routes.
	The command shows information about reachable network nodes, use option
	<cf/all/ to show information about all network nodes in the link-state
	database.

	<tag><label id="cli-show-ospf-topology">show ospf topology [all] [<m/name/]</tag>
	Show a topology of OSPF areas based on a content of the link-state
	database. It is just a stripped-down version of 'show ospf state'.

	<tag><label id="cli-show-ospf-lsadb">show ospf lsadb [global | area <m/id/ | link] [type <m/num/] [lsid <m/id/] [self | router <m/id/] [<m/name/] </tag>
	Show contents of an OSPF LSA database. Options could be used to filter
	entries.

	<tag><label id="cli-show-rip-interfaces">show rip interfaces [<m/name/] ["<m/interface/"]</tag>
	Show detailed information about RIP interfaces.

	<tag><label id="cli-show-rip-neighbors">show rip neighbors [<m/name/] ["<m/interface/"]</tag>
	Show a list of RIP neighbors and associated state.

	<tag><label id="cli-show-static">show static [<m/name/]</tag>
	Show detailed information about static routes.

	<tag><label id="cli-show-bfd-sessions">show bfd sessions [<m/name/]</tag>
	Show information about BFD sessions.

	<tag><label id="cli-show-symbols">show symbols [table|filter|function|protocol|template|roa|<m/symbol/]</tag>
	Show the list of symbols defined in the configuration (names of
	protocols, routing tables etc.).

	<tag><label id="cli-show-route">show route [[(for|in)] <m/prefix/|for <m/IP/] [table (<m/t/|all)] [(import|export) table <m/p/.<m/c/] [filter <m/f/|where <m/cond/] [(export|preexport|noexport) <m/p/] [protocol <m/p/] [(stats|count)] [<m/options/]</tag>
	Show contents of specified routing tables, that is routes, their metrics
	and (in case the <cf/all/ switch is given) all their attributes.

	<p>You can specify a <m/prefix/ if you want to print routes for a
	specific network. If you use <cf>for <m/prefix or IP/</cf>, you'll get
	the entry which will be used for forwarding of packets to the given
	destination. Finally, if you use <cf>in <m/prefix/</cf>, you get all
	prefixes covered by the given prefix.
	By default, all routes for each network are printed with
	the selected one at the top, unless <cf/primary/ is given in which case
	only the selected route is shown.

	<p>The <cf/show route/ command can process one or multiple routing
	tables. The set of selected tables is determined on three levels: First,
	tables can be explicitly selected by <cf/table/ switch, which could be
	used multiple times, all tables are specified by <cf/table all/. Second,
	tables can be implicitly selected by channels or protocols that are
	arguments of several other switches (e.g., <cf/export/, <cf/protocol/).
	Last, the set of default tables is used: <cf/master4/, <cf/master6/ and
	each first table of any other network type.

	<p>There are internal tables when <cf/(import|export) table/ options
	are used for some channels. They can be selected explicitly with
	<cf/(import|export) table/ switch, specifying protocol <m/p/ and
	channel name <m/c/.

	<p>You can also ask for printing only routes processed and accepted by
	a given filter (<cf>filter <m/name/</cf> or <cf>filter { <m/filter/ }
	</cf> or matching a given condition (<cf>where <m/condition/</cf>).

	The <cf/export/, <cf/preexport/ and <cf/noexport/ switches ask for
	printing of routes that are exported to the specified protocol or
	channel. With <cf/preexport/, the export filter of the channel is
	skipped. With <cf/noexport/, routes rejected by the export filter are
	printed instead. Note that routes not exported for other reasons
	(e.g. secondary routes or routes imported from that protocol) are not
	printed even with <cf/noexport/. These switches also imply that
	associated routing tables are selected instead of default ones.

	<p>You can also select just routes added by a specific protocol.
	<cf>protocol <m/p/</cf>. This switch also implies that associated
	routing tables are selected instead of default ones.

	<p>If BIRD is configured to keep filtered routes (see <cf/import keep
	filtered/ option), you can show them instead of routes by using
	<cf/filtered/ switch.

	<p>The <cf/stats/ switch requests showing of route statistics (the
	number of networks, number of routes before and after filtering). If
	you use <cf/count/ instead, only the statistics will be printed.

	<tag><label id="cli-mrt-dump">mrt dump table <m/name/|"<m/pattern/" to "<m/filename/" [filter <m/f/|where <m/c/]</tag>
	Dump content of a routing table to a specified file in MRT table dump
	format. See <ref id="mrt" name="MRT protocol"> for details.

	<tag><label id="cli-configure">configure [soft] ["<m/config file/"] [timeout [<m/num/]]</tag>
	Reload configuration from a given file. BIRD will smoothly switch itself
	to the new configuration, protocols are reconfigured if possible,
	restarted otherwise. Changes in filters usually lead to restart of
	affected protocols.

	The previous configuration is saved and the user can switch back to it
	with <ref id="cli-configure-undo" name="configure undo"> command. The
	old saved configuration is released (even if the reconfiguration attempt
	fails due to e.g. a syntax error).

	If <cf/soft/ option is used, changes in filters does not cause BIRD to
	restart affected protocols, therefore already accepted routes (according
	to old filters) would be still propagated, but new routes would be
	processed according to the new filters.

	If <cf/timeout/ option is used, config timer is activated. The new
	configuration could be either confirmed using <cf/configure confirm/
	command, or it will be reverted to the old one when the config timer
	expires. This is useful for cases when reconfiguration breaks current
	routing and a router becomes inaccessible for an administrator. The
	config timeout expiration is equivalent to <cf/configure undo/
	command. The timeout duration could be specified, default is 300 s.

	<tag><label id="cli-configure-confirm">configure confirm</tag>
	Deactivate the config undo timer and therefore confirm the current
	configuration.

	<tag><label id="cli-configure-undo">configure undo</tag>
	Undo the last configuration change and smoothly switch back to the
	previous (stored) configuration. If the last configuration change was
	soft, the undo change is also soft. There is only one level of undo, but
	in some specific cases when several reconfiguration requests are given
	immediately in a row and the intermediate ones are skipped then the undo
	also skips them back.

	<tag><label id="cli-configure-check">configure check ["<m/config file/"]</tag>
	Read and parse given config file, but do not use it. useful for checking
	syntactic and some semantic validity of an config file.

	<tag><label id="cli-enable-disable-restart">enable|disable|restart <m/name/|"<m/pattern/"|all</tag>
	Enable, disable or restart a given protocol instance, instances matching
	the <cf><m/pattern/</cf> or <cf/all/ instances.

	<tag><label id="cli-reload">reload [in|out] (<m/name/|"<m/pattern/"|all) [partial prefix] </tag>
	Reload a given protocol instance, that means re-import routes from the
	protocol instance and re-export preferred routes to the instance. If
	<cf/in/ or <cf/out/ options are used, the command is restricted to one
	direction (re-import or re-export).

	This command is useful if appropriate filters have changed but the
	protocol instance was not restarted (or reloaded), therefore it still
	propagates the old set of routes. For example when <cf/configure soft/
	command was used to change filters.

	If <cf/partial prefix/ option is used, only corresponding routes are reloaded.
	Protocol BGP does partial reload only if it has import table enabled, otherwise partial reload for BGP is refused.

	Re-export always succeeds, but re-import is protocol-dependent and might
	fail (for example, if BGP neighbor does not support route-refresh
	extension). In that case, re-export is also skipped. Note that for the
	pipe protocol, both directions are always reloaded together (<cf/in/ or
	<cf/out/ options are ignored in that case).

	<tag><label id="cli-down">down</tag>
	Shut BIRD down.

	<tag><label id="cli-graceful-restart">graceful restart</tag>
	Shut BIRD down for graceful restart. See <ref id="graceful-restart"
	name="graceful restart"> section for details.

	<tag><label id="cli-debug">debug <m/protocol/|<m/pattern/|all all|off|{ states|routes|filters|events|packets [, <m/.../] }</tag>
	Control protocol debugging.

	<tag><label id="cli-dump">dump resources|sockets|interfaces|neighbors|attributes|routes|protocols</tag>
	Dump contents of internal data structures to the debugging output.

	<tag><label id="cli-echo">echo all|off|{ <m/list of log classes/ } [ <m/buffer-size/ ]</tag>
	Control echoing of log messages to the command-line output.
	See <ref id="opt-log" name="log option"> for a list of log classes.

	<tag><label id="cli-eval">eval <m/expr/</tag>
	Evaluate given expression.
</descrip>


<chapt>Filters
<label id="filters">

<sect>Introduction
<label id="filters-intro">

<p>BIRD contains a simple programming language. (No, it can't yet read mail :-).
There are two objects in this language: filters and functions. Filters are
interpreted by BIRD core when a route is being passed between protocols and
routing tables. The filter language contains control structures such as if's and
switches, but it allows no loops. An example of a filter using many features can
be found in <file>filter/test.conf</file>.

<p>Filter gets the route, looks at its attributes and modifies some of them if
it wishes. At the end, it decides whether to pass the changed route through
(using <cf/accept/) or whether to <cf/reject/ it. A simple filter looks like
this:

<code>
filter not_too_far
{
	int var;
	if defined( rip_metric ) then
		var = rip_metric;
	else {
		var = 1;
		rip_metric = 1;
	}
	if rip_metric &gt; 10 then
		reject "RIP metric is too big";
	else
		accept "ok";
}
</code>

<p>As you can see, a filter has a header, a list of local variables, and a body.
The header consists of the <cf/filter/ keyword followed by a (unique) name of
filter. The list of local variables consists of <cf><M>type name</M>;</cf>
pairs where each pair declares one local variable. The body consists of <cf>
{ <M>statements</M> }</cf>. Each <m/statement/ is terminated by a <cf/;/. You
can group several statements to a single compound statement by using braces
(<cf>{ <M>statements</M> }</cf>) which is useful if you want to make a bigger
block of code conditional.

<p>BIRD supports functions, so that you don not have to repeat the same blocks
of code over and over. Functions can have zero or more parameters and they can
have local variables. If the function returns value, then you should always
specify its return type. Direct recursion is possible. Function definitions look
like this:

<code>
function name() -> int
{
	int local_variable;
	int another_variable = 5;
	return 42;
}

function with_parameters(int parameter) -> pair
{
	print parameter;
	return (1, 2);
}
</code>

<p>Like in C programming language, variables are declared inside function body,
either at the beginning, or mixed with other statements. Declarations may
contain initialization. You can also declare variables in nested blocks, such
variables have scope restricted to such block. There is a deprecated syntax to
declare variables after the <cf/function/ line, but before the first <cf/{/.
Functions are called like in C: <cf>name(); with_parameters(5);</cf>. Function
may return values using the <cf>return <m/[expr]/</cf> command. Returning a
value exits from current function (this is similar to C).

<p>Filters are defined in a way similar to functions except they cannot have
explicit parameters and cannot return. They get a route table entry as an implicit parameter, it
is also passed automatically to any functions called. The filter must terminate
with either <cf/accept/ or <cf/reject/ statement. If there is a runtime error in
filter, the route is rejected.

<p>A nice trick to debug filters is to use <cf>show route filter <m/name/</cf>
from the command line client. An example session might look like:

<code>
pavel@bug:~/bird$ ./birdc -s bird.ctl
BIRD 0.0.0 ready.
bird> show route
10.0.0.0/8         dev eth0 [direct1 23:21] (240)
195.113.30.2/32    dev tunl1 [direct1 23:21] (240)
127.0.0.0/8        dev lo [direct1 23:21] (240)
bird> show route ?
show route [<prefix>] [table <t>] [filter <f>] [all] [primary]...
bird> show route filter { if 127.0.0.5 &tilde; net then accept; }
127.0.0.0/8        dev lo [direct1 23:21] (240)
bird>
</code>


<sect>Data types
<label id="data-types">

<p>Each variable and each value has certain type. Booleans, integers and enums
are incompatible with each other (that is to prevent you from shooting oneself
in the foot).

<descrip>
	<tag><label id="type-bool">bool</tag>
	This is a boolean type, it can have only two values, <cf/true/ and
	<cf/false/. Boolean is the only type you can use in <cf/if/ statements.

	<tag><label id="type-int">int</tag>
	This is a general integer type. It is an unsigned 32bit type; i.e., you
	can expect it to store values from 0 to 4294967295. Overflows are not
	checked. You can use <cf/0x1234/ syntax to write hexadecimal values.

	<tag><label id="type-pair">pair</tag>
	This is a pair of two short integers. Each component can have values
	from 0 to 65535. Literals of this type are written as <cf/(1234,5678)/.
	The same syntax can also be used to construct a pair from two arbitrary
	integer expressions (for example <cf/(1+2,a)/).

	Operators <cf/.asn/ and <cf/.data/ can be used to extract corresponding
	components of a pair: <cf>(<m/asn/, <m/data/)</cf>.

	<tag><label id="type-quad">quad</tag>
	This is a dotted quad of numbers used to represent router IDs (and
	others). Each component can have a value from 0 to 255. Literals of
	this type are written like IPv4 addresses.

	<tag><label id="type-string">string</tag>
	This is a string of characters. There are no ways to modify strings in
	filters. You can pass them between functions, assign them to variables
	of type <cf/string/, print such variables, use standard string
	comparison operations (e.g. <cf/=, !=, &lt;, &gt;, &lt;=, &gt;=/), but
	you can't concatenate two strings. String literals are written as
	<cf/"This is a string constant"/. Additionally matching (<cf/&tilde;,
	!&tilde;/) operators could be used to match a string value against
	a shell pattern (represented also as a string).

	<tag><label id="type-bytestring">bytestring</tag>
	This is a sequences of arbitrary bytes. There are no ways to modify
	bytestrings in filters. You can pass them between function, assign
	them to variables of type <cf/bytestring/, print such values,
	compare bytestings (<cf/=, !=/). Bytestring literals are written
	in several ways:

	A sequence of hexadecimal digit pairs, optionally colon-separated.
	A bytestring specified this way must be at least 16 bytes (32 digits)
	long: <cf/01:23:45:67:89:ab:cd:ef:01:23:45:67:89:ab:cd:ef/ or
	<cf/0123456789abcdef0123456789abcdef/.

	A sequence of hexadecimal digit pairs of any lengh (including zero)
	with the <cf/hex:/ prefix. Colon separators can be inserted
	arbitrarily between any bytes: <cf/hex:/, <cf/hex:1234/,
	<cf/hex:1234:56:78/.

	A bytestring can be made from a hex string using <cf/from_hex()/
	function. Source strings can use any number of colons, hyphens and
	spaces as byte separators: <cf/from_hex(" 1234 56:78 ab-cd-ef ")/.

	<tag><label id="type-ip">ip</tag>
	This type can hold a single IP address. The IPv4 addresses are stored as
	IPv4-Mapped IPv6 addresses so one data type for both of them is used.
	Whether the address is IPv4 or not may be checked by <cf>.is_v4</cf>
	which returns a <cf/bool/. IP addresses are written in the standard
	notation (<cf/10.20.30.40/ or <cf/fec0:3:4::1/). You can apply special
	operator <cf>.mask(<M>num</M>)</cf> on values of type ip. It masks out
	all but first <cf><M>num</M></cf> bits from the IP address. So
	<cf/1.2.3.4.mask(8) = 1.0.0.0/ is true.

	<tag><label id="type-prefix">prefix</tag>
	This type can hold a network prefix consisting of IP address, prefix
	length and several other values. This is the key in route tables.

	Prefixes may be of several types, which can be determined by the special
	operator <cf/.type/. The type may be:

	<cf/NET_IP4/ and <cf/NET_IP6/ prefixes hold an IP prefix. The literals
	are written as <cf><m/ipaddress//<m/pxlen/</cf>. There are two special
	operators on these: <cf/.ip/ which extracts the IP address from the
	pair, and <cf/.len/, which separates prefix length from the pair.
	So <cf>1.2.0.0/16.len = 16</cf> is true.

	<cf/NET_IP6_SADR/ nettype holds both destination and source IPv6
	prefix. The literals are written as <cf><m/ipaddress//<m/pxlen/ from
	<m/ipaddress//<m/pxlen/</cf>, where the first part is the destination
	prefix and the second art is the source prefix. They support the same
	operators as IP prefixes, but just for the destination part. They also
	support <cf/.src/ and <cf/.dst/ operators to get respective parts of the
	address as separate <cf/NET_IP6/ values.

	<cf/NET_VPN4/ and <cf/NET_VPN6/ prefixes hold an IP prefix with VPN
	Route Distinguisher (<rfc id="4364">). They support the same special
	operators as IP prefixes, and also <cf/.rd/ which extracts the Route
	Distinguisher. Their literals are written
	as <cf><m/vpnrd/ <m/ipprefix/</cf>

	<cf/NET_ROA4/ and <cf/NET_ROA6/ prefixes hold an IP prefix range
	together with an ASN. They support the same special operators as IP
	prefixes, and also <cf/.maxlen/ which extracts maximal prefix length,
	and <cf/.asn/ which extracts the ASN.

	<cf/NET_FLOW4/ and <cf/NET_FLOW6/ hold an IP prefix together with a
	flowspec rule. Filters currently do not support much flowspec parsing,
	only <cf/.src/ and <cf/.dst/ operators to get source and destination
	parts of the flowspec as separate <cf/NET_IP4/ / <cf/NET_IP6/ values.

	<cf/NET_MPLS/ holds a single MPLS label and its handling is currently
	not implemented.

	<tag><label id="type-vpnrd">vpnrd</tag>
	This is a route distinguisher according to <rfc id="4364">. There are
	three kinds of RD's: <cf><m/asn/:<m/32bit int/</cf>, <cf><m/asn4/:<m/16bit int/</cf>
	and <cf><m/IPv4 address/:<m/32bit int/</cf>

	<tag><label id="type-ec">ec</tag>
	This is a specialized type used to represent BGP extended community
	values. It is essentially a 64bit value, literals of this type are
	usually written as <cf>(<m/kind/, <m/key/, <m/value/)</cf>, where
	<cf/kind/ is a kind of extended community (e.g. <cf/rt/ / <cf/ro/ for a
	route target / route origin communities), the format and possible values
	of <cf/key/ and <cf/value/ are usually integers, but it depends on the
	used kind. Similarly to pairs, ECs can be constructed using expressions
	for <cf/key/ and <cf/value/ parts, (e.g. <cf/(ro, myas, 3*10)/, where
	<cf/myas/ is an integer variable).

	<tag><label id="type-lc">lc</tag>
	This is a specialized type used to represent BGP large community
	values. It is essentially a triplet of 32bit values, where the first
	value is reserved for the AS number of the issuer, while meaning of
	remaining parts is defined by the issuer. Literals of this type are
	written as <cf/(123, 456, 789)/, with any integer values. Similarly to
	pairs, LCs can be constructed using expressions for its parts, (e.g.
	<cf/(myas, 10+20, 3*10)/, where <cf/myas/ is an integer variable).

	Operators <cf/.asn/, <cf/.data1/, and <cf/.data2/ can be used
	to extract corresponding components of LCs:
	<cf>(<m/asn/, <m/data1/, <m/data2/)</cf>.

	<tag><label id="type-set">int|pair|quad|ip|prefix|ec|lc|enum set</tag>
	Filters recognize four types of sets. Sets are similar to strings: you
	can pass them around but you can't modify them. Literals of type <cf>int
	set</cf> look like <cf> [ 1, 2, 5..7 ]</cf>. As you can see, both simple
	values and ranges are permitted in sets.

	For pair sets, expressions like <cf/(123,*)/ can be used to denote
	ranges (in that case <cf/(123,0)..(123,65535)/). You can also use
	<cf/(123,5..100)/ for range <cf/(123,5)..(123,100)/. You can also use
	<cf/*/ and <cf/a..b/ expressions in the first part of a pair, note that
	such expressions are translated to a set of intervals, which may be
	memory intensive. E.g. <cf/(*,4..20)/ is translated to <cf/(0,4..20),
	(1,4..20), (2,4..20), ... (65535, 4..20)/.

	EC sets use similar expressions like pair sets, e.g. <cf/(rt, 123,
	10..20)/ or <cf/(ro, 123, *)/. Expressions requiring the translation
	(like <cf/(rt, *, 3)/) are not allowed (as they usually have 4B range
	for ASNs).

	Also LC sets use similar expressions like pair sets. You can use ranges
	and wildcards, but if one field uses that, more specific (later) fields
	must be wildcards. E.g., <cf/(10, 20..30, *)/ or <cf/(10, 20, 30..40)/
	is valid, while <cf/(10, *, 20..30)/ or <cf/(10, 20..30, 40)/ is not
	valid.

	You can also use expressions for int, pair, EC and LC set values.
	However, it must be possible to evaluate these expressions before daemon
	boots. So you can use only constants inside them. E.g.

	<code>
	 define one=1;
	 define myas=64500;
	 int set odds;
	 pair set ps;
	 ec set es;

	 odds = [ one, 2+1, 6-one, 2*2*2-1, 9, 11 ];
	 ps = [ (1,one+one), (3,4)..(4,8), (5,*), (6,3..6), (7..9,*) ];
	 es = [ (rt, myas, 3*10), (rt, myas+one, 0..16*16*16-1), (ro, myas+2, *) ];
	</code>

	Sets of prefixes are special: their literals does not allow ranges, but
	allows prefix patterns that are written
	as <cf><M>ipaddress</M>/<M>pxlen</M>{<M>low</M>,<M>high</M>}</cf>.
	Prefix <cf><m>ip1</m>/<m>len1</m></cf> matches prefix
	pattern <cf><m>ip2</m>/<m>len2</m>{<m>l</m>,<m>h</m>}</cf> if the
	first <cf>min(len1, len2)</cf> bits of <cf/ip1/ and <cf/ip2/ are
	identical and <cf>len1 &lt;= ip1 &lt;= len2</cf>. A valid prefix pattern
	has to satisfy <cf>low &lt;= high</cf>, but <cf/pxlen/ is not
	constrained by <cf/low/ or <cf/high/. Obviously, a prefix matches a
	prefix set literal if it matches any prefix pattern in the prefix set
	literal.

	There are also two shorthands for prefix patterns: <cf><m/address//<m/len/+</cf>
	is a shorthand for <cf><m/address//<m/len/{<m/len/,<m/maxlen/}</cf>
	(where <cf><m/maxlen/</cf> is 32 for IPv4 and 128 for IPv6), that means
	network prefix <cf><m/address//<m/len/</cf> and all its	subnets.
	<cf><m/address//<m/len/-</cf> is a shorthand for
	<cf><m/address//<m/len/{0,<m/len/}</cf>, that means network prefix
	<cf><m/address//<m/len/</cf> and all its supernets (network prefixes
	that contain it).

	For example, <cf>[ 1.0.0.0/8, 2.0.0.0/8+, 3.0.0.0/8-, 4.0.0.0/8{16,24}
	]</cf> matches prefix <cf>1.0.0.0/8</cf>, all subprefixes of
	<cf>2.0.0.0/8</cf>, all superprefixes of <cf>3.0.0.0/8</cf> and prefixes
	<cf/4.X.X.X/ whose prefix length is 16 to 24. <cf>[ 0.0.0.0/0{20,24} ]</cf>
	matches all prefixes (regardless of IP address) whose prefix length is
	20 to 24, <cf>[ 1.2.3.4/32- ]</cf> matches any prefix that contains IP
	address <cf>1.2.3.4</cf>. <cf>1.2.0.0/16 &tilde; [ 1.0.0.0/8{15,17} ]</cf>
	is true, but <cf>1.0.0.0/16 &tilde; [ 1.0.0.0/8- ]</cf> is false.

	Cisco-style patterns like <cf>10.0.0.0/8 ge 16 le 24</cf> can be expressed
	in BIRD as <cf>10.0.0.0/8{16,24}</cf>, <cf>192.168.0.0/16 le 24</cf> as
	<cf>192.168.0.0/16{16,24}</cf> and <cf>192.168.0.0/16 ge 24</cf> as
	<cf>192.168.0.0/16{24,32}</cf>.

	It is not possible to mix IPv4 and IPv6 prefixes in a prefix set. It is
	currently possible to mix IPv4 and IPv6 addresses in an ip set, but that
	behavior may change between versions without any warning; don't do it
	unless you are more than sure what you are doing. (Really, don't do it.)

	<tag><label id="type-enum">enum</tag>
	Enumeration types are fixed sets of possibilities. You can't define your
	own variables of such type, but some route attributes are of enumeration
	type. Enumeration types are incompatible with each other.

	<tag><label id="type-bgppath">bgppath</tag>
	BGP path is a list of autonomous system numbers. You can't write
	literals of this type. There are several special operators on bgppaths:

	<cf><m/P/.first</cf> returns the first ASN (the neighbor ASN) in path <m/P/.

	<cf><m/P/.last</cf> returns the last ASN (the source ASN) in path <m/P/.

	<cf><m/P/.last_nonaggregated</cf> returns the last ASN in the non-aggregated part of the path <m/P/.

	Both <cf/first/ and <cf/last/ return zero if there is no appropriate
	ASN, for example if the path contains an AS set element as the first (or
	the last) part. If the path ends with an AS set, <cf/last_nonaggregated/
	may be used to get last ASN before any AS set.

	<cf><m/P/.len</cf> returns the length of path <m/P/.

	<cf><m/P/.empty</cf> makes the path <m/P/ empty. Can't be used as a value, always modifies the object.

	<cf><m/P/.prepend(<m/A/)</cf> prepends ASN <m/A/ to path <m/P/ and
	returns the result.

	<cf><m/P/.delete(<m/A/)</cf> deletes all instances of ASN <m/A/ from
	from path <m/P/ and returns the result. <m/A/ may also be an integer
	set, in that case the operator deletes all ASNs from path <m/P/ that are
	also members of set <m/A/.

	<cf><m/P/.filter(<m/A/)</cf> deletes all ASNs from path <m/P/ that are
	not members of integer set <m/A/, and returns the result.
	I.e., <cf/filter/ do the same as <cf/delete/ with inverted set <m/A/.

	Methods <cf>prepend</cf>, <cf>delete</cf> and <cf>filter</cf> keep the
	original object intact as long as you use the result in any way. You can
	also write e.g. <cf><m/P/.prepend(<m/A/);</cf> as a standalone statement.
	This variant does modify the original object with the result of the operation.

	<tag><label id="type-bgpmask">bgpmask</tag>
	BGP masks are patterns used for BGP path matching (using <cf>path
	&tilde; [= 2 3 5 * =]</cf> syntax). The masks resemble wildcard patterns
	as used by UNIX shells. Autonomous system numbers match themselves,
	<cf/*/ matches any (even empty) sequence of arbitrary AS numbers and
	<cf/?/ matches one arbitrary AS number. For example, if <cf>bgp_path</cf>
 	is 4 3 2 1, then: <tt>bgp_path &tilde; [= * 4 3 * =]</tt> is true,
	but <tt>bgp_path &tilde; [= * 4 5 * =]</tt> is false. There is also
	<cf/+/ operator which matches one or multiple instances of previous
	expression, e.g. <tt>[= 1 2+ 3 =]</tt> matches both path 1 2 3 and path
	1 2 2 2 3, but not 1 3 nor 1 2 4 3. Note that while <cf/*/ and <cf/?/
	are wildcard-style operators, <cf/+/ is regex-style operator.

	BGP mask expressions can also contain integer expressions enclosed in
	parenthesis and integer variables, for example <tt>[= * 4 (1+2) a =]</tt>.
	You can also use ranges (e.g. <tt>[= * 3..5 2 100..200 * =]</tt>)
	and sets (e.g. <tt>[= 1 2 [3, 5, 7] * =]</tt>).

	<tag><label id="type-clist">clist</tag>
	Clist is similar to a set, except that unlike other sets, it can be
	modified. The type is used for community list (a set of pairs) and for
	cluster list (a set of quads). There exist no literals of this type.
	There are special operators on clists:

	<cf><m/C/.len</cf> returns the length of clist <m/C/.

	<cf><m/C/.empty</cf> makes the list <m/C/ empty. Can't be used as a value, always modifies the object.

	<cf><m/C/.add(<m/P/)</cf> adds pair (or quad) <m/P/ to clist <m/C/ and
	returns the result. If item <m/P/ is already in clist <m/C/, it does
	nothing. <m/P/ may also be a clist, in that case all its members are
	added; i.e., it works as clist union.

	<cf><m/C/.delete(<m/P/)</cf> deletes pair (or quad) <m/P/ from clist
	<m/C/ and returns the result. If clist <m/C/ does not contain item
	<m/P/, it does nothing. <m/P/ may also be a pair (or quad) set, in that
	case the operator deletes all items from clist <m/C/ that are also
	members of set <m/P/. Moreover, <m/P/ may also be a clist, which works
	analogously; i.e., it works as clist difference.

	<cf><m/C/.filter(<m/P/)</cf> deletes all items from clist <m/C/ that are
	not members of pair (or quad) set <m/P/, and returns the result. I.e., <cf/filter/ do the same
	as <cf/delete/ with inverted set <m/P/. <m/P/ may also be a clist, which
	works analogously; i.e., it works as clist intersection.

	Methods <cf>add</cf>, <cf>delete</cf> and <cf>filter</cf> keep the
	original object intact as long as you use the result in any way. You can
	also write e.g. <cf><m/P/.add(<m/A/);</cf> as a standalone statement.
	This variant does modify the original object with the result of the operation.

	<cf><m/C/.min</cf> returns the minimum element of clist <m/C/.

	<cf><m/C/.max</cf> returns the maximum element of clist <m/C/.

	Operators <cf/.min/, <cf/.max/ can be used together with <cf/filter/
	to extract the community from the specific subset of communities
	(e.g. localpref or prepend) without the need to check every possible
	value (e.g. <cf/filter(bgp_community, [(23456, 1000..1099)]).min/).

	<tag><label id="type-eclist">eclist</tag>
	Eclist is a data type used for BGP extended community lists. Eclists
	are very similar to clists, but they are sets of ECs instead of pairs.
	The same operations (like <cf/add/, <cf/delete/ or <cf/&tilde;/ and
	<cf/!&tilde;/ membership operators) can be used to modify or test
	eclists, with ECs instead of pairs as arguments.

	<tag><label id="type-lclist">lclist</tag>
	Lclist is a data type used for BGP large community lists. Like eclists,
	lclists are very similar to clists, but they are sets of LCs instead of
	pairs. The same operations (like <cf/add/, <cf/delete/ or <cf/&tilde;/
	and <cf/!&tilde;/ membership operators) can be used to modify or test
	lclists, with LCs instead of pairs as arguments.
</descrip>


<sect>Operators
<label id="operators">

<p>The filter language supports common integer operators <cf>(+,-,*,/)</cf>,
parentheses <cf/(a*(b+c))/, comparison <cf/(a=b, a!=b, a&lt;b, a&gt;=b)/.
Logical operations include unary not (<cf/!/), and (<cf/&amp;&amp;/), and or
(<cf/&verbar;&verbar;/). Special operators include (<cf/&tilde;/,
<cf/!&tilde;/) for "is (not) element of a set" operation - it can be used on
element and set of elements of the same type (returning true if element is
contained in the given set), or on two strings (returning true if first string
matches a shell-like pattern stored in second string) or on IP and prefix
(returning true if IP is within the range defined by that prefix), or on prefix
and prefix (returning true if first prefix is more specific than second one) or
on bgppath and bgpmask (returning true if the path matches the mask) or on
number and bgppath (returning true if the number is in the path) or on bgppath
and int (number) set (returning true if any ASN from the path is in the set) or
on pair/quad and clist (returning true if the pair/quad is element of the
clist) or on clist and pair/quad set (returning true if there is an element of
the clist that is also a member of the pair/quad set).

<p>There is one operator related to ROA infrastructure - <cf/roa_check()/. It
examines a ROA table and does <rfc id="6483"> route origin validation for a
given network prefix. The basic usage is <cf>roa_check(<m/table/)</cf>, which
checks the current route (which should be from BGP to have AS_PATH argument) in
the specified ROA table and returns ROA_UNKNOWN if there is no relevant ROA,
ROA_VALID if there is a matching ROA, or ROA_INVALID if there are some relevant
ROAs but none of them match. There is also an extended variant
<cf>roa_check(<m/table/, <m/prefix/, <m/asn/)</cf>, which allows to specify a
prefix and an ASN as arguments.


<sect>Control structures
<label id="control-structures">

<p>Filters support several control structures: conditions, for loops and case
switches.

<p>Syntax of a condition is: <cf>if <M>boolean expression</M> then <m/commandT/;
else <m/commandF/;</cf> and you can use <cf>{ <m/command1/; <m/command2/;
<M>...</M> }</cf> instead of either command. The <cf>else</cf> clause may be
omitted. If the <cf><m>boolean expression</m></cf> is true, <m/commandT/ is
executed, otherwise <m/commandF/ is executed.

<p>For loops allow to iterate over elements in compound data like BGP paths or
community lists. The syntax is: <cf>for [ <m/type/ ] <m/variable/ in <m/expr/
do <m/command/;</cf> and you can also use compound command like in conditions.
The expression is evaluated to a compound data, then for each element from such
data the command is executed with the item assigned to the variable. A variable
may be an existing one (when just name is used) or a locally defined (when type
and name is used). In both cases, it must have the same type as elements.

<p>The <cf>case</cf> is similar to case from Pascal. Syntax is <cf>case
<m/expr/ { else: | <m/num_or_prefix [ .. num_or_prefix]/: <m/statement/ ; [
... ] }</cf>. The expression after <cf>case</cf> can be of any type which can be
on the left side of the &tilde; operator and anything that could be a member of
a set is allowed before <cf/:/. Multiple commands are allowed without <cf/{}/
grouping. If <cf><m/expr/</cf> matches one of the <cf/:/ clauses, statements
between it and next <cf/:/ statement are executed. If <cf><m/expr/</cf> matches
neither of the <cf/:/ clauses, the statements after <cf/else:/ are executed.

<p>Here is example that uses <cf/if/ and <cf/case/ structures:

<code>
if 1234 = i then printn "."; else {
	print "not 1234";
	print "You need {} around multiple commands";
}

for int asn in bgp_path do {
	printn "ASN: ", asn;
	if asn < 65536 then print " (2B)"; else print " (4B)";
}

case arg1 {
	2: print "two"; print "I can do more commands without {}";
	3 .. 5: print "three to five";
	else: print "something else";
}
</code>


<sect>Route attributes
<label id="route-attributes">

<p>A filter is implicitly passed a route, and it can access its attributes just
like it accesses variables. There are common route attributes, protocol-specific
route attributes and custom route attributes. Most common attributes are
mandatory (always defined), while remaining are optional.  Attempts to access
undefined attribute result in a runtime error; you can check if an attribute is
defined by using the <cf>defined( <m>attribute</m> )</cf> operator. One notable
exception to this rule are attributes of bgppath and *clist types, where
undefined value is regarded as empty bgppath/*clist for most purposes.

Attributes can be defined by just setting them in filters. Custom attributes
have to be first declared by <ref id="opt-attribute" name="attribute"> global
option. You can also undefine optional attribute back to non-existence by using
the <cf>unset( <m/attribute/ )</cf> operator.

Common route attributes are:

<descrip>
	<tag><label id="rta-net"><m/prefix/ net</tag>
	The network prefix or anything else the route is talking about. The
	primary key of the routing table. Read-only. (See the <ref id="routes"
	name="chapter about routes">.)

	<tag><label id="rta-preference"><m/int/ preference</tag>
	Preference of the route.

	<tag><label id="rta-from"><m/ip/ from</tag>
	The router which the route has originated from.

	<tag><label id="rta-gw"><m/ip/ gw</tag>
	Next hop packets routed using this route should be forwarded to.

	<tag><label id="rta-proto"><m/string/ proto</tag>
	The name of the protocol which the route has been imported from.
	Read-only.

	<tag><label id="rta-source"><m/enum/ source</tag>
	what protocol has told me about this route. Possible values:
	<cf/RTS_STATIC/, <cf/RTS_INHERIT/, <cf/RTS_DEVICE/,
	<cf/RTS_RIP/, <cf/RTS_OSPF/, <cf/RTS_OSPF_IA/, <cf/RTS_OSPF_EXT1/,
	<cf/RTS_OSPF_EXT2/, <cf/RTS_BGP/, <cf/RTS_PIPE/, <cf/RTS_BABEL/.

	<tag><label id="rta-dest"><m/enum/ dest</tag>
	Type of destination the packets should be sent to
	(<cf/RTD_ROUTER/ for forwarding to a neighboring router,
	<cf/RTD_DEVICE/ for routing to a directly-connected network,
	<cf/RTD_MULTIPATH/ for multipath destinations,
	<cf/RTD_BLACKHOLE/ for packets to be silently discarded,
	<cf/RTD_UNREACHABLE/, <cf/RTD_PROHIBIT/ for packets that should be
	returned with ICMP host unreachable / ICMP administratively prohibited
	messages). Can be changed, but only to <cf/RTD_BLACKHOLE/,
	<cf/RTD_UNREACHABLE/ or <cf/RTD_PROHIBIT/.

	<tag><label id="rta-ifname"><m/string/ ifname</tag>
	Name of the outgoing interface. Sink routes (like blackhole, unreachable
	or prohibit) and multipath routes have no interface associated with
	them, so <cf/ifname/ returns an empty string for such routes. Setting it
	would also change route to a direct one (remove gateway).

	<tag><label id="rta-ifindex"><m/int/ ifindex</tag>
	Index of the outgoing interface. System wide index of the interface. May
	be used for interface matching, however indexes might change on interface
	creation/removal. Zero is returned for routes with undefined outgoing
	interfaces. Read-only.

	<tag><label id="rta-weight"><m/int/ weight</tag>
	Multipath weight of route next hops. Valid values are 1-256. Reading
	returns the weight of the first next hop, setting it sets weights of all
	next hops to the specified value. Therefore, this attribute is not much
	useful for manipulating individual next hops of an ECMP route, but can
	be used in BGP multipath setup to set weights of individual routes that
	are merged to one ECMP route during export to the Kernel protocol
	(with active <ref id="krt-merge-paths" name="marge paths"> option).

	<tag><label id="rta-gw-mpls"><m/int/ gw_mpls</tag>
	Outgoing MPLS label attached to route (i.e., incoming MPLS label on the
	next hop router for this label-switched path). Reading returns the label
	value and setting it sets it to the start of the label stack. Setting
	implicit-NULL label (3) disables the MPLS label stack. Only the first
	next hop and only one label in the label stack supported right now. This
	is experimental option, will be likely changed in the future to handle
	full MPLS label stack.

	<tag><label id="rta-igp-metric"><m/int/ igp_metric</tag>
	The optional attribute that can be used to specify a distance to the
	network for routes that do not have a native protocol metric attribute
	(like <cf/ospf_metric1/ for OSPF routes). It is used mainly by BGP to
	compare internal distances to boundary routers (see below).

	<tag><label id="rta-mpls-label"><m/int/ mpls_label</tag>
	Local MPLS label attached to the route. This attribute is produced by
	MPLS-aware protocols for labeled routes. It can also be set in import
	filters to assign static labels, but that also requires static MPLS
	label policy.

	<tag><label id="rta-mpls-policy"><m/enum/ mpls_policy</tag>
	For MPLS-aware protocols, this attribute defines which
	<ref id="mpls-channel-label-policy" name="MPLS label policy"> will be
	used for the route. It can be set in import filters to change it on
	per-route basis. Valid values are <cf/MPLS_POLICY_NONE/ (no label),
	<cf/MPLS_POLICY_STATIC/ (static label), <cf/MPLS_POLICY_PREFIX/
	(per-prefix label), <cf/MPLS_POLICY_AGGREGATE/ (aggregated label),
	and <cf/MPLS_POLICY_VRF/ (per-VRF label). See <ref
	id="mpls-channel-label-policy" name="MPLS label policy"> for details.

	<tag><label id="rta-mpls-class"><m/int/ mpls_class</tag>
	When <ref id="mpls-channel-label-policy" name="MPLS label policy"> is
	set to <cf/aggregate/, it may be useful to apply more fine-grained
	aggregation than just one based on next hops. When routes have different
	value of this attribute, they will not be aggregated under one local
	label even if they have the same next hops.
</descrip>

<p>Protocol-specific route attributes are described in the corresponding
protocol sections.


<sect>Other statements
<label id="other-statements">

<p>The following statements are available:

<descrip>
	<tag><label id="assignment"><m/variable/ = <m/expr/</tag>
	Set variable (or route attribute) to a given value.

	<tag><label id="filter-accept-reject">accept|reject [ <m/expr/ ]</tag>
	Accept or reject the route, possibly printing <cf><m>expr</m></cf>.

	<tag><label id="return">return <m/expr/</tag>
	Return <cf><m>expr</m></cf> from the current function, the function ends
	at this point.

	<tag><label id="print">print|printn <m/expr/ [<m/, expr.../]</tag>
	Prints given expressions; useful mainly while debugging filters. The
	<cf/printn/ variant does not terminate the line.
</descrip>


<chapt>Protocols
<label id="protocols">

<sect>Aggregator
<label id="aggregator">

<sect1>Introduction
<label id="aggregator-intro">
<p>The Aggregator protocol explicitly merges routes by the given rules. There
   are four phases of aggregation. First routes are filtered, then sorted into buckets,
   then buckets are merged and finally the results are filtered once again.
   Aggregating an already aggregated route is forbidden.

<p>This is an experimental protocol, use with caution.

<sect1>Configuration
<label id="aggregator-config">
<p><descrip>
	<tag><label id="aggregator-table">table <m/table/</tag>
	The table from which routes are exported to get aggregated.

	<tag><label id="aggregator-export">export <m/.../</tag>
	A standard channel's <cf/export/ clause, defining which routes are accepted into aggregation.

	<tag><label id="aggregator-rule">aggregate on <m/expr/ | <m/attribute/ [<m/, .../]</tag>
	All the given filter expressions and route attributes are evaluated for each route. Then routes
	are sorted into buckets where <em/all/ values are the same. Note: due to performance reasons,
        all filter expressions must return a compact type, e.g. integer, a BGP
        (standard, extended, large) community or an IP address. If you need to compare e.g. modified
	AS Paths in the aggregation rule, you can define a custom route attribute and set this attribute
	in the export filter. For now, it's mandatory to say <cf/net/ here, we can't merge prefixes yet.

	<tag><label id="aggregation-merge">merge by { <m/filter code/ }</tag>
	The given filter code has an extra symbol defined: <cf/routes/. By iterating over <cf/routes/,
	you get all the routes in the bucket and you can construct your new route. All attributes
	selected in <cf/aggregate on/ are already set to the common values. For now, it's not possible
	to use a named filter here. You have to finalize the route by calling <cf/accept/.

	<tag><label id="aggregator-import">import <m/.../</tag>
	Filter applied to the route after <cf/merge by/. Here you can use a named filter.

	<tag><label id="aggregator-peer-table">peer table <m/table/</tag>
	The table to which aggregated routes are imported. It may be the same table
	as <cf/table/.
</descrip>

<sect1>Example
<label id="aggregator-example">

<p><code>
protocol aggregator {
  table master6;
  export where defined(bgp_path);
  /* Merge all routes with the same AS Path length */
  aggregate on net, bgp_path.len;
  merge by {
    for route r in routes do {
      if ! defined(bgp_path) then { bgp_path = r.bgp_path }
      bgp_community = bgp_community.add(r.bgp_community);
    }
    accept;
  };
  import all;
  peer table agr_result;
}
</code>

<sect>Babel
<label id="babel">

<sect1>Introduction
<label id="babel-intro">

<p>The Babel protocol
(<rfc id="8966">) is a loop-avoiding distance-vector routing protocol that is
robust and efficient both in ordinary wired networks and in wireless mesh
networks. Babel is conceptually very simple in its operation and "just works"
in its default configuration, though some configuration is possible and in some
cases desirable.

<p>The Babel protocol is dual stack; i.e., it can carry both IPv4 and IPv6
routes over the same IPv6 transport. For sending and receiving Babel packets,
only a link-local IPv6 address is needed.

<p>BIRD implements an extension for IPv6 source-specific routing (SSR or SADR),
but must be configured accordingly to use it. SADR-enabled Babel router can
interoperate with non-SADR Babel router, but the later would ignore routes
with specific (non-zero) source prefix.

<sect1>Configuration
<label id="babel-config">

<p>The Babel protocol support both IPv4 and IPv6 channels; both can be
configured simultaneously. It can also be configured with <ref
id="ip-sadr-routes" name="IPv6 SADR"> channel instead of regular IPv6
channel, in such case SADR support is enabled. Babel supports no global
configuration options apart from those common to all other protocols, but
supports the following per-interface configuration options:

<code>
protocol babel [<name>] {
	ipv4 { <channel config> };
	ipv6 [sadr] { <channel config> };
        randomize router id <switch>;
	interface <interface pattern> {
		type <wired|wireless|tunnel>;
		rxcost <number>;
		limit <number>;
		hello interval <time>;
		update interval <time>;
		port <number>;
		tx class|dscp <number>;
		tx priority <number>;
		rx buffer <number>;
		tx length <number>;
		check link <switch>;
		next hop ipv4 <address>;
		next hop ipv6 <address>;
		extended next hop <switch>;
		rtt cost <number>;
		rtt min <time>;
		rtt max <time>;
		rtt decay <number>;
		send timestamps <switch>;
		authentication none|mac [permissive];
		password "&lt;text&gt;";
		password "&lt;text&gt;" {
			id &lt;num&gt;;
			generate from "&lt;date&gt;";
			generate to "&lt;date&gt;";
			accept from "&lt;date&gt;";
			accept to "&lt;date&gt;";
			from "&lt;date&gt;";
			to "&lt;date&gt;";
			algorithm ( hmac sha1 | hmac sha256 | hmac sha384 |
	hmac sha512 | blake2s128 | blake2s256 | blake2b256 | blake2b512 );
		};
	};
}
</code>

<descrip>
      <tag><label id="babel-channel">ipv4 | ipv6 [sadr] <m/channel config/</tag>
      The supported channels are IPv4, IPv6, and IPv6 SADR.

      <tag><label id="babel-random-router-id">randomize router id <m/switch/</tag>
      If enabled, Bird will randomize the top 32 bits of its router ID whenever
      the protocol instance starts up. If a Babel node restarts, it loses its
      sequence number, which can cause its routes to be rejected by peers until
      the state is cleared out by other nodes in the network (which can take on
      the order of minutes). Enabling this option causes Bird to pick a random
      router ID every time it starts up, which avoids this problem at the cost
      of not having stable router IDs in the network. Default: no.

      <tag><label id="babel-type">type wired|wireless|tunnel </tag>
      This option specifies the interface type: Wired, wireless or tunnel. On
      wired interfaces a neighbor is considered unreachable after a small number
      of Hello packets are lost, as described by <cf/limit/ option. On wireless
      interfaces the ETX link quality estimation technique is used to compute
      the metrics of routes discovered over this interface. This technique will
      gradually degrade the metric of routes when packets are lost rather than
      the more binary up/down mechanism of wired type links. A tunnel is like a
      wired interface, but turns on RTT-based metrics with a default cost of 96.
      Default: <cf/wired/.

      <tag><label id="babel-rxcost">rxcost <m/num/</tag>
      This option specifies the nominal RX cost of the interface. The effective
      neighbor costs for route metrics will be computed from this value with a
      mechanism determined by the interface <cf/type/. Note that in contrast to
      other routing protocols like RIP or OSPF, the <cf/rxcost/ specifies the
      cost of RX instead of TX, so it affects primarily neighbors' route
      selection and not local route selection. Default: 96 for wired interfaces,
      256 for wireless.

      <tag><label id="babel-limit">limit <m/num/</tag>
      BIRD keeps track of received Hello messages from each neighbor to
      establish neighbor reachability. For wired type interfaces, this option
      specifies how many of last 16 hellos have to be correctly received in
      order to neighbor is assumed to be up. The option is ignored on wireless
      type interfaces, where gradual cost degradation is used instead of sharp
      limit. Default: 12.

      <tag><label id="babel-hello">hello interval <m/time/ s|ms</tag>
      Interval at which periodic Hello messages are sent on this interface,
      with time units. Default: 4 seconds.

      <tag><label id="babel-update">update interval <m/time/ s|ms</tag>
      Interval at which periodic (full) updates are sent, with time
      units. Default: 4 times the hello interval.

      <tag><label id="babel-port">port <m/number/</tag>
      This option selects an UDP port to operate on. The default is to operate
      on port 6696 as specified in the Babel RFC.

      <tag><label id="babel-tx-class">tx class|dscp|priority <m/number/</tag>
      These options specify the ToS/DiffServ/Traffic class/Priority of the
      outgoing Babel packets. See <ref id="proto-tx-class" name="tx class"> common
      option for detailed description.

      <tag><label id="babel-rx-buffer">rx buffer <m/number/</tag>
      This option specifies the size of buffers used for packet processing.
      The buffer size should be bigger than maximal size of received packets.
      The default value is the interface MTU, and the value will be clamped to a
      minimum of 512 bytes + IP packet overhead.

      <tag><label id="babel-tx-length">tx length <m/number/</tag>
      This option specifies the maximum length of generated Babel packets. To
      avoid IP fragmentation, it should not exceed the interface MTU value.
      The default value is the interface MTU value, and the value will be
      clamped to a minimum of 512 bytes + IP packet overhead.

      <tag><label id="babel-check-link">check link <m/switch/</tag>
      If set, the hardware link state (as reported by OS) is taken into
      consideration. When the link disappears (e.g. an ethernet cable is
      unplugged), neighbors are immediately considered unreachable and all
      routes received from them are withdrawn. It is possible that some
      hardware drivers or platforms do not implement this feature. Default:
      yes.

      <tag><label id="babel-next-hop-ipv4">next hop ipv4 <m/address/</tag>
      Set the next hop address advertised for IPv4 routes advertised on this
      interface. Default: the preferred IPv4 address of the interface.

      <tag><label id="babel-next-hop-ipv6">next hop ipv6 <m/address/</tag>
      Set the next hop address advertised for IPv6 routes advertised on this
      interface. If not set, the same link-local address that is used as the
      source for Babel packets will be used. In normal operation, it should not
      be necessary to set this option.

      <tag><label id="babel-extended-next-hop">extended next hop <m/switch/</tag>
      If enabled, BIRD will accept and emit IPv4 routes with an IPv6 next
      hop when IPv4 addresses are absent from the interface as described in
      <rfc id="9229">. Default: yes.

      <tag><label id="babel-rtt-cost">rtt cost <m/number/</tag>
      The RTT-based cost that will be applied to all routes from each neighbour
      based on the measured RTT to that neighbour. If this value is set,
      timestamps will be included in generated Babel Hello and IHU messages, and
      (if the neighbours also have timestamps enabled), the RTT to each
      neighbour will be computed. An additional cost is added to a neighbour if
      its RTT is above the <ref id="babel-rtt-min" name="rtt min"> value
      configured on the interface. The added cost scales linearly from 0 up to
      the RTT cost configured in this option; the full cost is applied if the
      neighbour RTT reaches the RTT configured in the <ref id="babel-rtt-max"
      name="rtt max"> option (and for all RTTs above this value). Default: 0
      (disabled), except for tunnel interfaces, where it is 96.

      <tag><label id="babel-rtt-min">rtt min <m/time/ s|ms</tag>
      The minimum RTT above which the RTT cost will start to be applied (scaling
      linearly from zero up to the full cost). Default: 10 ms

      <tag><label id="babel-rtt-max">rtt max <m/time/ s|ms</tag>
      The maximum RTT above which the full RTT cost will start be applied.
      Default: 120 ms

      <tag><label id="babel-rtt-decay">rtt decay <m/number/</tag>
      The decay factor used for the exponentional moving average of the RTT
      samples from each neighbour, in units of 1/256. Higher values discards old
      RTT samples faster. Must be between 1 and 256. Default: 42

      <tag><label id="babel-send-timestamps">send timestamps <m/switch/</tag>
      Whether to send the timestamps used for RTT calculation on this interface.
      Sending the timestamps enables peers to calculate an RTT to this node,
      even if no RTT cost is applied to the route metrics. Default: yes.

      <tag><label id="babel-authentication">authentication none|mac [permissive]</tag>
      Selects authentication method to be used. <cf/none/ means that packets
      are not authenticated at all, <cf/mac/ means MAC authentication is
      performed as described in <rfc id="8967">. If MAC authentication is
      selected, the <cf/permissive/ suffix can be used to select an operation
      mode where outgoing packets are signed, but incoming packets will be
      accepted even if they fail authentication. This can be useful for
      incremental deployment of MAC authentication across a network. If MAC
      authentication is selected, a key must be specified with the
      <cf/password/ configuration option. Default: none.

      <tag><label id="babel-password">password "<m/text/"</tag>
      Specifies a password used for authentication. See the <ref id="proto-pass"
      name="password"> common option for a detailed description. The Babel
      protocol will only accept HMAC-based algorithms or one of the Blake
      algorithms, and the length of the supplied password string must match the
      key size used by the selected algorithm.
</descrip>

<sect1>Attributes
<label id="babel-attr">

<p>Babel defines just one attribute: the internal babel metric of the route. It
is exposed as the <cf/babel_metric/ attribute and has range from 1 to infinity
(65535).

<sect1>Example
<label id="babel-exam">

<p><code>
protocol babel {
	interface "eth*" {
		type wired;
	};
	interface "wlan0", "wlan1" {
		type wireless;
		hello interval 1;
		rxcost 512;
	};
	interface "tap0";

	# This matches the default of babeld: redistribute all addresses
	# configured on local interfaces, plus re-distribute all routes received
	# from other babel peers.

	ipv4 {
		export where (source = RTS_DEVICE) || (source = RTS_BABEL);
	};
	ipv6 {
		export where (source = RTS_DEVICE) || (source = RTS_BABEL);
	};
}
</code>

<sect1>Known issues
<label id="babel-issues">

<p>When retracting a route, Babel generates an unreachable route for a little
while (according to RFC). The interaction of this behavior with other protocols
is not well tested and strange things may happen.


<sect>BFD
<label id="bfd">

<sect1>Introduction
<label id="bfd-intro">

<p>Bidirectional Forwarding Detection (BFD) is not a routing protocol itself, it
is an independent tool providing liveness and failure detection. Routing
protocols like OSPF and BGP use integrated periodic "hello" messages to monitor
liveness of neighbors, but detection times of these mechanisms are high (e.g. 40
seconds by default in OSPF, could be set down to several seconds). BFD offers
universal, fast and low-overhead mechanism for failure detection, which could be
attached to any routing protocol in an advisory role.

<p>BFD consists of mostly independent BFD sessions. Each session monitors an
unicast bidirectional path between two BFD-enabled routers. This is done by
periodically sending control packets in both directions. BFD does not handle
neighbor discovery, BFD sessions are created on demand by request of other
protocols (like OSPF or BGP), which supply appropriate information like IP
addresses and associated interfaces. When a session changes its state, these
protocols are notified and act accordingly (e.g. break an OSPF adjacency when
the BFD session went down).

<p>BIRD implements basic BFD behavior as defined in <rfc id="5880"> (some
advanced features like the echo mode or authentication are not implemented), IP
transport for BFD as defined in <rfc id="5881"> and <rfc id="5883"> and
interaction with client protocols as defined in <rfc id="5882">.

<p>BFD packets are sent with a dynamic source port number. Linux systems use by
default a bit different dynamic port range than the IANA approved one
(49152-65535). If you experience problems with compatibility, please adjust
<cf>/proc/sys/net/ipv4/ip_local_port_range</cf>.

<sect1>Configuration
<label id="bfd-config">

<p>BFD configuration consists mainly of multiple definitions of interfaces.
Most BFD config options are session specific. When a new session is requested
and dynamically created, it is configured from one of these definitions. For
sessions to directly connected neighbors, <cf/interface/ definitions are chosen
based on the interface associated with the session, while <cf/multihop/
definition is used for multihop sessions. If no definition is relevant, the
session is just created with the default configuration. Therefore, an empty BFD
configuration is often sufficient.

<p>Note that to use BFD for other protocols like OSPF or BGP, these protocols
also have to be configured to request BFD sessions, usually by <cf/bfd/ option.
In BGP case, it is also possible to specify per-peer BFD session options (e.g.
rx/tx intervals) as a part of the <cf/bfd/ option.

<p>A BFD instance not associated with any VRF handles session requests from all
other protocols, even ones associated with a VRF. Such setup would work for
single-hop BFD sessions if <cf/net.ipv4.udp_l3mdev_accept/ sysctl is enabled,
but does not currently work for multihop sessions. Another approach is to
configure multiple BFD instances, one for each VRF (including the default VRF).
Each BFD instance associated with a VRF (regular or default) only handles
session requests from protocols in the same VRF.

<p>Some of BFD session options require <m/time/ value, which has to be specified
with the appropriate unit: <m/num/ <cf/s/|<cf/ms/|<cf/us/. Although microseconds
are allowed as units, practical minimum values are usually in order of tens of
milliseconds.

<code>
protocol bfd [&lt;name&gt;] {
	accept [ipv4|ipv6] [direct|multihop];
	interface &lt;interface pattern&gt; {
		interval &lt;time&gt;;
		min rx interval &lt;time&gt;;
		min tx interval &lt;time&gt;;
		idle tx interval &lt;time&gt;;
		multiplier &lt;num&gt;;
		passive &lt;switch&gt;;
		authentication none;
		authentication simple;
		authentication [meticulous] keyed md5|sha1;
		password "&lt;text&gt;";
		password "&lt;text&gt;" {
			id &lt;num&gt;;
			generate from "&lt;date&gt;";
			generate to "&lt;date&gt;";
			accept from "&lt;date&gt;";
			accept to "&lt;date&gt;";
			from "&lt;date&gt;";
			to "&lt;date&gt;";
		};
	};
	multihop {
		interval &lt;time&gt;;
		min rx interval &lt;time&gt;;
		min tx interval &lt;time&gt;;
		idle tx interval &lt;time&gt;;
		multiplier &lt;num&gt;;
		passive &lt;switch&gt;;
	};
	neighbor &lt;ip&gt; [dev "&lt;interface&gt;"] [local &lt;ip&gt;] [multihop &lt;switch&gt;];
}
</code>

<descrip>
	<tag><label id="bfd-accept">accept [ipv4|ipv6] [direct|multihop]</tag>
	A BFD protocol instance accepts (by default) all BFD session requests
	(with regard to VRF restrictions, see above). This option controls
	whether IPv4 / IPv6 and direct / multihop session requests are accepted
	(and which listening sockets are opened). It can be used, for example,
	to configure separate BFD protocol instances for IPv4 and for IPv6
	sessions.

	<tag><label id="bfd-strict-bind">strict bind <m/switch/</tag>
	Specify whether each BFD interface should use a separate listening
	socket bound to its local address, or just use a shared listening socket
	accepting all addresses. Binding to a specific address could be useful
	in cases like running multiple BIRD instances on a machine, each
	handling a different set of interfaces. Default: disabled.

	<tag><label id="bfd-iface">interface <m/pattern/ [, <m/.../] { <m/options/ }</tag>
	Interface definitions allow to specify options for sessions associated
	with such interfaces and also may contain interface specific options.
	See <ref id="proto-iface" name="interface"> common option for a detailed
	description of interface patterns. Note that contrary to the behavior of
	<cf/interface/ definitions of other protocols, BFD protocol would accept
	sessions (in default configuration) even on interfaces not covered by
	such definitions.

	<tag><label id="bfd-multihop">multihop { <m/options/ }</tag>
	Multihop definitions allow to specify options for multihop BFD sessions,
	in the same manner as <cf/interface/ definitions are used for directly
	connected sessions. Currently only one such definition (for all multihop
	sessions) could be used.

	<tag><label id="bfd-neighbor">neighbor <m/ip/ [dev "<m/interface/"] [local <m/ip/] [multihop <m/switch/]</tag>
	BFD sessions are usually created on demand as requested by other
	protocols (like OSPF or BGP). This option allows to explicitly add
	a BFD session to the specified neighbor regardless of such requests.

	The session is identified by the IP address of the neighbor, with
	optional specification of used interface and local IP. By default
	the neighbor must be directly connected, unless the session is
	configured as multihop. Note that local IP must be specified for
	multihop sessions.
</descrip>

<p>Session specific options (part of <cf/interface/ and <cf/multihop/ definitions):

<descrip>
	<tag><label id="bfd-interval">interval <m/time/</tag>
	BFD ensures availability of the forwarding path associated with the
	session by periodically sending BFD control packets in both
	directions. The rate of such packets is controlled by two options,
	<cf/min rx interval/ and <cf/min tx interval/ (see below). This option
	is just a shorthand to set both of these options together.

	<tag><label id="bfd-min-rx-interval">min rx interval <m/time/</tag>
	This option specifies the minimum RX interval, which is announced to the
	neighbor and used there to limit the neighbor's rate of generated BFD
	control packets. Default: 10 ms.

	<tag><label id="bfd-min-tx-interval">min tx interval <m/time/</tag>
	This option specifies the desired TX interval, which controls the rate
	of generated BFD control packets (together with <cf/min rx interval/
	announced by the neighbor). Note that this value is used only if the BFD
	session is up, otherwise the value of <cf/idle tx interval/ is used
	instead. Default: 100 ms.

	<tag><label id="bfd-idle-tx-interval">idle tx interval <m/time/</tag>
	In order to limit unnecessary traffic in cases where a neighbor is not
	available or not running BFD, the rate of generated BFD control packets
	is lower when the BFD session is not up. This option specifies the
	desired TX interval in such cases instead of <cf/min tx interval/.
	Default: 1 s.

	<tag><label id="bfd-multiplier">multiplier <m/num/</tag>
	Failure detection time for BFD sessions is based on established rate of
	BFD control packets (<cf>min rx/tx interval</cf>) multiplied by this
	multiplier, which is essentially (ignoring jitter) a number of missed
	packets after which the session is declared down. Note that rates and
	multipliers could be different in each direction of a BFD session.
	Default: 5.

	<tag><label id="bfd-passive">passive <m/switch/</tag>
	Generally, both BFD session endpoints try to establish the session by
	sending control packets to the other side. This option allows to enable
	passive mode, which means that the router does not send BFD packets
	until it has received one from the other side. Default: disabled.

	<tag>authentication none</tag>
	No passwords are sent in BFD packets. This is the default value.

	<tag>authentication simple</tag>
	Every packet carries 16 bytes of password. Received packets lacking this
	password are ignored. This authentication mechanism is very weak.

	<tag>authentication [meticulous] keyed md5|sha1</tag>
	An authentication code is appended to each packet. The cryptographic
	algorithm is keyed MD5 or keyed SHA-1. Note that the algorithm is common
	for all keys (on one interface), in contrast to OSPF or RIP, where it
	is a per-key option. Passwords (keys) are not sent open via network.

	The <cf/meticulous/ variant means that cryptographic sequence numbers
	are increased for each sent packet, while in the basic variant they are
	increased about once per second. Generally, the <cf/meticulous/ variant
	offers better resistance to replay attacks but may require more
	computation.

	<tag>password "<M>text</M>"</tag>
	Specifies a password used for authentication. See <ref id="proto-pass"
	name="password"> common option for detailed description. Note that
	password option <cf/algorithm/ is not available in BFD protocol. The
	algorithm is selected by <cf/authentication/ option for all passwords.

</descrip>

<sect1>Example
<label id="bfd-exam">

<p><code>
protocol bfd {
	interface "eth*" {
		min rx interval 20 ms;
		min tx interval 50 ms;
		idle tx interval 300 ms;
	};
	interface "gre*" {
		interval 200 ms;
		multiplier 10;
		passive;
	};
	multihop {
		interval 200 ms;
		multiplier 10;
	};

	neighbor 192.168.1.10;
	neighbor 192.168.2.2 dev "eth2";
	neighbor 192.168.10.1 local 192.168.1.1 multihop;
}
</code>


<sect>BGP
<label id="bgp">

<p>The Border Gateway Protocol is the routing protocol used for backbone level
routing in the today's Internet. Contrary to other protocols, its convergence
does not rely on all routers following the same rules for route selection,
making it possible to implement any routing policy at any router in the network,
the only restriction being that if a router advertises a route, it must accept
and forward packets according to it.

<p>BGP works in terms of autonomous systems (often abbreviated as AS). Each AS
is a part of the network with common management and common routing policy. It is
identified by a unique 16-bit number (ASN). Routers within each AS usually
exchange AS-internal routing information with each other using an interior
gateway protocol (IGP, such as OSPF or RIP). Boundary routers at the border of
the AS communicate global (inter-AS) network reachability information with their
neighbors in the neighboring AS'es via exterior BGP (eBGP) and redistribute
received information to other routers in the AS via interior BGP (iBGP).

<p>Each BGP router sends to its neighbors updates of the parts of its routing
table it wishes to export along with complete path information (a list of AS'es
the packet will travel through if it uses the particular route) in order to
avoid routing loops.

<sect1>Supported standards
<label id="bgp-standards">

<p>
<itemize>
<item> <rfc id="4271"> - Border Gateway Protocol 4 (BGP)
<item> <rfc id="1997"> - BGP Communities Attribute
<item> <rfc id="2385"> - Protection of BGP Sessions via TCP MD5 Signature
<item> <rfc id="2545"> - Use of BGP Multiprotocol Extensions for IPv6
<item> <rfc id="2918"> - Route Refresh Capability
<item> <rfc id="3107"> - Carrying Label Information in BGP
<item> <rfc id="4360"> - BGP Extended Communities Attribute
<item> <rfc id="4364"> - BGP/MPLS IPv4 Virtual Private Networks
<item> <rfc id="4456"> - BGP Route Reflection
<item> <rfc id="4486"> - Subcodes for BGP Cease Notification Message
<item> <rfc id="4659"> - BGP/MPLS IPv6 Virtual Private Networks
<item> <rfc id="4724"> - Graceful Restart Mechanism for BGP
<item> <rfc id="4760"> - Multiprotocol extensions for BGP
<item> <rfc id="4798"> - Connecting IPv6 Islands over IPv4 MPLS
<item> <rfc id="5065"> - AS confederations for BGP
<item> <rfc id="5082"> - Generalized TTL Security Mechanism
<item> <rfc id="5492"> - Capabilities Advertisement with BGP
<item> <rfc id="5575"> - Dissemination of Flow Specification Rules
<item> <rfc id="5668"> - 4-Octet AS Specific BGP Extended Community
<item> <rfc id="6286"> - AS-Wide Unique BGP Identifier
<item> <rfc id="6608"> - Subcodes for BGP Finite State Machine Error
<item> <rfc id="6793"> - BGP Support for 4-Octet AS Numbers
<item> <rfc id="7311"> - Accumulated IGP Metric Attribute for BGP
<item> <rfc id="7313"> - Enhanced Route Refresh Capability for BGP
<item> <rfc id="7606"> - Revised Error Handling for BGP UPDATE Messages
<item> <rfc id="7911"> - Advertisement of Multiple Paths in BGP
<item> <rfc id="7947"> - Internet Exchange BGP Route Server
<item> <rfc id="8092"> - BGP Large Communities Attribute
<item> <rfc id="8203"> - BGP Administrative Shutdown Communication
<item> <rfc id="8212"> - Default EBGP Route Propagation Behavior without Policies
<item> <rfc id="8654"> - Extended Message Support for BGP
<item> <rfc id="8950"> - Advertising IPv4 NLRI with an IPv6 Next Hop
<item> <rfc id="9072"> - Extended Optional Parameters Length for BGP OPEN Message
<item> <rfc id="9117"> - Revised Validation Procedure for BGP Flow Specifications
<item> <rfc id="9234"> - Route Leak Prevention and Detection Using Roles
</itemize>

<sect1>Route selection rules
<label id="bgp-route-select-rules">

<p>BGP doesn't have any simple metric, so the rules for selection of an optimal
route among multiple BGP routes with the same preference are a bit more complex
and they are implemented according to the following algorithm. It starts the
first rule, if there are more "best" routes, then it uses the second rule to
choose among them and so on.

<itemize>
	<item>Prefer route with the highest Local Preference attribute.
	<item>Prefer route with the shortest AS path.
	<item>Prefer IGP origin over EGP and EGP origin over incomplete.
	<item>Prefer the lowest value of the Multiple Exit Discriminator.
	<item>Prefer routes received via eBGP over ones received via iBGP.
	<item>Prefer routes with lower internal distance to a boundary router.
	<item>Prefer the route with the lowest value of router ID of the
	advertising router.
</itemize>

<sect1>IGP routing table
<label id="bgp-igp-routing-table">

<p>BGP is mainly concerned with global network reachability and with routes to
other autonomous systems. When such routes are redistributed to routers in the
AS via BGP, they contain IP addresses of a boundary routers (in route attribute
NEXT_HOP). BGP depends on existing IGP routing table with AS-internal routes to
determine immediate next hops for routes and to know their internal distances to
boundary routers for the purpose of BGP route selection. In BIRD, there is
usually one routing table used for both IGP routes and BGP routes.

<sect1>Protocol configuration
<label id="bgp-proto-config">

<p>Each instance of the BGP corresponds to one neighboring router. This allows
to set routing policy and all the other parameters differently for each neighbor
using the following configuration parameters:

<descrip>
	<tag><label id="bgp-local">local [<m/ip/] [port <m/number/] [as <m/number/]</tag>
	Define which AS we are part of. (Note that contrary to other IP routers,
	BIRD is able to act as a router located in multiple AS'es simultaneously,
	but in such cases you need to tweak the BGP paths manually in the filters
	to get consistent behavior.) Optional <cf/ip/ argument specifies a source
	address, equivalent to the <cf/source address/ option (see below).
	Optional <cf/port/ argument specifies the local BGP port instead of
	standard port 179. The parameter may be used multiple times with
	different sub-options (e.g., both <cf/local 10.0.0.1 as 65000;/ and
	<cf/local 10.0.0.1; local as 65000;/ are valid). This parameter is
	mandatory.

	<tag><label id="bgp-neighbor">neighbor [<m/ip/ | range <m/prefix/] [port <m/number/] [as <m/number/] [internal|external]</tag>
	Define neighboring router this instance will be talking to and what AS
	it is located in. In case the neighbor is in the same AS as we are, we
	automatically switch to IBGP. Alternatively, it is possible to specify
	just <cf/internal/ or <cf/external/ instead of AS number, in that case
	either local AS number, or any external AS number is accepted.
	Optionally, the remote port may also be specified. Like <cf/local/
	parameter, this parameter may also be used multiple times with different
	sub-options. This parameter is mandatory.

	It is possible to specify network prefix (with <cf/range/ keyword)
	instead of explicit neighbor IP address. This enables dynamic BGP
	behavior, where the BGP instance listens on BGP port, but new BGP
	instances are spawned for incoming BGP connections (if source address
	matches the network prefix). It is possible to mix regular BGP instances
	with dynamic BGP instances and have multiple dynamic BGP instances with
	different ranges.

	<tag><label id="bgp-iface">interface <m/string/</tag>
	Define interface we should use for link-local BGP IPv6 sessions.
	Interface can also be specified as a part of <cf/neighbor address/
	(e.g., <cf/neighbor fe80::1234%eth0 as 65000;/). The option may also be
	used for non link-local sessions when it is necessary to explicitly
	specify an interface, but only for direct (not multihop) sessions.

	<tag><label id="bgp-direct">direct</tag>
	Specify that the neighbor is directly connected. The IP address of the
	neighbor must be from a directly reachable IP range (i.e. associated
	with one of your router's interfaces), otherwise the BGP session
	wouldn't start but it would wait for such interface to appear. The
	alternative is the <cf/multihop/ option. Default: enabled for eBGP.

	<tag><label id="bgp-multihop">multihop [<m/number/]</tag>
	Configure multihop BGP session to a neighbor that isn't directly
	connected. Accurately, this option should be used if the configured
	neighbor IP address does not match with any local network subnets. Such
	IP address have to be reachable through system routing table. The
	alternative is the <cf/direct/ option. For multihop BGP it is
	recommended to explicitly configure the source address to have it
	stable. Optional <cf/number/ argument can be used to specify the number
	of hops (used for TTL). Note that the number of networks (edges) in a
	path is counted; i.e., if two BGP speakers are separated by one router,
	the number of hops is 2. Default: enabled for iBGP.

	<tag><label id="bgp-source-address">source address <m/ip/</tag>
	Define local address we should use as a source address for the BGP
	session. Default: the address of the local end of the interface our
	neighbor is connected to.

	<tag><label id="bgp-dynamic-name">dynamic name "<m/text/"</tag>
	Define common prefix of names used for new BGP instances spawned when
	dynamic BGP behavior is active. Actual names also contain numeric
	index to distinguish individual instances.  Default: "dynbgp".

	<tag><label id="bgp-dynamic-name-digits">dynamic name digits <m/number/</tag>
	Define minimum number of digits for index in names of spawned dynamic
	BGP instances. E.g., if set to 2, then the first name would be
	"dynbgp01". Default: 0.

	<tag><label id="bgp-strict-bind">strict bind <m/switch/</tag>
	Specify whether BGP listening socket should be bound to a specific local
	address (the same as the <cf/source address/) and associated interface,
	or to all addresses. Binding to a specific address could be useful in
	cases like running multiple BIRD instances on a machine, each using its
	IP address. Note that listening sockets bound to a specific address and
	to all addresses collide, therefore either all BGP protocols (of the
	same address family and using the same local port) should have set
	<cf/strict bind/, or none of them. Default: disabled.

	<tag><label id="bgp-free-bind">free bind <m/switch/</tag>
	Use IP_FREEBIND socket option for the listening socket, which allows
	binding to an IP address not (yet) assigned to an interface. Note that
	all BGP instances that share a listening socket should have the same
	value of the <cf/freebind/ option. Default: disabled.

	<tag><label id="bgp-check-link">check link <M>switch</M></tag>
	BGP could use hardware link state into consideration.  If enabled,
	BIRD tracks the link state of the associated interface and when link
	disappears (e.g. an ethernet cable is unplugged), the BGP session is
	immediately shut down. Note that this option cannot be used with
	multihop BGP. Default: enabled for direct BGP, disabled otherwise.

	<tag><label id="bgp-bfd">bfd <M>switch</M>|graceful| { <m/options/ }</tag>
	BGP could use BFD protocol as an advisory mechanism for neighbor
	liveness and failure detection. If enabled, BIRD setups a BFD session
	for the BGP neighbor and tracks its liveness by it. This has an
	advantage of an order of magnitude lower detection times in case of
	failure. When a neighbor failure is detected, the BGP session is
	restarted. Optionally, it can be configured (by <cf/graceful/ argument)
	to trigger graceful restart instead of regular restart. It is also
	possible to specify section with per-peer BFD session options instead of
	just switch argument. Most BFD session specific options are allowed here
	with the exception of authentication options. here Note that BFD
	protocol also has to be configured, see <ref id="bfd" name="BFD">
	section for details. Default: disabled.

	<tag><label id="bgp-ttl-security">ttl security <m/switch/</tag>
	Use GTSM (<rfc id="5082"> - the generalized TTL security mechanism). GTSM
	protects against spoofed packets by ignoring received packets with a
	smaller than expected TTL. To work properly, GTSM have to be enabled on
	both sides of a BGP session. If both <cf/ttl security/ and
	<cf/multihop/ options are enabled, <cf/multihop/ option should specify
	proper hop value to compute expected TTL. Kernel support required:
	Linux: 2.6.34+ (IPv4), 2.6.35+ (IPv6), BSD: since long ago, IPv4 only.
	Note that full (ICMP protection, for example) <rfc id="5082"> support is
	provided by Linux only. Default: disabled.

	<tag><label id="bgp-password">password <m/string/</tag>
	Use this password for MD5 authentication of BGP sessions (<rfc id="2385">). When
	used on BSD systems, see also <cf/setkey/ option below. Default: no
	authentication.

	<tag><label id="bgp-setkey">setkey <m/switch/</tag>
	On BSD systems, keys for TCP MD5 authentication are stored in the global
	SA/SP database, which can be accessed by external utilities (e.g.
	setkey(8)). BIRD configures security associations in the SA/SP database
	automatically based on <cf/password/ options (see above), this option
	allows to disable automatic updates by BIRD when manual configuration by
	external utilities is preferred. Note that automatic SA/SP database
	updates are currently implemented only for FreeBSD. Passwords have to be
	set manually by an external utility on NetBSD and OpenBSD. Default:
	enabled (ignored on non-FreeBSD).

	<tag><label id="bgp-passive">passive <m/switch/</tag>
	Standard BGP behavior is both initiating outgoing connections and
	accepting incoming connections. In passive mode, outgoing connections
	are not initiated. Default: off.

	<tag><label id="bgp-confederation">confederation <m/number/</tag>
	BGP confederations (<rfc id="5065">) are collections of autonomous
	systems that act as one entity to external systems, represented by one
	confederation identifier (instead of AS numbers). This option allows to
	enable BGP confederation behavior and to specify the local confederation
	identifier. When BGP confederations are used, all BGP speakers that are
	members of the BGP confederation should have the same confederation
	identifier configured. Default: 0 (no confederation).

	<tag><label id="bgp-confederation-member">confederation member <m/switch/</tag>
	When BGP confederations are used, this option allows to specify whether
	the BGP neighbor is a member of the same confederation as the local BGP
	speaker. The option is unnecessary (and ignored) for IBGP sessions, as
	the same AS number implies the same confederation. Default: no.

	<tag><label id="bgp-rr-client">rr client</tag>
	Be a route reflector and treat the neighbor as a route reflection
	client. Default: disabled.

	<tag><label id="bgp-rr-cluster-id">rr cluster id <m/IPv4 address/</tag>
	Route reflectors use cluster id to avoid route reflection loops. When
	there is one route reflector in a cluster it usually uses its router id
	as a cluster id, but when there are more route reflectors in a cluster,
	these need to be configured (using this option) to use a common cluster
	id. Clients in a cluster need not know their cluster id and this option
	is not allowed for them. Default: the same as router id.

	<tag><label id="bgp-rs-client">rs client</tag>
	Be a route server and treat the neighbor as a route server client.
	A route server is used as a replacement for full mesh EBGP routing in
	Internet exchange points in a similar way to route reflectors used in
	IBGP routing. BIRD does not implement obsoleted <rfc id="1863">, but
	uses ad-hoc implementation, which behaves like plain EBGP but reduces
	modifications to advertised route attributes to be transparent (for
	example does not prepend its AS number to AS PATH attribute and
	keeps MED attribute). Default: disabled.

	<tag><label id="bgp-allow-local-pref">allow bgp_local_pref <m/switch/</tag>
	Standard BGP implementations do not send the Local Preference attribute
	to EBGP neighbors and ignore this attribute if received from EBGP
	neighbors, as per <rfc id="4271">.  When this option is enabled on an
	EBGP session, this attribute will be sent to and accepted from the peer,
	which is useful for example if you have a setup like in <rfc id="7938">.
	The option does not affect IBGP sessions. Default: off.

	<tag><label id="bgp-allow-med">allow bgp_med <m/switch/</tag>
	Standard BGP implementations do not propagate the MULTI_EXIT_DESC
	attribute unless it is configured locally. When this option is enabled
	on an EBGP session, this attribute will be sent to the peer regardless,
	which is useful for example if you have a setup like in <rfc id="7938">.
	The option does not affect IBGP sessions. Default: off.

	<tag><label id="bgp-allow-local-as">allow local as [<m/number/]</tag>
	BGP prevents routing loops by rejecting received routes with the local
	AS number in the AS path. This option allows to loose or disable the
	check. Optional <cf/number/ argument can be used to specify the maximum
	number of local ASNs in the AS path that is allowed for received
	routes. When the option is used without the argument, the check is
	completely disabled and you should ensure loop-free behavior by some
	other means. Default: 0 (no local AS number allowed).

	<tag><label id="bgp-allow-as-sets">allow as sets [<m/switch/]</tag>
	AS path attribute received with BGP routes may contain not only
	sequences of AS numbers, but also sets of AS numbers. These rarely used
	artifacts are results of inter-AS route aggregation. AS sets are
	deprecated (<rfc id="6472">), and likely to be rejected in the future,
	as they complicate security features like RPKI validation. When this
	option is disabled, then received AS paths with AS sets are rejected as
	malformed and corresponding BGP updates are treated as withdraws.
	Default: on.

	<tag><label id="bgp-enforce-first-as">enforce first as [<m/switch/]</tag>
	Routes received from an EBGP neighbor are generally expected to have the
	first (leftmost) AS number in their AS path equal to the neighbor AS
	number. This is not enforced by default as there are legitimate cases
	where it is not true, e.g. connections to route servers. When this
	option is enabled, routes with non-matching first AS number are rejected
	and corresponding updates are treated as withdraws. The option is valid
	on EBGP sessions only. Default: off.

	<tag><label id="bgp-enable-route-refresh">enable route refresh <m/switch/</tag>
	After the initial route exchange, BGP protocol uses incremental updates
	to keep BGP speakers synchronized. Sometimes (e.g., if BGP speaker
	changes its import filter, or if there is suspicion of inconsistency) it
	is necessary to do a new complete route exchange. BGP protocol extension
	Route Refresh (<rfc id="2918">) allows BGP speaker to request
	re-advertisement of all routes from its neighbor. This option
	specifies whether BIRD advertises this capability and supports
	related procedures. Note that even when disabled, BIRD can send route
	refresh requests. Disabling Route Refresh also disables Enhanced Route Refresh.
	Default: on.

	<tag><label id="bgp-enable-enhanced-route-refresh">enable enhanced route refresh <m/switch/</tag>
	BGP protocol extension Enhanced Route Refresh (<rfc id="7313">) specifies explicit
	begin and end for Route Refresh (see previous option),
	therefore the receiver can remove
	stale routes that were not advertised during the exchange. This option
	specifies whether BIRD advertises this capability and supports
	related procedures.
	Default: on.

	<tag><label id="bgp-graceful-restart">graceful restart <m/switch/|aware</tag>
	When a BGP speaker restarts or crashes, neighbors will discard all
	received paths from the speaker, which disrupts packet forwarding even
	when the forwarding plane of the speaker remains intact. <rfc id="4724">
	specifies an optional graceful restart mechanism to alleviate this
	issue. This option controls the mechanism. It has three states:
	Disabled, when no support is provided. Aware, when the graceful restart
	support is announced and the support for restarting neighbors is
	provided, but no local graceful restart is allowed (i.e. receiving-only
	role). Enabled, when the full graceful restart support is provided
	(i.e. both restarting and receiving role). Restarting role could be also
	configured per-channel. Note that proper support for local graceful
	restart requires also configuration of other protocols. Default: aware.

	<tag><label id="bgp-graceful-restart-time">graceful restart time <m/number/</tag>
	The restart time is announced in the BGP graceful restart capability
	and specifies how long the neighbor would wait for the BGP session to
	re-establish after a restart before deleting stale routes. Default:
	120 seconds.

	<tag><label id="bgp-long-lived-graceful-restart">long lived graceful restart <m/switch/|aware</tag>
	The long-lived graceful restart is an extension of the traditional
	<ref id="bgp-graceful-restart" name="BGP graceful restart">, where stale
	routes are kept even after the <ref id="bgp-graceful-restart-time"
	name="restart time"> expires for additional long-lived stale time, but
	they are marked with the LLGR_STALE community, depreferenced, and
	withdrawn from routers not supporting LLGR. Like traditional BGP
	graceful restart, it has three states: disabled, aware (receiving-only),
	and enabled. Note that long-lived graceful restart requires at least
	aware level of traditional BGP graceful restart. Default: aware, unless
	graceful restart is disabled.

	<tag><label id="bgp-long-lived-stale-time">long lived stale time <m/number/</tag>
	The long-lived stale time is announced in the BGP long-lived graceful
	restart capability and specifies how long the neighbor would keep stale
	routes depreferenced during long-lived graceful restart until either the
	session is re-stablished and synchronized or the stale time expires and
	routes are removed. Default: 3600 seconds.

	<tag><label id="bgp-interpret-communities">interpret communities <m/switch/</tag>
	<rfc id="1997"> demands that BGP speaker should process well-known
	communities like no-export (65535, 65281) or no-advertise (65535,
	65282). For example, received route carrying a no-adverise community
	should not be advertised to any of its neighbors. If this option is
	enabled (which is by default), BIRD has such behavior automatically (it
	is evaluated when a route is exported to the BGP protocol just before
	the export filter).  Otherwise, this integrated processing of
	well-known communities is disabled. In that case, similar behavior can
	be implemented in the export filter.  Default: on.

	<tag><label id="bgp-enable-as4">enable as4 <m/switch/</tag>
	BGP protocol was designed to use 2B AS numbers and was extended later to
	allow 4B AS number. BIRD supports 4B AS extension, but by disabling this
	option it can be persuaded not to advertise it and to maintain old-style
	sessions with its neighbors. This might be useful for circumventing bugs
	in neighbor's implementation of 4B AS extension. Even when disabled
	(off), BIRD behaves internally as AS4-aware BGP router. Default: on.

	<tag><label id="bgp-enable-extended-messages">enable extended messages <m/switch/</tag>
	The BGP protocol uses maximum message length of 4096 bytes. This option
	provides an extension (<rfc id="8654">) to allow extended messages with
	length up to 65535 bytes. Default: off.

	<tag><label id="bgp-capabilities">capabilities <m/switch/</tag>
	Use capability advertisement to advertise optional capabilities. This is
	standard behavior for newer BGP implementations, but there might be some
	older BGP implementations that reject such connection attempts. When
	disabled (off), features that request it (4B AS support) are also
	disabled. Default: on, with automatic fallback to off when received
	capability-related error.

	<tag><label id="bgp-advertise-hostname">advertise hostname <m/switch/</tag>
	Advertise hostname capability along with the hostname. Default: off.

	<tag><label id="bgp-disable-after-error">disable after error <m/switch/</tag>
	When an error is encountered (either locally or by the other side),
	disable the instance automatically and wait for an administrator to fix
	the problem manually. Default: off.

	<tag><label id="bgp-disable-after-cease">disable after cease <m/switch/|<m/set-of-flags/</tag>
	When a Cease notification is received, disable the instance
	automatically and wait for an administrator to fix the problem manually.
	When used with <m/switch/ argument, it means handle every Cease subtype
	with the exception of <cf/connection collision/. Default: off.

	The <m/set-of-flags/ allows to narrow down relevant Cease subtypes. The
	syntax is <cf>{<m/flag/ [, <m/.../] }</cf>, where flags are: <cf/cease/,
	<cf/prefix limit hit/, <cf/administrative shutdown/,
	<cf/peer deconfigured/, <cf/administrative reset/,
	<cf/connection rejected/, <cf/configuration change/,
	<cf/connection collision/, <cf/out of resources/.

	<tag><label id="bgp-hold-time">hold time <m/number/</tag>
	Time in seconds to wait for a Keepalive message from the other side
	before considering the connection stale. The effective value is
	negotiated during session establishment and it is a minimum of this
	configured value and the value proposed by the peer. The zero value has
	a special meaning, signifying that no keepalives are used. Default: 240
	seconds.

	<tag><label id="bgp-min-hold-time">min hold time <m/number/</tag>
	Minimum value of the hold time that is accepted during session negotiation.
	If the peer proposes a lower value, the session is rejected with error.
	Default: none.

	<tag><label id="bgp-startup-hold-time">startup hold time <m/number/</tag>
	Value of the hold timer used before the routers have a chance to exchange
	open messages and agree on the real value. Default: 240	seconds.

	<tag><label id="bgp-keepalive-time">keepalive time <m/number/</tag>
	Delay in seconds between sending of two consecutive Keepalive messages.
	The effective value depends on the negotiated hold time, as it is scaled
	to maintain proportion between the keepalive time and the hold time.
	Default: One third of the hold time.

	<tag><label id="bgp-min-keepalive-time">min keepalive time <m/number/</tag>
	Minimum value of the keepalive time that is accepted during session
	negotiation. If the proposed hold time would lead to a lower value of
	the keepalive time, the session is rejected with error. Default: none.

	<tag><label id="bgp-connect-delay-time">connect delay time <m/number/</tag>
	Delay in seconds between protocol startup and the first attempt to
	connect. Default: 5 seconds.

	<tag><label id="bgp-connect-retry-time">connect retry time <m/number/</tag>
	Time in seconds to wait before retrying a failed attempt to connect.
	Default: 120 seconds.

	<tag><label id="bgp-error-wait-time">error wait time <m/number/,<m/number/</tag>
	Minimum and maximum delay in seconds between a protocol failure (either
	local or reported by the peer) and automatic restart. Doesn not apply
	when <cf/disable after error/ is configured. If consecutive errors
	happen, the delay is increased exponentially until it reaches the
	maximum. Default: 60, 300.

	<tag><label id="bgp-error-forget-time">error forget time <m/number/</tag>
	Maximum time in seconds between two protocol failures to treat them as a
	error sequence which makes <cf/error wait time/ increase exponentially.
	Default: 300 seconds.

	<tag><label id="bgp-path-metric">path metric <m/switch/</tag>
	Enable comparison of path lengths when deciding which BGP route is the
	best one. Default: on.

	<tag><label id="bgp-med-metric">med metric <m/switch/</tag>
	Enable comparison of MED attributes (during best route selection) even
	between routes received from different ASes. This may be useful if all
	MED attributes contain some consistent metric, perhaps enforced in
	import filters of AS boundary routers. If this option is disabled, MED
	attributes are compared only if routes are received from the same AS
	(which is the standard behavior). Default: off.

	<tag><label id="bgp-deterministic-med">deterministic med <m/switch/</tag>
	BGP route selection algorithm is often viewed as a comparison between
	individual routes (e.g. if a new route appears and is better than the
	current best one, it is chosen as the new best one). But the proper
	route selection, as specified by <rfc id="4271">, cannot be fully
	implemented in that way. The problem is mainly in handling the MED
	attribute. BIRD, by default, uses an simplification based on individual
	route comparison, which in some cases may lead to temporally dependent
	behavior (i.e. the selection is dependent on the order in which routes
	appeared). This option enables a different (and slower) algorithm
	implementing proper <rfc id="4271"> route selection, which is
	deterministic. Alternative way how to get deterministic behavior is to
	use <cf/med metric/ option. This option is incompatible with <ref
	id="dsc-table-sorted" name="sorted tables">.  Default: off.

	<tag><label id="bgp-igp-metric">igp metric <m/switch/</tag>
	Enable comparison of internal distances to boundary routers during best
	route selection. Default: on.

	<tag><label id="bgp-prefer-older">prefer older <m/switch/</tag>
	Standard route selection algorithm breaks ties by comparing router IDs.
	This changes the behavior to prefer older routes (when both are external
	and from different peer). For details, see <rfc id="5004">. Default: off.

	<tag><label id="bgp-default-med">default bgp_med <m/number/</tag>
	Value of the Multiple Exit Discriminator to be used during route
	selection when the MED attribute is missing. Default: 0.

	<tag><label id="bgp-default-local-pref">default bgp_local_pref <m/number/</tag>
	A default value for the Local Preference attribute. It is used when
	a new Local Preference attribute is attached to a route by the BGP
	protocol itself (for example, if a route is received through eBGP and
	therefore does not have such attribute). Default: 100 (0 in pre-1.2.0
	versions of BIRD).

	<tag><label id="bgp-local-role">local role <m/role-name/</tag>
	BGP roles are a mechanism for route leak prevention and automatic route
	filtering based on common BGP topology relationships. They are defined
	in <rfc id="9234">. Instead of manually configuring filters and
	communities, automatic filtering is done with the help of the OTC
	attribute - a flag for routes that should be sent only to customers.
	The same attribute is also used to automatically detect and filter route
	leaks created by third parties.

	This option is valid for EBGP sessions, but it is not recommended to be
	used within AS confederations (which would require manual filtering of
	<cf/bgp_otc/ attribute on confederation boundaries).

	Possible <cf><m/role-name/</cf> values are: <cf/provider/,
	<cf/rs_server/, <cf/rs_client/, <cf/customer/ and <cf/peer/.
	Default: No local role assigned.

	<tag><label id="bgp-require-roles">require roles <m/switch/</tag>
	If this option is set, the BGP roles must be defined on both sides,
	otherwise the session will not be established. This behavior is defined
	in <rfc id="9234"> as "strict mode" and is used to enforce corresponding
	configuration at your conterpart side. Default: disabled.
</descrip>

<sect1>Channel configuration
<label id="bgp-channel-config">

<p>BGP supports several AFIs and SAFIs over one connection. Every AFI/SAFI
announced to the peer corresponds to one channel. The table of supported AFI/SAFIs
together with their appropriate channels follows.

<table loc="h">
<tabular ca="l|l|l|r|r">
  <bf/Channel name/   | <bf/Table nettype/ | <bf/IGP table allowed/  | <bf/AFI/ | <bf/SAFI/
@<hline>
  <cf/ipv4/	      | <cf/ipv4/          | <cf/ipv4/ and <cf/ipv6/ | 1        | 1
@ <cf/ipv6/           | <cf/ipv6/          | <cf/ipv4/ and <cf/ipv6/ | 2        | 1
@ <cf/ipv4 multicast/ | <cf/ipv4/          | <cf/ipv4/ and <cf/ipv6/ | 1        | 2
@ <cf/ipv6 multicast/ | <cf/ipv6/          | <cf/ipv4/ and <cf/ipv6/ | 2        | 2
@ <cf/ipv4 mpls/      | <cf/ipv4/          | <cf/ipv4/ and <cf/ipv6/ | 1        | 4
@ <cf/ipv6 mpls/      | <cf/ipv6/          | <cf/ipv4/ and <cf/ipv6/ | 2        | 4
@ <cf/vpn4 mpls/      | <cf/vpn4/          | <cf/ipv4/ and <cf/ipv6/ | 1        | 128
@ <cf/vpn6 mpls/      | <cf/vpn6/          | <cf/ipv4/ and <cf/ipv6/ | 2        | 128
@ <cf/vpn4 multicast/ | <cf/vpn4/          | <cf/ipv4/ and <cf/ipv6/ | 1        | 129
@ <cf/vpn6 multicast/ | <cf/vpn6/          | <cf/ipv4/ and <cf/ipv6/ | 2        | 129
@ <cf/flow4/	      | <cf/flow4/         | ---                     | 1        | 133
@ <cf/flow6/          | <cf/flow6/         | ---                     | 2        | 133
</tabular>
</table>

<p>The BGP protocol can be configured as MPLS-aware (by defining both AFI/SAFI
channels and the MPLS channel). In such case the BGP protocol assigns labels to
routes imported from MPLS-aware SAFIs (i.e. <cf/ipvX mpls/ and <cf/vpnX mpls/)
and automatically announces corresponding MPLS route for each labeled route. As
BGP generally processes a large amount of routes, it is suggested to set MPLS
label policy to <cf/aggregate/.

<p>Note that even BGP instances without MPLS channel and without local MPLS
configuration can still propagate third-party MPLS labels, e.g. as route
reflectors, they just will not assign local labels to imported routes and will
not announce MPLS routes for local MPLS forwarding.

<p>Due to <rfc id="8212">, external BGP protocol requires explicit configuration
of import and export policies (in contrast to other protocols, where default
policies of <cf/import all/ and <cf/export none/ are used in absence of explicit
configuration). Note that blanket policies like <cf/all/ or <cf/none/ can still
be used in explicit configuration.

<p>BGP channels have additional config options (together with the common ones):

<descrip>
	<tag><label id="bgp-mandatory">mandatory <m/switch/</tag>
	When local and neighbor sets of configured AFI/SAFI pairs differ,
	capability negotiation ensures that a common subset is used. For
	mandatory channels their associated AFI/SAFI must be negotiated
	(i.e., also announced by the neighbor), otherwise BGP session
	negotiation fails with <it/'Required capability missing'/ error.
	Regardless, at least one AFI/SAFI must be negotiated in order to BGP
	session be successfully established. Default: off.

	<tag><label id="bgp-next-hop-keep">next hop keep <m/switch/|ibgp|ebgp</tag>
	Do not modify the Next Hop attribute and advertise the current one
	unchanged even in cases where our own local address should be used
	instead. This is necessary when the BGP speaker does not forward network
	traffic (route servers and some route reflectors) and also can be useful
	in some other cases (e.g. multihop EBGP sessions). Can be enabled for
	all routes, or just for routes received from IBGP / EBGP neighbors.
	Default: disabled for regular BGP, enabled for route servers,
	<cf/ibgp/ for route reflectors.

	<tag><label id="bgp-next-hop-self">next hop self <m/switch/|ibgp|ebgp</tag>
	Always advertise our own local address as a next hop, even in cases
	where the current Next Hop attribute should be used unchanged. This is
	sometimes used for routes propagated from EBGP to IBGP when IGP routing
	does not cover inter-AS links, therefore IP addreses of EBGP neighbors
	are not resolvable through IGP. Can be enabled for all routes, or just
	for routes received from IBGP / EBGP neighbors. Default: disabled.

	<tag><label id="bgp-next-hop-address">next hop address <m/ip/</tag>
	Specify which address to use when our own local address should be
	announced in the Next Hop attribute. Default: the source address of the
	BGP session (if acceptable), or the preferred address of an associated
	interface.

	<tag><label id="bgp-next-hop-prefer">next hop prefer global</tag>
	Prefer global IPv6 address to link-local IPv6 address for immediate next
	hops of received routes. For IPv6 routes, the Next Hop attribute may
	contain both a global IP address and a link-local IP address. For IBGP
	sessions, the global IP address is resolved (<ref id="bgp-gateway"
	name="gateway recursive">) through an IGP routing table
	(<ref id="bgp-igp-table" name="igp table">) to get an immediate next
	hop. If the resulting IGP route is a direct route (i.e., the next hop is
	a direct neighbor), then the link-local IP address from the Next Hop
	attribute is used as the immediate next hop. This option change it to
	use the global IP address instead. Note that even with this option
	enabled a route may end with a link-local immediate next hop when the
	IGP route has one. Default: disabled.

	<tag><label id="bgp-gateway">gateway direct|recursive</tag>
	For received routes, their <cf/gw/ (immediate next hop) attribute is
	computed from received <cf/bgp_next_hop/ attribute. This option
	specifies how it is computed. Direct mode means that the IP address from
	<cf/bgp_next_hop/ is used and must be directly reachable. Recursive mode
	means that the gateway is computed by an IGP routing table lookup for
	the IP address from <cf/bgp_next_hop/. Note that there is just one level
	of indirection in recursive mode - the route obtained by the lookup must
	not be recursive itself, to prevent mutually recursive routes.

	Recursive mode is the behavior specified by the BGP
	standard. Direct mode is simpler, does not require any routes in a
	routing table, and was used in older versions of BIRD, but does not
	handle well nontrivial iBGP setups and multihop. Recursive mode is
	incompatible with <ref id="dsc-table-sorted" name="sorted tables">. Default:
	<cf/direct/ for direct sessions, <cf/recursive/ for multihop sessions.

	<tag><label id="bgp-igp-table">igp table <m/name/</tag>
	Specifies a table that is used as an IGP routing table. The type of this
	table must be as allowed in the table above. This option is allowed once
	for every allowed table type. Default: the same as the main table
	the channel is connected to (if eligible).

	<tag><label id="bgp-import-table">import table <m/switch/</tag>
	A BGP import table contains all received routes from given BGP neighbor,
	before application of import filters. It is also called <em/Adj-RIB-In/
	in BGP terminology. BIRD BGP by default operates without import tables,
	in which case received routes are just processed by import filters,
	accepted ones are stored in the master table, and the rest is forgotten.
	Enabling <cf/import table/ allows to store unprocessed routes, which can
	be examined later by <cf/show route/, and can be used to reconfigure
	import filters without full route refresh. Default: off.

	<tag><label id="bgp-export-table">export table <m/switch/</tag>
	A BGP export table contains all routes sent to given BGP neighbor, after
	application of export filters. It is also called <em/Adj-RIB-Out/ in BGP
	terminology. BIRD BGP by default operates without export tables, in
	which case routes from master table are just processed by export filters
	and then announced by BGP. Enabling <cf/export table/ allows to store
	routes after export filter processing, so they can be examined later by
	<cf/show route/, and can be used to eliminate unnecessary updates or
	withdraws. Default: off.

	<tag><label id="bgp-secondary">secondary <m/switch/</tag>
	Usually, if an export filter rejects a selected route, no other route is
	propagated for that network. This option allows to try the next route in
	order until one that is accepted is found or all routes for that network
	are rejected. This can be used for route servers that need to propagate
	different tables to each client but do not want to have these tables
	explicitly (to conserve memory). This option requires that the connected
	routing table is <ref id="dsc-table-sorted" name="sorted">. Default: off.

	<tag><label id="bgp-validate">validate <m/switch/</tag>
	Apply flowspec validation procedure as described in <rfc id="8955">
	section 6 and <rfc id="9117">. The Validation procedure enforces that
	only routers in the forwarding path for a network can originate flowspec
	rules for that network. The validation procedure should be used for EBGP
	to prevent injection of malicious flowspec rules from outside, but it
	should also be used for IBGP to ensure that selected flowspec rules are
	consistent with selected IP routes. The validation procedure uses an IP
	routing table (<ref id="bgp-base-table" name="base table">, see below)
	against which flowspec rules are validated. This option is limited to
	flowspec channels. Default: off (for compatibility reasons).

	Note that currently the flowspec validation does not work reliably
	together with <ref id="bgp-import-table" name="import table"> option
	enabled on flowspec channels.

	<tag><label id="bgp-base-table">base table <m/name/</tag>
	Specifies an IP table used for the flowspec validation procedure. The
	table must have enabled <cf/trie/ option, otherwise the validation
	procedure would not work. The type of the table must be <cf/ipv4/ for
	<cf/flow4/ channels and <cf/ipv6/ for <cf/flow6/ channels. This option
	is limited to flowspec channels. Default: the main table of the
	<cf/ipv4/ / <cf/ipv6/ channel of the same BGP instance, or the
	<cf/master4/ / <cf/master6/ table if there is no such channel.

	<tag><label id="bgp-extended-next-hop">extended next hop <m/switch/</tag>
	BGP expects that announced next hops have the same address family as
	associated network prefixes. This option provides an extension to use
	IPv4 next hops with IPv6 prefixes and vice versa. For IPv4 / VPNv4
	channels, the behavior is controlled by the Extended Next Hop Encoding
	capability, as described in <rfc id="8950">. For IPv6 / VPNv6 channels,
	just IPv4-mapped IPv6 addresses are used, as described in
	<rfc id="4798"> and <rfc id="4659">. Default: off.

	<tag><label id="bgp-add-paths">add paths <m/switch/|rx|tx</tag>
	Standard BGP can propagate only one path (route) per destination network
	(usually the selected one). This option controls the add-path protocol
	extension, which allows to advertise any number of paths to a
	destination. Note that to be active, add-path has to be enabled on both
	sides of the BGP session, but it could be enabled separately for RX and
	TX direction. When active, all available routes accepted by the export
	filter are advertised to the neighbor. Default: off.

	<tag><label id="bgp-aigp">aigp <m/switch/|originate</tag>
	The BGP protocol does not use a common metric like other routing
	protocols, instead it uses a set of criteria for route selection
	consisting both overall AS path length and a distance to the nearest AS
	boundary router. Assuming that metrics of different autonomous systems
	are incomparable, once a route is propagated from an AS to a next one,
	the distance in the old AS does not matter.

	The AIGP extension (<rfc id="7311">) allows to propagate accumulated
	IGP metric (in the AIGP attribute) through both IBGP and EBGP links,
	computing total distance through multiple autonomous systems (assuming
	they use comparable IGP metric). The total AIGP metric is compared in
	the route selection process just after Local Preference comparison (and
	before AS path length comparison).

	This option controls whether AIGP attribute propagation is allowed on
	the session. Optionally, it can be set to <cf/originate/, which not only
	allows AIGP attribute propagation, but also new AIGP attributes are
	automatically attached to non-BGP routes with valid IGP metric (e.g.
	<cf/ospf_metric1/) as they are exported to the BGP session. Default:
	enabled for IBGP (and intra-confederation EBGP), disabled for regular
	EBGP.

	<tag><label id="bgp-cost">cost <m/number/</tag>
	When BGP <ref id="bgp-gateway" name="gateway mode"> is <cf/recursive/
	(mainly multihop IBGP sessions), then the distance to BGP next hop is
	based on underlying IGP metric. This option specifies the distance to
	BGP next hop for BGP sessions in direct gateway mode (mainly direct
	EBGP sessions).

	<tag><label id="bgp-graceful-restart-c">graceful restart <m/switch/</tag>
	Although BGP graceful restart is configured mainly by protocol-wide
	<ref id="bgp-graceful-restart" name="options">, it is possible to
	configure restarting role per AFI/SAFI pair by this channel option.
	The option is ignored if graceful restart is disabled by protocol-wide
	option. Default: off in aware mode, on in full mode.

	<tag><label id="bgp-long-lived-graceful-restart-c">long lived graceful restart <m/switch/</tag>
	BGP long-lived graceful restart is configured mainly by protocol-wide
	<ref id="bgp-long-lived-graceful-restart" name="options">, but the
	restarting role can be set per AFI/SAFI pair by this channel option.
	The option is ignored if long-lived graceful restart is disabled by
	protocol-wide option. Default: off in aware mode, on in full mode.

	<tag><label id="bgp-long-lived-stale-time-c">long lived stale time <m/number/</tag>
	Like previous graceful restart channel options, this option allows to
	set <ref id="bgp-long-lived-stale-time" name="long lived stale time">
	per AFI/SAFI pair instead of per protocol. Default: set by protocol-wide
	option.
</descrip>

<sect1>Attributes
<label id="bgp-attr">

<p>BGP defines several route attributes. Some of them (those marked with
`<tt/I/' in the table below) are available on internal BGP connections only,
some of them (marked with `<tt/O/') are optional.

<descrip>
	<tag><label id="rta-bgp-path">bgppath bgp_path</tag>
	Sequence of AS numbers describing the AS path the packet will travel
	through when forwarded according to the particular route. In case of
	internal BGP it doesn't contain the number of the local AS.

	<tag><label id="rta-bgp-local-pref">int bgp_local_pref [I]</tag>
	Local preference value used for selection among multiple BGP routes (see
	the selection rules above). It's used as an additional metric which is
	propagated through the whole local AS.

	<tag><label id="rta-bgp-med">int bgp_med [O]</tag>
	The Multiple Exit Discriminator of the route is an optional attribute
	which is used on external (inter-AS) links to convey to an adjacent AS
	the optimal entry point into the local AS. The received attribute is
	also propagated over internal BGP links. The attribute value is zeroed
	when a route is exported to an external BGP instance to ensure that the
	attribute received from a neighboring AS is not propagated to other
	neighboring ASes. A new value might be set in the export filter of an
	external BGP instance. See <rfc id="4451"> for further discussion of
	BGP MED attribute.

	<tag><label id="rta-bgp-origin">enum bgp_origin</tag>
	Origin of the route: either <cf/ORIGIN_IGP/ if the route has originated
	in an interior routing protocol or <cf/ORIGIN_EGP/ if it's been imported
	from the <tt>EGP</tt> protocol (nowadays it seems to be obsolete) or
	<cf/ORIGIN_INCOMPLETE/ if the origin is unknown.

	<tag><label id="rta-bgp-next-hop">ip bgp_next_hop</tag>
	Next hop to be used for forwarding of packets to this destination. On
	internal BGP connections, it's an address of the originating router if
	it's inside the local AS or a boundary router the packet will leave the
	AS through if it's an exterior route, so each BGP speaker within the AS
	has a chance to use the shortest interior path possible to this point.

	<tag><label id="rta-bgp-atomic-aggr">void bgp_atomic_aggr [O]</tag>
	This is an optional attribute which carries no value, but the sole
	presence of which indicates that the route has been aggregated from
	multiple routes by some router on the path from the originator.

	<tag><label id="rta-bgp-aggregator">void bgp_aggregator [O]</tag>
	This is an optional attribute specifying AS number and IP address of the
	BGP router that created the route by aggregating multiple BGP routes.
	Currently, the attribute is not accessible from filters.

	<tag><label id="rta-bgp-community">clist bgp_community [O]</tag>
	List of community values associated with the route. Each such value is a
	pair (represented as a <cf/pair/ data type inside the filters) of 16-bit
	integers, the first of them containing the number of the AS which
	defines the community and the second one being a per-AS identifier.
	There are lots of uses of the community mechanism, but generally they
	are used to carry policy information like "don't export to USA peers".
	As each AS can define its own routing policy, it also has a complete
	freedom about which community attributes it defines and what will their
	semantics be.

	<tag><label id="rta-bgp-ext-community">eclist bgp_ext_community [O]</tag>
	List of extended community values associated with the route. Extended
	communities have similar usage as plain communities, but they have an
	extended range (to allow 4B ASNs) and a nontrivial structure with a type
	field. Individual community values are represented using an <cf/ec/ data
	type inside the filters.

	<tag><label id="rta-bgp-large-community">lclist bgp_large_community [O]</tag>
	List of large community values associated with the route. Large BGP
	communities is another variant of communities, but contrary to extended
	communities they behave very much the same way as regular communities,
	just larger -- they are uniform untyped triplets of 32bit numbers.
	Individual community values are represented using an <cf/lc/ data type
	inside the filters.

	<tag><label id="rta-bgp-originator-id">quad bgp_originator_id [I, O]</tag>
	This attribute is created by the route reflector when reflecting the
	route and contains the router ID of the originator of the route in the
	local AS.

	<tag><label id="rta-bgp-cluster-list">clist bgp_cluster_list [I, O]</tag>
	This attribute contains a list of cluster IDs of route reflectors. Each
	route reflector prepends its cluster ID when reflecting the route.

	<tag><label id="rta-bgp-aigp">void bgp_aigp [O]</tag>
	This attribute contains accumulated IGP metric, which is a total
	distance to the destination through multiple autonomous systems.
	Currently, the attribute is not accessible from filters.

	<tag><label id="bgp-otc">int bgp_otc [O]</tag>
	This attribute is defined in <rfc id="9234">. OTC is a flag that marks
	routes that should be sent only to customers. If <ref id="bgp-role"
	name="local Role"> is configured it set automatically.
</descrip>

<p>For attributes unknown by BIRD, the user can assign a name (on top level)
to an attribute by its number. This defined name can be used then to both set
(by a bytestring literal, transitive) or unset the given attribute even though
BIRD knows nothing about it:

<tt><label id="bgp-attribute-custom">attribute bgp <m/number/ bytestring <m/name/;</tt>

<sect1>Example
<label id="bgp-exam">

<p><code>
protocol bgp {
	local 198.51.100.14 as 65000;	     # Use a private AS number
	neighbor 198.51.100.130 as 64496;    # Our neighbor ...
	multihop;			     # ... which is connected indirectly
	ipv4 {
		export filter {			     # We use non-trivial export rules
			if source = RTS_STATIC then { # Export only static routes
				# Assign our community
				bgp_community.add((65000,64501));
				# Artificially increase path length
				# by advertising local AS number twice
				if bgp_path ~ [= 65000 =] then
					bgp_path.prepend(65000);
				accept;
			}
			reject;
		};
		import all;
		next hop self; # advertise this router as next hop
		igp table myigptable4; # IGP table for routes with IPv4 nexthops
		igp table myigptable6; # IGP table for routes with IPv6 nexthops
	};
	ipv6 {
		export filter mylargefilter; # We use a named filter
		import all;
		missing lladdr self;
		igp table myigptable4; # IGP table for routes with IPv4 nexthops
		igp table myigptable6; # IGP table for routes with IPv6 nexthops
	};
	ipv4 multicast {
		import all;
		export filter someotherfilter;
		table mymulticasttable4; # Another IPv4 table, dedicated for multicast
		igp table myigptable4;
	};
}
</code>


<sect>BMP
<label id="bmp">

<p>The BGP Monitoring Protocol is used for monitoring BGP sessions and obtaining
routing table data. The current implementation in BIRD is a preliminary release
with a limited feature set, it will be subject to significant changes in the
future. It is not ready for production usage and therefore it is not compiled
by default and have to be enabled during installation by the configure option
<tt/--with-protocols=/.

<p>The implementation is limited to monitor protocol state changes and routes
in <ref id="bgp-import-table" name="BGP import tables"> (not regular routing
tables), therefore import table must be enabled in BGP protocols. All BGP
protocols are monitored automatically.

<sect1>Example
<label id="bmp-exam">

<p><code>
protocol bmp {
	# The monitoring station to connect to
	station address ip 198.51.100.10 port 1790;

	# required option
	monitoring rib in pre_policy;
}
</code>


<sect>Device
<label id="device">

<p>The Device protocol is not a real routing protocol. It doesn't generate any
routes and it only serves as a module for getting information about network
interfaces from the kernel. This protocol supports no channel.

<p>Except for very unusual circumstances, you probably should include this
protocol in the configuration since almost all other protocols require network
interfaces to be defined for them to work with.

<sect1>Configuration
<label id="device-config">

<p><descrip>
	<tag><label id="device-scan-time">scan time <m/number/</tag>
	Time in seconds between two scans of the network interface list. On
	systems where we are notified about interface status changes
	asynchronously (such as newer versions of Linux), we need to scan the
	list only in order to avoid confusion by lost notification messages,
	so the default time is set to a large value.

	<tag><label id="device-iface">interface <m/pattern/ [, <m/.../]</tag>
	By default, the Device protocol handles all interfaces without any
	configuration. Interface definitions allow to specify optional
	parameters for specific interfaces. See <ref id="proto-iface"
	name="interface"> common option for detailed description. Currently only
	one interface option is available:

	<tag><label id="device-preferred">preferred <m/ip/</tag>
	If a network interface has more than one IP address, BIRD chooses one of
	them as a preferred one. Preferred IP address is used as source address
	for packets or announced next hop by routing protocols. Precisely, BIRD
	chooses one preferred IPv4 address, one preferred IPv6 address and one
	preferred link-local IPv6 address. By default, BIRD chooses the first
	found IP address as the preferred one.

	This option allows to specify which IP address should be preferred. May
	be used multiple times for different address classes (IPv4, IPv6, IPv6
	link-local). In all cases, an address marked by operating system as
	secondary cannot be chosen as the primary one.
</descrip>

<p>As the Device protocol doesn't generate any routes, it cannot have
any attributes. Example configuration looks like this:

<p><code>
protocol device {
	scan time 10;		# Scan the interfaces often
	interface "eth0" {
		preferred 192.168.1.1;
		preferred 2001:db8:1:10::1;
	};
}
</code>


<sect>Direct
<label id="direct">

<p>The Direct protocol is a simple generator of device routes for all the
directly connected networks according to the list of interfaces provided by the
kernel via the Device protocol. The Direct protocol supports both IPv4 and IPv6
channels; both can be configured simultaneously. It can also be configured with
<ref id="ip-sadr-routes" name="IPv6 SADR"> channel instead of regular IPv6
channel in order to be used together with SADR-enabled Babel protocol.

<p>The question is whether it is a good idea to have such device routes in BIRD
routing table. OS kernel usually handles device routes for directly connected
networks by itself so we don't need (and don't want) to export these routes to
the kernel protocol. OSPF protocol creates device routes for its interfaces
itself and BGP protocol is usually used for exporting aggregate routes. But the
Direct protocol is necessary for distance-vector protocols like RIP or Babel to
announce local networks.

<p>There are just few configuration options for the Direct protocol:

<p><descrip>
	<tag><label id="direct-iface">interface <m/pattern/ [, <m/.../]</tag>
	By default, the Direct protocol will generate device routes for all the
	interfaces available. If you want to restrict it to some subset of
	interfaces or addresses (e.g. if you're using multiple routing tables
	for policy routing and some of the policy domains don't contain all
	interfaces), just use this clause. See <ref id="proto-iface" name="interface">
	common option for detailed description. The Direct protocol uses
	extended interface clauses.

	<tag><label id="direct-check-link">check link <m/switch/</tag>
	If enabled, a hardware link state (reported by OS) is taken into
	consideration. Routes for directly connected networks are generated only
	if link up is reported and they are withdrawn when link disappears
	(e.g., an ethernet cable is unplugged). Default value is no.
</descrip>

<p>Direct device routes don't contain any specific attributes.

<p>Example config might look like this:

<p><code>
protocol direct {
	ipv4;
	ipv6;
	interface "-arc*", "*";		# Exclude the ARCnets
}
</code>


<sect>Kernel
<label id="krt">

<p>The Kernel protocol is not a real routing protocol. Instead of communicating
with other routers in the network, it performs synchronization of BIRD's routing
tables with the OS kernel. Basically, it sends all routing table updates to the
kernel and from time to time it scans the kernel tables to see whether some
routes have disappeared (for example due to unnoticed up/down transition of an
interface) or whether an `alien' route has been added by someone else (depending
on the <cf/learn/ switch, such routes are either ignored or accepted to our
table).

<p>Note that routes created by OS kernel itself, namely direct routes
representing IP subnets of associated interfaces, are not imported even with
<cf/learn/ enabled. You can use <ref id="direct" name="Direct protocol"> to
generate these direct routes.

<p>If your OS supports only a single routing table, you can configure only one
instance of the Kernel protocol. If it supports multiple tables (in order to
allow policy routing; such an OS is for example Linux), you can run as many
instances as you want, but each of them must be connected to a different BIRD
routing table and to a different kernel table.

<p>Because the kernel protocol is partially integrated with the connected
routing table, there are two limitations - it is not possible to connect more
kernel protocols to the same routing table and changing route destination
(gateway) in an export filter of a kernel protocol does not work. Both
limitations can be overcome using another routing table and the pipe protocol.

<p>The Kernel protocol supports both IPv4 and IPv6 channels; only one channel
can be configured in each protocol instance. On Linux, it also supports <ref
id="ip-sadr-routes" name="IPv6 SADR"> and <ref id="mpls-routes" name="MPLS">
channels.

<sect1>Configuration
<label id="krt-config">

<p><descrip>
	<tag><label id="krt-persist">persist <m/switch/</tag>
	Tell BIRD to leave all its routes in the routing tables when it exits
	(instead of cleaning them up).

	<tag><label id="krt-scan-time">scan time <m/number/</tag>
	Time in seconds between two consecutive scans of the kernel routing
	table.

	<tag><label id="krt-learn">learn <m/switch/</tag>
	Enable learning of routes added to the kernel routing tables by other
	routing daemons or by the system administrator. This is possible only on
	systems which support identification of route authorship.

	<tag><label id="krt-kernel-table">kernel table <m/number/</tag>
	Select which kernel table should this particular instance of the Kernel
	protocol work with. Available only on systems supporting multiple
	routing tables.

	<tag><label id="krt-metric">metric <m/number/</tag> (Linux)
	Use specified value as a kernel metric (priority) for all routes sent to
	the kernel. When multiple routes for the same network are in the kernel
	routing table, the Linux kernel chooses one with lower metric. Also,
	routes with different metrics do not clash with each other, therefore
	using dedicated metric value is a reliable way to avoid overwriting
	routes from other sources (e.g. kernel device routes). Metric 0 has a
	special meaning of undefined metric, in which either OS default is used,
	or per-route metric can be set using <cf/krt_metric/ attribute. Default:
	32.

	<tag><label id="krt-graceful-restart">graceful restart <m/switch/</tag>
	Participate in graceful restart recovery. If this option is enabled and
	a graceful restart recovery is active, the Kernel protocol will defer
	synchronization of routing tables until the end of the recovery. Note
	that import of kernel routes to BIRD is not affected.

	<tag><label id="krt-merge-paths">merge paths <M>switch</M> [limit <M>number</M>]</tag>
	Usually, only best routes are exported to the kernel protocol. With path
	merging enabled, both best routes and equivalent non-best routes are
	merged during export to generate one ECMP (equal-cost multipath) route
	for each network. This is useful e.g. for BGP multipath. Note that best
	routes are still pivotal for route export (responsible for most
	properties of resulting ECMP routes), while exported non-best routes are
	responsible just for additional multipath next hops. This option also
	allows to specify a limit on maximal number of nexthops in one route. By
	default, multipath merging is disabled. If enabled, default value of the
	limit is 16.

	<tag><label id="krt-netlink-rx-buffer">netlink rx buffer <m/number/</tag> (Linux)
	Set kernel receive buffer size (in bytes) for the netlink socket. The default
	value is OS-dependent (from the <file>/proc/sys/net/core/rmem_default</file>
	file), If you get some "Kernel dropped some netlink message ..." warnings,
	you may increase this value.
</descrip>

<sect1>Attributes
<label id="krt-attr">

<p>The Kernel protocol defines several attributes. These attributes are
translated to appropriate system (and OS-specific) route attributes. We support
these attributes:

<descrip>
	<tag><label id="rta-krt-source">int krt_source</tag>
	The original source of the imported kernel route. The value is
	system-dependent. On Linux, it is a value of the protocol field of the
	route. See /etc/iproute2/rt_protos for common values. On BSD, it is
	based on STATIC and PROTOx flags. The attribute is read-only.

	<tag><label id="rta-krt-metric">int krt_metric</tag> (Linux)
	The kernel metric of the route. When multiple same routes are in a
	kernel routing table, the Linux kernel chooses one with lower metric.
	Note that preferred way to set kernel metric is to use protocol option
	<cf/metric/, unless per-route metric values are needed.

	<tag><label id="rta-krt-prefsrc">ip krt_prefsrc</tag> (Linux)
	The preferred source address. Used in source address selection for
	outgoing packets. Has to be one of the IP addresses of the router.

	<tag><label id="rta-krt-realm">int krt_realm</tag> (Linux)
	The realm of the route. Can be used for traffic classification.

	<tag><label id="rta-krt-scope">int krt_scope</tag> (Linux IPv4)
	The scope of the route. Valid values are 0-254, although Linux kernel
	may reject some values depending on route type and nexthop. It is
	supposed to represent `indirectness' of the route, where nexthops of
	routes are resolved through routes with a higher scope, but in current
	kernels anything below <it/link/ (253) is treated as <it/global/ (0).
	When not present, global scope is implied for all routes except device
	routes, where link scope is used by default.
</descrip>

<p>In Linux, there is also a plenty of obscure route attributes mostly focused
on tuning TCP performance of local connections. BIRD supports most of these
attributes, see Linux or iproute2 documentation for their meaning. Attributes
<cf/krt_lock_*/ and <cf/krt_feature_*/ have type bool, others have type int.
Supported attributes are:

<cf/krt_mtu/, <cf/krt_lock_mtu/, <cf/krt_window/, <cf/krt_lock_window/,
<cf/krt_rtt/, <cf/krt_lock_rtt/, <cf/krt_rttvar/, <cf/krt_lock_rttvar/,
<cf/krt_sstresh/, <cf/krt_lock_sstresh/, <cf/krt_cwnd/, <cf/krt_lock_cwnd/,
<cf/krt_advmss/, <cf/krt_lock_advmss/, <cf/krt_reordering/, <cf/krt_lock_reordering/,
<cf/krt_hoplimit/, <cf/krt_lock_hoplimit/, <cf/krt_rto_min/, <cf/krt_lock_rto_min/,
<cf/krt_initcwnd/, <cf/krt_initrwnd/, <cf/krt_quickack/,
<cf/krt_feature_ecn/, <cf/krt_feature_allfrag/

<sect1>Example
<label id="krt-exam">

<p>A simple configuration can look this way:

<p><code>
protocol kernel {
	export all;
}
</code>

<p>Or for a system with two routing tables:

<p><code>
protocol kernel {		# Primary routing table
	learn;			# Learn alien routes from the kernel
	persist;		# Do not remove routes on bird shutdown
	scan time 10;		# Scan kernel routing table every 10 seconds
	ipv4 {
		import all;
		export all;
	};
}

protocol kernel {		# Secondary routing table
	kernel table 100;
	ipv4 {
		table auxtable;
		export all;
	};
}
</code>


<sect>L3VPN
<label id="l3vpn">

<sect1>Introduction
<label id="l3vpn-intro">

<p>The L3VPN protocol serves as a translator between IP routes and VPN
routes. It is a component for BGP/MPLS IP VPNs (<rfc id="4364">) and implements
policies defined there. In import direction (VPN -&gt; IP), VPN routes matching
import target specification are stripped of route distinguisher and MPLS labels
and announced as IP routes, In export direction (IP -&gt; VPN), IP routes are
expanded with specific route distinguisher, export target communities and MPLS
label and announced as labeled VPN routes. Unlike the Pipe protocol, the L3VPN
protocol propagates just the best route for each network.

<p>In BGP/MPLS IP VPNs, route distribution is controlled by Route Targets (RT).
VRFs are associated with one or more RTs. Routes are also associated with one or
more RTs, which are encoded as route target extended communities
in <ref id="rta-bgp-ext-community" name="bgp_ext_community">. A route is then
imported into each VRF that shares an associated Route Target. The L3VPN
protocol implements this mechanism through mandatory <cf/import target/ and
<cf/export target/ protocol options.

<sect1>Configuration
<label id="l3vpn-config">

<p>L3VPN configuration consists of a few mandatory options and multiple channel
definitions. For convenience, the default export filter in L3VPN channels is
<cf/all/, as the primary way to control import and export of routes is through
protocol options <cf/import target/ and <cf/export target/. If custom filters
are used, note that the export filter of the input channel is applied before
the route translation, while the import filter of the output channel is applied
after that.

<p>In contrast to the Pipe protocol, the L3VPN protocol can handle both IPv4 and
IPv6 routes in one instance, also both IP side and VPN side are represented as
separate channels, although that may change in the future. The L3VPN is always
MPLS-aware protocol, therefore a MPLS channel is mandatory. Altogether, L3VPN
could have up to 5 channels: <cf/ipv4/, <cf/ipv6/, <cf/vpn4/, <cf/vpn6/, and
<cf/mpls/.

<p><descrip>
	<tag><label id="l3vpn-route-distinguisher">route distinguisher <m/vpn-rd/</tag>
	<tag><label id="l3vpn-rd">rd <m/vpn-rd/</tag>
	The route distinguisher that is attached to routes in the export
	direction. Mandatory.

	<tag><label id="l3vpn-import-target">import target <m/ec/|<m/ec-set/</tag>
	Route target extended communities specifying which routes should be
	imported. Either one community or a set. A route is imported if there is
	non-empty intersection between extended communities of the route and the
	import target of the L3VPN protocol. Mandatory.

	<tag><label id="l3vpn-export-target">export target <m/ec/|<m/ec-set/</tag>
	Route target extended communities that are attached to the route in the
	export direction. Either one community or a set. Other route target
	extended communities are removed. Mandatory.

	<tag><label id="l3vpn-route-target">route target <m/ec/|<m/ec-set/</tag>
	A shorthand for both <cf/import target/ and <cf/export target/.
</descrip>

<sect1>Attributes
<label id="l3vpn-attr">

<p>The L3VPN protocol does not define any route attributes.

<sect1>Example
<label id="l3vpn-exam">

<p>Here is an example of L3VPN setup with one VPN and BGP uplink. IP routes
learned from a customer in the VPN are stored in <cf/vrf0vX/ tables, which are
mapped to kernel VRF vrf0. Routes can also be exchanged through BGP with
different sites hosting that VPN. Forwarding of VPN traffic through the network
is handled by MPLS.

<p>Omitted from the example are some routing protocol to exchange routes with
the customer and some sort of MPLS-aware IGP to resolve next hops for BGP VPN
routes.

<code>
# MPLS basics
mpls domain mdom;
mpls table  mtab;

protocol kernel krt_mpls {
	mpls { table mtab; export all; };
}

vpn4 table vpntab4;
vpn6 table vpntab6;

# Exchange VPN routes through BGP
protocol bgp {
	vpn4 { table vpntab4; import all; export all; };
	vpn6 { table vpntab6; import all; export all; };
	mpls { label policy aggregate; };
	local 10.0.0.1 as 10;
	neighbor 10.0.0.2 as 10;
}

# VRF 0
ipv4 table vrf0v4;
ipv6 table vrf0v6;

protocol kernel kernel0v4 {
        vrf "vrf0";
        ipv4 { table vrf0v4; export all; };
        kernel table 100;
}

protocol kernel kernel0v6 {
        vrf "vrf0";
        ipv6 { table vrf0v6; export all; };
        kernel table 100;
}

protocol l3vpn l3vpn0 {
        vrf "vrf0";
        ipv4 { table vrf0v4; };
        ipv6 { table vrf0v6; };
        vpn4 { table vpntab4; };
        vpn6 { table vpntab6; };
        mpls { label policy vrf; };

        rd 10:12;
        import target [(rt, 10, 32..40)];
        export target [(rt, 10, 30), (rt, 10, 31)];
}
</code>


<sect>MRT
<label id="mrt">

<sect1>Introduction
<label id="mrt-intro">

<p>The MRT protocol is a component responsible for handling the Multi-Threaded
Routing Toolkit (MRT) routing information export format, which is mainly used
for collecting and analyzing of routing information from BGP routers. The MRT
protocol can be configured to do periodic dumps of routing tables, created MRT
files can be analyzed later by other tools. Independent MRT table dumps can also
be requested from BIRD client. There is also a feature to save incoming BGP
messages in MRT files, but it is controlled by <ref id="proto-mrtdump"
name="mrtdump"> options independently of MRT protocol, although that might
change in the future.

BIRD implements the main MRT format specification as defined in <rfc id="6396">
and the ADD_PATH extension (<rfc id="8050">).

<sect1>Configuration
<label id="mrt-config">

<p>MRT configuration consists of several statements describing routing table
dumps. Multiple independent periodic dumps can be done as multiple MRT protocol
instances. The MRT protocol does not use channels. There are two mandatory
statements: <cf/filename/ and <cf/period/.

The behavior can be modified by following configuration parameters:

<descrip>
	<tag><label id="mrt-table">table <m/name/ | "<m/pattern/"</tag>
	Specify a routing table (or a set of routing tables described by a
	wildcard pattern) that are to be dumped by the MRT protocol instance.
	Default: the master table.

	<tag><label id="mrt-filter">filter { <m/filter commands/ }</tag>
	The MRT protocol allows to specify a filter that is applied to routes as
	they are dumped. Rejected routes are ignored and not saved to the MRT
	dump file. Default: no filter.

	<tag><label id="mrt-where">where <m/filter expression/</tag>
	An alternative way to specify a filter for the MRT protocol.

	<tag><label id="mrt-filename">filename "<m/filename/"</tag>
	Specify a filename for MRT dump files. The filename may contain time
	format sequences with <it/strftime(3)/ notation (see <it/man strftime/
	for details), there is also a sequence "%N" that is expanded to the name
	of dumped table. Therefore, each periodic dump of each table can be
	saved to a different file. Mandatory, see example below.

	<tag><label id="mrt-period">period <m/number/</tag>
	Specify the time interval (in seconds) between periodic dumps.
	Mandatory.

	<tag><label id="mrt-always-add-path">always add path <m/switch/</tag>
	The MRT format uses special records (specified in <rfc id="8050">) for
	routes received using BGP ADD_PATH extension to keep Path ID, while
	other routes use regular records. This has advantage of better
	compatibility with tools that do not know special records, but it loses
	information about which route is the best route. When this option is
	enabled, both ADD_PATH and non-ADD_PATH routes are stored in ADD_PATH
	records and order of routes for network is preserved. Default: disabled.
</descrip>

<sect1>Example
<label id="mrt-exam">

<p><code>
protocol mrt {
	table "tab*";
	where source = RTS_BGP;
	filename "/var/log/bird/%N_%F_%T.mrt";
	period 300;
}
</code>


<sect>OSPF
<label id="ospf">

<sect1>Introduction
<label id="ospf-intro">

<p>Open Shortest Path First (OSPF) is a quite complex interior gateway
protocol. The current IPv4 version (OSPFv2) is defined in <rfc id="2328"> and
the current IPv6 version (OSPFv3) is defined in <rfc id="5340"> It's a link
state (a.k.a. shortest path first) protocol -- each router maintains a database
describing the autonomous system's topology. Each participating router has an
identical copy of the database and all routers run the same algorithm
calculating a shortest path tree with themselves as a root. OSPF chooses the
least cost path as the best path.

<p>In OSPF, the autonomous system can be split to several areas in order to
reduce the amount of resources consumed for exchanging the routing information
and to protect the other areas from incorrect routing data. Topology of the area
is hidden to the rest of the autonomous system.

<p>Another very important feature of OSPF is that it can keep routing information
from other protocols (like Static or BGP) in its link state database as external
routes. Each external route can be tagged by the advertising router, making it
possible to pass additional information between routers on the boundary of the
autonomous system.

<p>OSPF quickly detects topological changes in the autonomous system (such as
router interface failures) and calculates new loop-free routes after a short
period of convergence. Only a minimal amount of routing traffic is involved.

<p>Each router participating in OSPF routing periodically sends Hello messages
to all its interfaces. This allows neighbors to be discovered dynamically. Then
the neighbors exchange theirs parts of the link state database and keep it
identical by flooding updates. The flooding process is reliable and ensures that
each router detects all changes.

<sect1>Configuration
<label id="ospf-config">

<p>First, the desired OSPF version can be specified by using <cf/ospf v2/ or
<cf/ospf v3/ as a protocol type. By default, OSPFv2 is used. In the main part of
configuration, there can be multiple definitions of OSPF areas, each with a
different id. These definitions includes many other switches and multiple
definitions of interfaces. Definition of interface may contain many switches and
constant definitions and list of neighbors on nonbroadcast networks.

<p>OSPFv2 needs one IPv4 channel. OSPFv3 needs either one IPv6 channel, or one
IPv4 channel (<rfc id="5838">). Therefore, it is possible to use OSPFv3 for both
IPv4 and Pv6 routing, but it is necessary to have two protocol instances anyway.
If no channel is configured, appropriate channel is defined with default
parameters.

<code>
protocol ospf [v2|v3] &lt;name&gt; {
	rfc1583compat &lt;switch&gt;;
	rfc5838 &lt;switch&gt;;
	instance id &lt;num&gt;;
	stub router &lt;switch&gt;;
	tick &lt;num&gt;;
	ecmp &lt;switch&gt; [limit &lt;num&gt;];
	merge external &lt;switch&gt;;
	graceful restart &lt;switch&gt;|aware;
	graceful restart time &lt;num&gt;;
	area &lt;id&gt; {
		stub;
		nssa;
		summary &lt;switch&gt;;
		default nssa &lt;switch&gt;;
		default cost &lt;num&gt;;
		default cost2 &lt;num&gt;;
		translator &lt;switch&gt;;
		translator stability &lt;num&gt;;

                networks {
			&lt;prefix&gt;;
			&lt;prefix&gt; hidden;
		};
                external {
			&lt;prefix&gt;;
			&lt;prefix&gt; hidden;
			&lt;prefix&gt; tag &lt;num&gt;;
		};
		stubnet &lt;prefix&gt;;
		stubnet &lt;prefix&gt; {
			hidden &lt;switch&gt;;
			summary &lt;switch&gt;;
			cost &lt;num&gt;;
		};
		interface &lt;interface pattern&gt; [instance &lt;num&gt;] {
			cost &lt;num&gt;;
			stub &lt;switch&gt;;
			hello &lt;num&gt;;
			poll &lt;num&gt;;
			retransmit &lt;num&gt;;
			priority &lt;num&gt;;
			wait &lt;num&gt;;
			dead count &lt;num&gt;;
			dead &lt;num&gt;;
			secondary &lt;switch&gt;;
			rx buffer [normal|large|&lt;num&gt;];
			tx length &lt;num&gt;;
			type [broadcast|bcast|pointopoint|ptp|
				nonbroadcast|nbma|pointomultipoint|ptmp];
			link lsa suppression &lt;switch&gt;;
			strict nonbroadcast &lt;switch&gt;;
			real broadcast &lt;switch&gt;;
			ptp netmask &lt;switch&gt;;
			ptp address &lt;switch&gt;;
			check link &lt;switch&gt;;
			bfd &lt;switch&gt;;
			ecmp weight &lt;num&gt;;
			ttl security [&lt;switch&gt;; | tx only]
			tx class|dscp &lt;num&gt;;
			tx priority &lt;num&gt;;
			authentication none|simple|cryptographic;
			password "&lt;text&gt;";
			password "&lt;text&gt;" {
				id &lt;num&gt;;
				generate from "&lt;date&gt;";
				generate to "&lt;date&gt;";
				accept from "&lt;date&gt;";
				accept to "&lt;date&gt;";
				from "&lt;date&gt;";
				to "&lt;date&gt;";
				algorithm ( keyed md5 | keyed sha1 | hmac sha1 | hmac sha256 | hmac sha384 | hmac sha512 );
			};
			neighbors {
				&lt;ip&gt;;
				&lt;ip&gt; eligible;
			};
		};
		virtual link &lt;id&gt; [instance &lt;num&gt;] {
			hello &lt;num&gt;;
			retransmit &lt;num&gt;;
			wait &lt;num&gt;;
			dead count &lt;num&gt;;
			dead &lt;num&gt;;
			authentication none|simple|cryptographic;
			password "&lt;text&gt;";
			password "&lt;text&gt;" {
				id &lt;num&gt;;
				generate from "&lt;date&gt;";
				generate to "&lt;date&gt;";
				accept from "&lt;date&gt;";
				accept to "&lt;date&gt;";
				from "&lt;date&gt;";
				to "&lt;date&gt;";
				algorithm ( keyed md5 | keyed sha1 | hmac sha1 | hmac sha256 | hmac sha384 | hmac sha512 );
			};
		};
	};
}
</code>

<descrip>
	<tag><label id="ospf-rfc1583compat">rfc1583compat <M>switch</M></tag>
	This option controls compatibility of routing table calculation with
	<rfc id="1583">. Default value is no.

	<tag><label id="ospf-rfc5838">rfc5838 <m/switch/</tag>
	Basic OSPFv3 is limited to IPv6 unicast routing. The <rfc id="5838">
	extension defines support for more address families (IPv4, IPv6, both
	unicast and multicast). The extension is enabled by default, but can be
	disabled if necessary, as it restricts the range of available instance
	IDs. Default value is yes.

	<tag><label id="ospf-instance-id">instance id <m/num/</tag>
	When multiple OSPF protocol instances are active on the same links, they
	should use different instance IDs to distinguish their packets. Although
	it could be done on per-interface basis, it is often preferred to set
	one instance ID to whole OSPF domain/topology (e.g., when multiple
	instances are used to represent separate logical topologies on the same
	physical network). This option specifies the instance ID for all
	interfaces of the OSPF instance, but can be overridden by
	<cf/interface/ option. Default value is 0 unless OSPFv3-AF extended
	address families are used, see <rfc id="5838"> for that case.

	<tag><label id="ospf-stub-router">stub router <M>switch</M></tag>
	This option configures the router to be a stub router, i.e., a router
	that participates in the OSPF topology but does not allow transit
	traffic. In OSPFv2, this is implemented by advertising maximum metric
	for outgoing links. In OSPFv3, the stub router behavior is announced by
	clearing the R-bit in the router LSA. See <rfc id="6987"> for details.
	Default value is no.

	<tag><label id="ospf-tick">tick <M>num</M></tag>
	The routing table calculation and clean-up of areas' databases is not
	performed when a single link state change arrives. To lower the CPU
	utilization, it's processed later at periodical intervals of <m/num/
	seconds. The default value is 1.

	<tag><label id="ospf-ecmp">ecmp <M>switch</M> [limit <M>number</M>]</tag>
	This option specifies whether OSPF is allowed to generate ECMP
	(equal-cost multipath) routes. Such routes are used when there are
	several directions to the destination, each with the same (computed)
	cost. This option also allows to specify a limit on maximum number of
	nexthops in one route. By default, ECMP is enabled if supported by
	Kernel. Default value of the limit is 16.

	<tag><label id="ospf-merge-external">merge external <M>switch</M></tag>
	This option specifies whether OSPF should merge external routes from
	different routers/LSAs for the same destination. When enabled together
	with <cf/ecmp/, equal-cost external routes will be combined to multipath
	routes in the same way as regular routes. When disabled, external routes
	from different LSAs are treated as separate even if they represents the
	same destination. Default value is no.

	<tag><label id="ospf-graceful-restart">graceful restart <m/switch/|aware</tag>
	When an OSPF instance is restarted, neighbors break adjacencies and
	recalculate their routing tables, which disrupts packet forwarding even
	when the forwarding plane of the restarting router remains intact.
	<rfc id="3623"> specifies a graceful restart mechanism to alleviate this
	issue. For OSPF graceful restart, restarting router originates
	Grace-LSAs, announcing intent to do graceful restart. Neighbors
	receiving these LSAs enter helper mode, in which they ignore breakdown
	of adjacencies, behave as if nothing is happening and keep old routes.
	When adjacencies are reestablished, the restarting router flushes
	Grace-LSAs and graceful restart is ended.

	This option controls the graceful restart mechanism. It has three
	states: Disabled, when no support is provided. Aware, when graceful
	restart helper mode is supported, but no local graceful restart is
	allowed (i.e. helper-only role). Enabled, when the full graceful restart
	support is provided (i.e. both restarting and helper role). Note that
	proper support for local graceful restart requires also configuration of
	other protocols. Default: aware.

	<tag><label id="ospf-graceful-restart-time">graceful restart time <m/num/</tag>
	The restart time is announced in the Grace-LSA and specifies how long
	neighbors should wait for proper end of the graceful restart before
	exiting helper mode prematurely. Default: 120 seconds.

	<tag><label id="ospf-area">area <M>id</M></tag>
	This defines an OSPF area with given area ID (an integer or an IPv4
	address, similarly to a router ID). The most important area is the
	backbone (ID 0) to which every other area must be connected.

	<tag><label id="ospf-stub">stub</tag>
	This option configures the area to be a stub area. External routes are
	not flooded into stub areas. Also summary LSAs can be limited in stub
	areas (see option <cf/summary/). By default, the area is not a stub
	area.

	<tag><label id="ospf-nssa">nssa</tag>
	This option configures the area to be a NSSA (Not-So-Stubby Area). NSSA
	is a variant of a stub area which allows a limited way of external route
	propagation. Global external routes are not propagated into a NSSA, but
	an external route can be imported into NSSA as a (area-wide) NSSA-LSA
	(and possibly translated and/or aggregated on area boundary). By
	default, the area is not NSSA.

	<tag><label id="ospf-summary">summary <M>switch</M></tag>
	This option controls propagation of summary LSAs into stub or NSSA
	areas. If enabled, summary LSAs are propagated as usual, otherwise just
	the default summary route (0.0.0.0/0) is propagated (this is sometimes
	called totally stubby area). If a stub area has more area boundary
	routers, propagating summary LSAs could lead to more efficient routing
	at the cost of larger link state database. Default value is no.

	<tag><label id="ospf-default-nssa">default nssa <M>switch</M></tag>
	When <cf/summary/ option is enabled, default summary route is no longer
	propagated to the NSSA. In that case, this option allows to originate
	default route as NSSA-LSA to the NSSA. Default value is no.

	<tag><label id="ospf-default-cost">default cost <M>num</M></tag>
	This option controls the cost of a default route propagated to stub and
	NSSA areas. Default value is 1000.

	<tag><label id="ospf-default-cost2">default cost2 <M>num</M></tag>
	When a default route is originated as NSSA-LSA, its cost can use either
	type 1 or type 2 metric. This option allows to specify the cost of a
	default route in type 2 metric. By default, type 1 metric (option
	<cf/default cost/) is used.

	<tag><label id="ospf-translator">translator <M>switch</M></tag>
	This option controls translation of NSSA-LSAs into external LSAs. By
	default, one translator per NSSA is automatically elected from area
	boundary routers. If enabled, this area boundary router would
	unconditionally translate all NSSA-LSAs regardless of translator
	election. Default value is no.

	<tag><label id="ospf-translator-stability">translator stability <M>num</M></tag>
	This option controls the translator stability interval (in seconds).
	When the new translator is elected, the old one keeps translating until
	the interval is over. Default value is 40.

	<tag><label id="ospf-networks">networks { <m/set/ }</tag>
	Definition of area IP ranges. This is used in summary LSA origination.
	Hidden networks are not propagated into other areas.

	<tag><label id="ospf-external">external { <m/set/ }</tag>
	Definition of external area IP ranges for NSSAs. This is used for
	NSSA-LSA translation. Hidden networks are not translated into external
	LSAs. Networks can have configured route tag.

	<tag><label id="ospf-stubnet">stubnet <m/prefix/ { <m/options/ }</tag>
	Stub networks are networks that are not transit networks between OSPF
	routers. They are also propagated through an OSPF area as a part of a
	link state database. By default, BIRD generates a stub network record
	for each primary network address on each OSPF interface that does not
	have any OSPF neighbors, and also for each non-primary network address
	on each OSPF interface. This option allows to alter a set of stub
	networks propagated by this router.

	Each instance of this option adds a stub network with given network
	prefix to the set of propagated stub network, unless option <cf/hidden/
	is used. It also suppresses default stub networks for given network
	prefix. When option <cf/summary/ is used, also default stub networks
	that are subnetworks of given stub network are suppressed. This might be
	used, for example, to aggregate generated stub networks.

	<tag><label id="ospf-iface">interface <M>pattern</M> [instance <m/num/]</tag>
	Defines that the specified interfaces belong to the area being defined.
	See <ref id="proto-iface" name="interface"> common option for detailed
	description. In OSPFv2, extended interface clauses are used, because
	each network prefix is handled as a separate virtual interface.

	You can specify alternative instance ID for the interface definition,
	therefore it is possible to have several instances of that interface
	with different options or even in different areas. For OSPFv2, instance
	ID support is an extension (<rfc id="6549">) and is supposed to be set
	per-protocol. For OSPFv3, it is an integral feature.

	<tag><label id="ospf-virtual-link">virtual link <M>id</M> [instance <m/num/]</tag>
	Virtual link to router with the router id. Virtual link acts as a
	point-to-point interface belonging to backbone. The actual area is used
	as a transport area. This item cannot be in the backbone. Like with
	<cf/interface/ option, you could also use several virtual links to one
	destination with different instance IDs.

	<tag><label id="ospf-cost">cost <M>num</M></tag>
	Specifies output cost (metric) of an interface. Default value is 10.

	<tag><label id="ospf-stub-iface">stub <M>switch</M></tag>
	If set to interface it does not listen to any packet and does not send
	any hello. Default value is no.

	<tag><label id="ospf-hello">hello <M>num</M></tag>
	Specifies interval in seconds between sending of Hello messages. Beware,
	all routers on the same network need to have the same hello interval.
	Default value is 10.

	<tag><label id="ospf-poll">poll <M>num</M></tag>
	Specifies interval in seconds between sending of Hello messages for some
	neighbors on NBMA network. Default value is 20.

	<tag><label id="ospf-retransmit">retransmit <M>num</M></tag>
	Specifies interval in seconds between retransmissions of unacknowledged
	updates. Default value is 5.

	<tag><label id="ospf-transmit-delay">transmit delay <M>num</M></tag>
	Specifies estimated transmission delay of link state updates send over
	the interface. The value is added to LSA age of LSAs propagated through
	it. Default value is 1.

	<tag><label id="ospf-priority">priority <M>num</M></tag>
	On every multiple access network (e.g., the Ethernet) Designated Router
	and Backup Designated router are elected. These routers have some special
	functions in the flooding process. Higher priority increases preferences
	in this election. Routers with priority 0 are not eligible. Default
	value is 1.

	<tag><label id="ospf-wait">wait <M>num</M></tag>
	After start, router waits for the specified number of seconds between
	starting election and building adjacency. Default value is 4*<m/hello/.

	<tag><label id="ospf-dead-count">dead count <M>num</M></tag>
	When the router does not receive any messages from a neighbor in
	<m/dead count/*<m/hello/ seconds, it will consider the neighbor down.

	<tag><label id="ospf-dead">dead <M>num</M></tag>
	When the router does not receive any messages from a neighbor in
	<m/dead/ seconds, it will consider the neighbor down. If both directives
	<cf/dead count/ and <cf/dead/ are used, <cf/dead/ has precedence.

	<tag><label id="ospf-rx-buffer">rx buffer <M>num</M></tag>
	This option allows to specify the size of buffers used for packet
	processing. The buffer size should be bigger than maximal size of any
	packets. By default, buffers are dynamically resized as needed, but a
	fixed value could be specified. Value <cf/large/ means maximal allowed
	packet size - 65535.

	<tag><label id="ospf-tx-length">tx length <M>num</M></tag>
	Transmitted OSPF messages that contain large amount of information are
	segmented to separate OSPF packets to avoid IP fragmentation. This
	option specifies the soft ceiling for the length of generated OSPF
	packets. Default value is the MTU of the network interface. Note that
	larger OSPF packets may still be generated if underlying OSPF messages
	cannot be splitted (e.g. when one large LSA is propagated).

	<tag><label id="ospf-type-bcast">type broadcast|bcast</tag>
	BIRD detects a type of a connected network automatically, but sometimes
	it's convenient to force use of a different type manually. On broadcast
	networks (like ethernet), flooding and Hello messages are sent using
	multicasts (a single packet for all the neighbors). A designated router
	is elected and it is responsible for synchronizing the link-state
	databases and originating network LSAs. This network type cannot be used
	on physically NBMA networks and on unnumbered networks (networks without
	proper IP prefix).

	<tag><label id="ospf-type-ptp">type pointopoint|ptp</tag>
	Point-to-point networks connect just 2 routers together. No election is
	performed and no network LSA is originated, which makes it simpler and
	faster to establish. This network type is useful not only for physically
	PtP ifaces (like PPP or tunnels), but also for broadcast networks used
	as PtP links. This network type cannot be used on physically NBMA
	networks.

	<tag><label id="ospf-type-nbma">type nonbroadcast|nbma</tag>
	On NBMA networks, the packets are sent to each neighbor separately
	because of lack of multicast capabilities. Like on broadcast networks,
	a designated router is elected, which plays a central role in propagation
	of LSAs. This network type cannot be used on unnumbered networks.

	<tag><label id="ospf-type-ptmp">type pointomultipoint|ptmp</tag>
	This is another network type designed to handle NBMA networks. In this
	case the NBMA network is treated as a collection of PtP links. This is
	useful if not every pair of routers on the NBMA network has direct
	communication, or if the NBMA network is used as an (possibly
	unnumbered) PtP link.

	<tag><label id="ospf-link-lsa-suppression">link lsa suppression <m/switch/</tag>
	In OSPFv3, link LSAs are generated for each link, announcing link-local
	IPv6 address of the router to its local neighbors. These are useless on
	PtP or PtMP networks and this option allows to suppress the link LSA
	origination for such interfaces. The option is ignored on other than PtP
	or PtMP interfaces. Default value is no.

	<tag><label id="ospf-strict-nonbroadcast">strict nonbroadcast <m/switch/</tag>
	If set, don't send hello to any undefined neighbor. This switch is
	ignored on other than NBMA or PtMP interfaces. Default value is no.

	<tag><label id="ospf-real-broadcast">real broadcast <m/switch/</tag>
	In <cf/type broadcast/ or <cf/type ptp/ network configuration, OSPF
	packets are sent as IP multicast packets. This option changes the
	behavior to using old-fashioned IP broadcast packets. This may be useful
	as a workaround if IP multicast for some reason does not work or does
	not work reliably. This is a non-standard option and probably is not
	interoperable with other OSPF implementations. Default value is no.

	<tag><label id="ospf-ptp-netmask">ptp netmask <m/switch/</tag>
	In <cf/type ptp/ network configurations, OSPFv2 implementations should
	ignore received netmask field in hello packets and should send hello
	packets with zero netmask field on unnumbered PtP links. But some OSPFv2
	implementations perform netmask checking even for PtP links.

	This option specifies whether real netmask will be used in hello packets
	on <cf/type ptp/ interfaces. You should ignore this option unless you
	meet some compatibility problems related to this issue. Default value is
	no for unnumbered PtP links, yes otherwise.

	<tag><label id="ospf-ptp-address">ptp address <m/switch/</tag>
	In <cf/type ptp/ network configurations, OSPFv2 implementations should
	use IP address for regular PtP links and interface id for unnumbered PtP
	links in data field of link description records in router LSA. This data
	field has only local meaning for PtP links, but some broken OSPFv2
	implementations assume there is an IP address and use it as a next hop
	in SPF calculations. Note that interface id for unnumbered PtP links is
	necessary when graceful restart is enabled to distinguish PtP links with
	the same local IP address.

	This option specifies whether an IP address will be used in data field
	for <cf/type ptp/ interfaces, it is ignored for other interfaces. You
	should ignore this option unless you meet some compatibility problems
	related to this issue. Default value is no for unnumbered PtP links when
	graceful restart is enabled, yes otherwise.

	<tag><label id="ospf-check-link">check link <M>switch</M></tag>
	If set, a hardware link state (reported by OS) is taken into consideration.
	When a link disappears (e.g. an ethernet cable is unplugged), neighbors
	are immediately considered unreachable and only the address of the iface
	(instead of whole network prefix) is propagated. It is possible that
	some hardware drivers or platforms do not implement this feature.
	Default value is yes.

	<tag><label id="ospf-bfd">bfd <M>switch</M></tag>
	OSPF could use BFD protocol as an advisory mechanism for neighbor
	liveness and failure detection. If enabled, BIRD setups a BFD session
	for each OSPF neighbor and tracks its liveness by it. This has an
	advantage of an order of magnitude lower detection times in case of
	failure. Note that BFD protocol also has to be configured, see
	<ref id="bfd" name="BFD"> section for details. Default value is no.

	<tag><label id="ospf-ttl-security">ttl security [<m/switch/ | tx only]</tag>
	TTL security is a feature that protects routing protocols from remote
	spoofed packets by using TTL 255 instead of TTL 1 for protocol packets
	destined to neighbors. Because TTL is decremented when packets are
	forwarded, it is non-trivial to spoof packets with TTL 255 from remote
	locations. Note that this option would interfere with OSPF virtual
	links.

	If this option is enabled, the router will send OSPF packets with TTL
	255 and drop received packets with TTL less than 255. If this option si
	set to <cf/tx only/, TTL 255 is used for sent packets, but is not
	checked for received packets. Default value is no.

	<tag><label id="ospf-tx-class">tx class|dscp|priority <m/num/</tag>
	These options specify the ToS/DiffServ/Traffic class/Priority of the
	outgoing OSPF packets. See <ref id="proto-tx-class" name="tx class"> common
	option for detailed description.

	<tag><label id="ospf-ecmp-weight">ecmp weight <M>num</M></tag>
	When ECMP (multipath) routes are allowed, this value specifies a
	relative weight used for nexthops going through the iface. Allowed
	values are 1-256. Default value is 1.

	<tag><label id="ospf-auth-none">authentication none</tag>
	No passwords are sent in OSPF packets. This is the default value.

	<tag><label id="ospf-auth-simple">authentication simple</tag>
	Every packet carries 8 bytes of password. Received packets lacking this
	password are ignored. This authentication mechanism is very weak.
	This option is not available in OSPFv3.

	<tag><label id="ospf-auth-cryptographic">authentication cryptographic</tag>
	An authentication code is appended to every packet. The specific
	cryptographic algorithm is selected by option <cf/algorithm/ for each
	key. The default cryptographic algorithm for OSPFv2 keys is Keyed-MD5
	and for OSPFv3 keys is HMAC-SHA-256. Passwords are not sent open via
	network, so this mechanism is quite secure. Packets can still be read by
	an attacker.

	<tag><label id="ospf-pass">password "<M>text</M>"</tag>
	Specifies a password used for authentication. See
	<ref id="proto-pass" name="password"> common option for detailed
	description.

	<tag><label id="ospf-neighbors">neighbors { <m/set/ } </tag>
	A set of neighbors to which Hello messages on NBMA or PtMP networks are
	to be sent. For NBMA networks, some of them could be marked as eligible.
	In OSPFv3, link-local addresses should be used, using global ones is
	possible, but it is nonstandard and might be problematic. And definitely,
	link-local and global addresses should not be mixed.
</descrip>

<sect1>Attributes
<label id="ospf-attr">

<p>OSPF defines four route attributes. Each internal route has a <cf/metric/.

<p>Metric is ranging from 1 to infinity (65535). External routes use
<cf/metric type 1/ or <cf/metric type 2/. A <cf/metric of type 1/ is comparable
with internal <cf/metric/, a <cf/metric of type 2/ is always longer than any
<cf/metric of type 1/ or any <cf/internal metric/. <cf/Internal metric/ or
<cf/metric of type 1/ is stored in attribute <cf/ospf_metric1/, <cf/metric type
2/ is stored in attribute <cf/ospf_metric2/.

When both metrics are specified then <cf/metric of type 2/ is used. This is
relevant e.g. when a type 2 external route is propagated from one OSPF domain to
another and <cf/ospf_metric1/ is an internal distance to the original ASBR,
while <cf/ospf_metric2/ stores the type 2 metric. Note that in such cases if
<cf/ospf_metric1/ is non-zero then <cf/ospf_metric2/ is increased by one to
ensure monotonicity of metric, as internal distance is reset to zero when an
external route is announced.

<p>Each external route can also carry attribute <cf/ospf_tag/ which is a 32-bit
integer which is used when exporting routes to other protocols; otherwise, it
doesn't affect routing inside the OSPF domain at all. The fourth attribute
<cf/ospf_router_id/ is a router ID of the router advertising that route /
network. This attribute is read-only. Default is <cf/ospf_metric2 = 10000/ and
<cf/ospf_tag = 0/.

<sect1>Example
<label id="ospf-exam">

<p><code>
protocol ospf MyOSPF {
	ipv4 {
		export filter {
			if source = RTS_BGP then {
				ospf_metric1 = 100;
				accept;
			}
			reject;
		};
	};
	area 0.0.0.0 {
		interface "eth*" {
			cost 11;
			hello 15;
			priority 100;
			retransmit 7;
			authentication simple;
			password "aaa";
		};
		interface "ppp*" {
			cost 100;
			authentication cryptographic;
			password "abc" {
				id 1;
				generate to "22-04-2003 11:00:06";
				accept from "17-01-2001 12:01:05";
				algorithm hmac sha384;
			};
			password "def" {
				id 2;
				generate to "22-07-2005 17:03:21";
				accept from "22-02-2001 11:34:06";
				algorithm hmac sha512;
			};
		};
		interface "arc0" {
			cost 10;
			stub yes;
		};
		interface "arc1";
	};
	area 120 {
		stub yes;
		networks {
			172.16.1.0/24;
			172.16.2.0/24 hidden;
		};
		interface "-arc0" , "arc*" {
			type nonbroadcast;
			authentication none;
			strict nonbroadcast yes;
			wait 120;
			poll 40;
			dead count 8;
			neighbors {
				192.168.120.1 eligible;
				192.168.120.2;
				192.168.120.10;
			};
		};
	};
}
</code>

<sect>Perf
<label id="perf">

<sect1>Introduction
<label id="perf-intro">

<p>The Perf protocol is a generator of fake routes together with a time measurement
framework. Its purpose is to check BIRD performance and to benchmark filters.

<p>Import mode of this protocol runs in several steps. In each step, it generates 2^x routes,
imports them into the appropriate table and withdraws them. The exponent x is configurable.
It runs the benchmark several times for the same x, then it increases x by one
until it gets too high, then it stops.

<p>Export mode of this protocol repeats route refresh from table and measures how long it takes.

<p>Output data is logged on info level. There is a Perl script <cf>proto/perf/parse.pl</cf>
which may be handy to parse the data and draw some plots.

<p>Implementation of this protocol is experimental. Use with caution and do not keep
any instance of Perf in production configs for long time. The config interface is also unstable
and may change in future versions without warning.

<sect1>Configuration
<label id="perf-config">

<p><descrip>
	<tag><label id="perf-mode">mode import|export</tag>
	Set perf mode. Default: import

	<tag><label id="perf-repeat">repeat <m/number/</tag>
	Run this amount of iterations of the benchmark for every amount step. Default: 4

	<tag><label id="perf-from">exp from <m/number/</tag>
	Begin benchmarking on this exponent for number of generated routes in one step.
	Default: 10

	<tag><label id="perf-to">exp to <m/number/</tag>
	Stop benchmarking on this exponent. Default: 20

	<tag><label id="perf-threshold-min">threshold min <m/time/</tag>
	If a run for the given exponent took less than this time for route import,
	increase the exponent immediately. Default: 1 ms

	<tag><label id="perf-threshold-max">threshold max <m/time/</tag>
	If every run for the given exponent took at least this time for route import,
	stop benchmarking. Default: 500 ms
</descrip>

<sect>Pipe
<label id="pipe">

<sect1>Introduction
<label id="pipe-intro">

<p>The Pipe protocol serves as a link between two routing tables, allowing
routes to be passed from a table declared as primary (i.e., the one the pipe is
connected to using the <cf/table/ configuration keyword) to the secondary one
(declared using <cf/peer table/) and vice versa, depending on what's allowed by
the filters. Export filters control export of routes from the primary table to
the secondary one, import filters control the opposite direction. Both tables
must be of the same nettype.

<p>The Pipe protocol retransmits all routes from one table to the other table,
retaining their original source and attributes. If import and export filters
are set to accept, then both tables would have the same content.

<p>The primary use of multiple routing tables and the Pipe protocol is for
policy routing, where handling of a single packet doesn't depend only on its
destination address, but also on its source address, source interface, protocol
type and other similar parameters. In many systems (Linux being a good example),
the kernel allows to enforce routing policies by defining routing rules which
choose one of several routing tables to be used for a packet according to its
parameters. Setting of these rules is outside the scope of BIRD's work (on
Linux, you can use the <tt/ip/ command), but you can create several routing
tables in BIRD, connect them to the kernel ones, use filters to control which
routes appear in which tables and also you can employ the Pipe protocol for
exporting a selected subset of one table to another one.

<sect1>Configuration
<label id="pipe-config">

<p>Essentially, the Pipe protocol is just a channel connected to a table on both
sides. Therefore, the configuration block for <cf/protocol pipe/ shall directly
include standard channel config options; see the example below.

<p><descrip>
	<tag><label id="pipe-peer-table">peer table <m/table/</tag>
	Defines secondary routing table to connect to. The primary one is
	selected by the <cf/table/ keyword.

	<tag><label id="pipe-max-generation">max generation <m/expr/</tag>
	Sets maximal generation of route that may pass through this pipe.
	The generation value is increased by one by each pipe on its path.
	Not meeting this requirement causes an error message complaining about
	an overpiped route. If you have long chains of pipes, you probably want
	to raise this value; anyway the default of 16 should be enough for even
	most strange uses. Maximum is 254.
</descrip>

<sect1>Attributes
<label id="pipe-attr">

<p>The Pipe protocol doesn't define any route attributes.

<sect1>Example
<label id="pipe-exam">

<p>Let's consider a router which serves as a boundary router of two different
autonomous systems, each of them connected to a subset of interfaces of the
router, having its own exterior connectivity and wishing to use the other AS as
a backup connectivity in case of outage of its own exterior line.

<p>Probably the simplest solution to this situation is to use two routing tables
(we'll call them <cf/as1/ and <cf/as2/) and set up kernel routing rules, so that
packets having arrived from interfaces belonging to the first AS will be routed
according to <cf/as1/ and similarly for the second AS. Thus we have split our
router to two logical routers, each one acting on its own routing table, having
its own routing protocols on its own interfaces. In order to use the other AS's
routes for backup purposes, we can pass the routes between the tables through a
Pipe protocol while decreasing their preferences and correcting their BGP paths
to reflect the AS boundary crossing.

<code>
ipv4 table as1;				# Define the tables
ipv4 table as2;

protocol kernel kern1 {			# Synchronize them with the kernel
	ipv4 { table as1; export all; };
	kernel table 1;
}

protocol kernel kern2 {
	ipv4 { table as2; export all; };
	kernel table 2;
}

protocol bgp bgp1 {			# The outside connections
	ipv4 { table as1; import all; export all; };
	local as 1;
	neighbor 192.168.0.1 as 1001;
}

protocol bgp bgp2 {
	ipv4 { table as2; import all; export all; };
	local as 2;
	neighbor 10.0.0.1 as 1002;
}

protocol pipe {				# The Pipe
	table as1;
	peer table as2;
	export filter {
		if net ~ [ 1.0.0.0/8+] then {	# Only AS1 networks
			if preference>10 then preference = preference-10;
			if source=RTS_BGP then bgp_path.prepend(1);
			accept;
		}
		reject;
	};
	import filter {
		if net ~ [ 2.0.0.0/8+] then {	# Only AS2 networks
			if preference>10 then preference = preference-10;
			if source=RTS_BGP then bgp_path.prepend(2);
			accept;
		}
		reject;
	};
}
</code>


<sect>RAdv
<label id="radv">

<sect1>Introduction
<label id="radv-intro">

<p>The RAdv protocol is an implementation of Router Advertisements, which are
used in the IPv6 stateless autoconfiguration. IPv6 routers send (in irregular
time intervals or as an answer to a request) advertisement packets to connected
networks. These packets contain basic information about a local network (e.g. a
list of network prefixes), which allows network hosts to autoconfigure network
addresses and choose a default route. BIRD implements router behavior as defined
in <rfc id="4861">, router preferences and specific routes (<rfc id="4191">),
and DNS extensions (<rfc id="6106">).

<p>The RAdv protocols supports just IPv6 channel.

<sect1>Configuration
<label id="radv-config">

<p>There are several classes of definitions in RAdv configuration -- interface
definitions, prefix definitions and DNS definitions:

<descrip>
	<tag><label id="radv-iface">interface <m/pattern/ [, <m/.../] { <m/options/ }</tag>
	Interface definitions specify a set of interfaces on which the
	protocol is activated and contain interface specific options.
	See <ref id="proto-iface" name="interface"> common options for
	detailed description.

	<tag><label id="radv-prefix">prefix <m/prefix/ { <m/options/ }</tag>
	Prefix definitions allow to modify a list of advertised prefixes. By
	default, the advertised prefixes are the same as the network prefixes
	assigned to the interface. For each network prefix, the matching prefix
	definition is found and its options are used. If no matching prefix
	definition is found, the prefix is used with default options.

	Prefix definitions can be either global or interface-specific. The
	second ones are part of interface options. The prefix definition
	matching is done in the first-match style, when interface-specific
	definitions are processed before global definitions. As expected, the
	prefix definition is matching if the network prefix is a subnet of the
	prefix in prefix definition.

	<tag><label id="radv-rdnss">rdnss { <m/options/ }</tag>
	RDNSS definitions allow to specify a list of advertised recursive DNS
	servers together with their options. As options are seldom necessary,
	there is also a short variant <cf>rdnss <m/address/</cf> that just
	specifies one DNS server. Multiple definitions are cumulative. RDNSS
	definitions may also be interface-specific when used inside interface
	options. By default, interface uses both global and interface-specific
	options, but that can be changed by <cf/rdnss local/ option.

	<tag><label id="radv-dnssl">dnssl { <m/options/ }</tag>
	DNSSL definitions allow to specify a list of advertised DNS search
	domains together with their options. Like <cf/rdnss/ above, multiple
	definitions are cumulative, they can be used also as interface-specific
	options and there is a short variant <cf>dnssl <m/domain/</cf> that just
	specifies one DNS search domain.

	<tag><label id="radv-custom-option">custom option type <m/number/ value <m/bytestring/</tag>
	Custom option definitions allow to define an arbitrary option to
	advertise. You need to specify the option type number and the binary
	payload of the option. The length field is calculated automatically.
	Like <cf/rdnss/ above, multiple definitions are cumulative, they can
	be used also as interface-specific options.

	The following example advertises PREF64 option (<rfc id="8781">) with
	prefix <cf>2001:db8:a:b::/96</cf> and the lifetime of <cf/1 hour/:

	<label id="radv-custom-option-exam">
<p><code>
custom option type 38 value hex:0e10:2001:0db8:000a:000b:0000:0000;
</code>

	<tag><label id="radv-trigger">trigger <m/prefix/</tag>
	RAdv protocol could be configured to change its behavior based on
	availability of routes. When this option is used, the protocol waits in
	suppressed state until a <it/trigger route/ (for the specified network)
	is exported to the protocol, the protocol also returns to suppressed
	state if the <it/trigger route/ disappears. Note that route export
	depends on specified export filter, as usual. This option could be used,
	e.g., for handling failover in multihoming scenarios.

	During suppressed state, router advertisements are generated, but with
	some fields zeroed. Exact behavior depends on which fields are zeroed,
	this can be configured by <cf/sensitive/ option for appropriate
	fields. By default, just <cf/default lifetime/ (also called <cf/router
	lifetime/) is zeroed, which means hosts cannot use the router as a
	default router. <cf/preferred lifetime/ and <cf/valid lifetime/ could
	also be configured as <cf/sensitive/ for a prefix, which would cause
	autoconfigured IPs to be deprecated or even removed.

	<tag><label id="radv-propagate-routes">propagate routes <m/switch/</tag>
	This option controls propagation of more specific routes, as defined in
	<rfc id="4191">. If enabled, all routes exported to the RAdv protocol,
	with the exception of the trigger prefix, are added to advertisments as
	additional options. The lifetime and preference of advertised routes can
	be set individually by <cf/ra_lifetime/ and <cf/ra_preference/ route
	attributes, or per interface by <cf/route lifetime/ and
	<cf/route preference/ options. Default: disabled.

	Note that the RFC discourages from sending more than 17 routes and
	recommends the routes to be configured manually.
</descrip>

<p>Interface specific options:

<descrip>
	<tag><label id="radv-iface-max-ra-interval">max ra interval <m/expr/</tag>
	Unsolicited router advertisements are sent in irregular time intervals.
	This option specifies the maximum length of these intervals, in seconds.
	Valid values are 4-1800. Default: 600

	<tag><label id="radv-iface-min-ra-interval">min ra interval <m/expr/</tag>
	This option specifies the minimum length of that intervals, in seconds.
	Must be at least 3 and at most 3/4 * <cf/max ra interval/. Default:
	about 1/3 * <cf/max ra interval/.

	<tag><label id="radv-iface-min-delay">min delay <m/expr/</tag>
	The minimum delay between two consecutive router advertisements, in
	seconds. Default: 3

	<tag><label id="radv-solicited-ra-unicast">solicited ra unicast <m/switch/</tag>
	Solicited router advertisements are usually sent to all-nodes multicast
	group like unsolicited ones, but the router can be configured to send
	them as unicast directly to soliciting nodes instead. This is especially
	useful on wireless networks (see <rfc id="7772">). Default: no

	<tag><label id="radv-iface-managed">managed <m/switch/</tag>
	This option specifies whether hosts should use DHCPv6 for IP address
	configuration. Default: no

	<tag><label id="radv-iface-other-config">other config <m/switch/</tag>
	This option specifies whether hosts should use DHCPv6 to receive other
	configuration information. Default: no

	<tag><label id="radv-iface-link-mtu">link mtu <m/expr/</tag>
	This option specifies which value of MTU should be used by hosts. 0
	means unspecified. Default: 0

	<tag><label id="radv-iface-reachable-time">reachable time <m/expr/</tag>
	This option specifies the time (in milliseconds) how long hosts should
	assume a neighbor is reachable (from the last confirmation). Maximum is
	3600000, 0 means unspecified. Default 0.

	<tag><label id="radv-iface-retrans-timer">retrans timer <m/expr/</tag>
	This option specifies the time (in milliseconds) how long hosts should
	wait before retransmitting Neighbor Solicitation messages. 0 means
	unspecified. Default 0.

	<tag><label id="radv-iface-current-hop-limit">current hop limit <m/expr/</tag>
	This option specifies which value of Hop Limit should be used by
	hosts. Valid values are 0-255, 0 means unspecified. Default: 64

	<tag><label id="radv-iface-default-lifetime">default lifetime <m/expr/ [sensitive <m/switch/]</tag>
	This option specifies the time (in seconds) how long (since the receipt
	of RA) hosts may use the router as a default router. 0 means do not use
	as a default router. For <cf/sensitive/ option, see <ref id="radv-trigger" name="trigger">.
	Default: 3 * <cf/max ra	interval/, <cf/sensitive/ yes.

	<tag><label id="radv-iface-default-preference">default preference low|medium|high</tag>
	This option specifies the Default Router Preference value to advertise
	to hosts. Default: medium.

	<tag><label id="radv-iface-route-lifetime">route lifetime <m/expr/ [sensitive <m/switch/]</tag>
	This option specifies the default value of advertised lifetime for
	specific routes; i.e., the time (in seconds) for how long (since the
	receipt of RA) hosts should consider these routes valid. A special value
	0xffffffff represents infinity. The lifetime can be overriden on a per
	route basis by the <ref id="rta-ra-lifetime" name="ra_lifetime"> route
	attribute. Default: 3 * <cf/max ra interval/, <cf/sensitive/ no.

	For the <cf/sensitive/ option, see <ref id="radv-trigger" name="trigger">.
	If <cf/sensitive/ is enabled, even the routes with the <cf/ra_lifetime/
	attribute become sensitive to the trigger.

	<tag><label id="radv-iface-route-preference">route preference low|medium|high</tag>
	This option specifies the default value of advertised route preference
	for specific routes. The value can be overriden on a per route basis by
	the <ref id="rta-ra-preference" name="ra_preference"> route attribute.
	Default: medium.

	<tag><label id="radv-prefix-linger-time">prefix linger time <m/expr/</tag>
	When a prefix or a route disappears, it is advertised for some time with
	zero lifetime, to inform clients it is no longer valid. This option
	specifies the time (in seconds) for how long prefixes are advertised
	that way. Default: 3 * <cf/max ra interval/.

	<tag><label id="radv-route-linger-time">route linger time <m/expr/</tag>
	When a prefix or a route disappears, it is advertised for some time with
	zero lifetime, to inform clients it is no longer valid. This option
	specifies the time (in seconds) for how long routes are advertised
	that way. Default: 3 * <cf/max ra interval/.

	<tag><label id="radv-iface-rdnss-local">rdnss local <m/switch/</tag>
	Use only local (interface-specific) RDNSS definitions for this
	interface. Otherwise, both global and local definitions are used. Could
	also be used to disable RDNSS for given interface if no local definitons
	are specified. Default: no.

	<tag><label id="radv-iface-dnssl-local">dnssl local <m/switch/</tag>
	Use only local DNSSL definitions for this interface. See <cf/rdnss local/
	option above. Default: no.

	<tag><label id="radv-iface-custom-local">custom option local <m/switch/</tag>
	Use only local custom option definitions for this interface. See <cf/rdnss local/
	option above. Default: no.
</descrip>

<p>Prefix specific options

<descrip>
	<tag><label id="radv-prefix-skip">skip <m/switch/</tag>
	This option allows to specify that given prefix should not be
	advertised. This is useful for making exceptions from a default policy
	of advertising all prefixes. Note that for withdrawing an already
	advertised prefix it is more useful to advertise it with zero valid
	lifetime. Default: no

	<tag><label id="radv-prefix-onlink">onlink <m/switch/</tag>
	This option specifies whether hosts may use the advertised prefix for
	onlink determination. Default: yes

	<tag><label id="radv-prefix-autonomous">autonomous <m/switch/</tag>
	This option specifies whether hosts may use the advertised prefix for
	stateless autoconfiguration. Default: yes

	<tag><label id="radv-prefix-valid-lifetime">valid lifetime <m/expr/ [sensitive <m/switch/]</tag>
	This option specifies the time (in seconds) how long (after the
	receipt of RA) the prefix information is valid, i.e., autoconfigured
	IP addresses can be assigned and hosts with that IP addresses are
	considered directly reachable. 0 means the prefix is no longer
	valid. For <cf/sensitive/ option, see <ref id="radv-trigger" name="trigger">.
	Default: 86400 (1 day), <cf/sensitive/ no.

	<tag><label id="radv-prefix-preferred-lifetime">preferred lifetime <m/expr/ [sensitive <m/switch/]</tag>
	This option specifies the time (in seconds) how long (after the
	receipt of RA) IP addresses generated from the prefix using stateless
	autoconfiguration remain preferred. For <cf/sensitive/ option,
	see <ref id="radv-trigger" name="trigger">. Default: 14400 (4 hours),
	<cf/sensitive/ no.
</descrip>

<p>RDNSS specific options:

<descrip>
	<tag><label id="radv-rdnss-ns">ns <m/address/</tag>
	This option specifies one recursive DNS server. Can be used multiple
	times for multiple servers. It is mandatory to have at least one
	<cf/ns/ option in <cf/rdnss/ definition.

	<tag><label id="radv-rdnss-lifetime">lifetime [mult] <m/expr/</tag>
	This option specifies the time how long the RDNSS information may be
	used by clients after the receipt of RA. It is expressed either in
	seconds or (when <cf/mult/ is used) in multiples of <cf/max ra
	interval/. Note that RDNSS information is also invalidated when
	<cf/default lifetime/ expires. 0 means these addresses are no longer
	valid DNS servers. Default: 3 * <cf/max ra interval/.
</descrip>

<p>DNSSL specific options:

<descrip>
	<tag><label id="radv-dnssl-domain">domain <m/address/</tag>
	This option specifies one DNS search domain. Can be used multiple times
	for multiple domains. It is mandatory to have at least one <cf/domain/
	option in <cf/dnssl/ definition.

	<tag><label id="radv-dnssl-lifetime">lifetime [mult] <m/expr/</tag>
	This option specifies the time how long the DNSSL information may be
	used by clients after the receipt of RA. Details are the same as for
	RDNSS <cf/lifetime/ option above. Default: 3 * <cf/max ra interval/.
</descrip>

<sect1>Attributes
<label id="radv-attr">

<p>RAdv defines two route attributes:

<descrip>
	<tag><label id="rta-ra-preference">enum ra_preference</tag>
	The preference of the route. The value can be <it/RA_PREF_LOW/,
	<it/RA_PREF_MEDIUM/ or <it/RA_PREF_HIGH/. If the attribute is not set,
	the <ref id="radv-iface-route-preference" name="route preference">
	option is used.

	<tag><label id="rta-ra-lifetime">int ra_lifetime</tag>
	The advertised lifetime of the route, in seconds. The special value of
	0xffffffff represents infinity. If the attribute is not set, the
	<ref id="radv-iface-route-lifetime" name="route lifetime">
	option is used.
</descrip>

<sect1>Example
<label id="radv-exam">

<p><code>
ipv6 table radv_routes;			# Manually configured routes go here

protocol static {
	ipv6 { table radv_routes; };

	route 2001:0DB8:4000::/48 unreachable;
	route 2001:0DB8:4010::/48 unreachable;

	route 2001:0DB8:4020::/48 unreachable {
		ra_preference = RA_PREF_HIGH;
		ra_lifetime = 3600;
	};
}

protocol radv {
	propagate routes yes;		# Propagate the routes from the radv_routes table
	ipv6 { table radv_routes; export all; };

	interface "eth2" {
		max ra interval 5;	# Fast failover with more routers
		managed yes;		# Using DHCPv6 on eth2
		prefix ::/0 {
			autonomous off;	# So do not autoconfigure any IP
		};
	};

	interface "eth*";		# No need for any other options

	prefix 2001:0DB8:1234::/48 {
		preferred lifetime 0;	# Deprecated address range
	};

	prefix 2001:0DB8:2000::/48 {
		autonomous off;		# Do not autoconfigure
	};

	rdnss 2001:0DB8:1234::10;	# Short form of RDNSS

	rdnss {
		lifetime mult 10;
		ns 2001:0DB8:1234::11;
		ns 2001:0DB8:1234::12;
	};

	dnssl {
		lifetime 3600;
		domain "abc.com";
		domain "xyz.com";
	};
}
</code>


<sect>RIP
<label id="rip">

<sect1>Introduction
<label id="rip-intro">

<p>The RIP protocol (also sometimes called Rest In Pieces) is a simple protocol,
where each router broadcasts (to all its neighbors) distances to all networks it
can reach. When a router hears distance to another network, it increments it and
broadcasts it back. Broadcasts are done in regular intervals. Therefore, if some
network goes unreachable, routers keep telling each other that its distance is
the original distance plus 1 (actually, plus interface metric, which is usually
one). After some time, the distance reaches infinity (that's 15 in RIP) and all
routers know that network is unreachable. RIP tries to minimize situations where
counting to infinity is necessary, because it is slow. Due to infinity being 16,
you can't use RIP on networks where maximal distance is higher than 15
hosts.

<p>BIRD supports RIPv1 (<rfc id="1058">), RIPv2 (<rfc id="2453">), RIPng (<rfc
id="2080">), Triggered RIP for demand circuits (<rfc id="2091">), and RIP
cryptographic authentication (<rfc id="4822">).

<p>RIP is a very simple protocol, and it has a lot of shortcomings. Slow
convergence, big network load and inability to handle larger networks makes it
pretty much obsolete. It is still usable on very small networks.

<sect1>Configuration
<label id="rip-config">

<p>RIP configuration consists mainly of common protocol options and interface
definitions, most RIP options are interface specific. RIPng (RIP for IPv6)
protocol instance can be configured by using <cf/rip ng/ instead of just
<cf/rip/ as a protocol type.

<p>RIP needs one IPv4 channel. RIPng needs one IPv6 channel. If no channel is
configured, appropriate channel is defined with default parameters.

<code>
protocol rip [ng] [&lt;name&gt;] {
	infinity &lt;number&gt;;
	ecmp &lt;switch&gt; [limit &lt;number&gt;];
	interface &lt;interface pattern&gt; {
		metric &lt;number&gt;;
		mode multicast|broadcast;
		passive &lt;switch&gt;;
		address &lt;ip&gt;;
		port &lt;number&gt;;
		version 1|2;
		split horizon &lt;switch&gt;;
		poison reverse &lt;switch&gt;;
		demand circuit &lt;switch&gt;;
		check zero &lt;switch&gt;;
		update time &lt;number&gt;;
		timeout time &lt;number&gt;;
		garbage time &lt;number&gt;;
		ecmp weight &lt;number&gt;;
		ttl security &lt;switch&gt;; | tx only;
		tx class|dscp &lt;number&gt;;
		tx priority &lt;number&gt;;
		rx buffer &lt;number&gt;;
		tx length &lt;number&gt;;
		check link &lt;switch&gt;;
		authentication none|plaintext|cryptographic;
		password "&lt;text&gt;";
		password "&lt;text&gt;" {
			id &lt;num&gt;;
			generate from "&lt;date&gt;";
			generate to "&lt;date&gt;";
			accept from "&lt;date&gt;";
			accept to "&lt;date&gt;";
			from "&lt;date&gt;";
			to "&lt;date&gt;";
			algorithm ( keyed md5 | keyed sha1 | hmac sha1 | hmac sha256 | hmac sha384 | hmac sha512 );
		};
	};
}
</code>

<descrip>
	<tag><label id="rip-infinity">infinity <M>number</M></tag>
	Selects the distance of infinity. Bigger values will make
	protocol convergence even slower. The default value is 16.

	<tag><label id="rip-ecmp">ecmp <M>switch</M> [limit <M>number</M>]</tag>
	This option specifies whether RIP is allowed to generate ECMP
	(equal-cost multipath) routes. Such routes are used when there are
	several directions to the destination, each with the same (computed)
	cost. This option also allows to specify a limit on maximum number of
	nexthops in one route. By default, ECMP is enabled if supported by
	Kernel. Default value of the limit is 16.

	<tag><label id="rip-iface">interface <m/pattern/ [, <m/.../] { <m/options/ }</tag>
	Interface definitions specify a set of interfaces on which the
	protocol is activated and contain interface specific options.
	See <ref id="proto-iface" name="interface"> common options for
	detailed description.
</descrip>

<p>Interface specific options:

<descrip>
	<tag><label id="rip-iface-metric">metric <m/num/</tag>
	This option specifies the metric of the interface. When a route is
	received from the interface, its metric is increased by this value
	before further processing. Valid values are 1-255, but values higher
	than infinity has no further meaning. Default: 1.

	<tag><label id="rip-iface-mode">mode multicast|broadcast</tag>
	This option selects the mode for RIP to use on the interface. The
	default is multicast mode for RIPv2 and broadcast mode for RIPv1.
	RIPng always uses the multicast mode.

	<tag><label id="rip-iface-passive">passive <m/switch/</tag>
	Passive interfaces receive routing updates but do not transmit any
	messages. Default: no.

	<tag><label id="rip-iface-address">address <m/ip/</tag>
	This option specifies a destination address used for multicast or
	broadcast messages, the default is the official RIP (224.0.0.9) or RIPng
	(ff02::9) multicast address, or an appropriate broadcast address in the
	broadcast mode.

	<tag><label id="rip-iface-port">port <m/number/</tag>
	This option selects an UDP port to operate on, the default is the
	official RIP (520) or RIPng (521) port.

	<tag><label id="rip-iface-version">version 1|2</tag>
	This option selects the version of RIP used on the interface. For RIPv1,
	automatic subnet aggregation is not implemented, only classful network
	routes and host routes are propagated. Note that BIRD allows RIPv1 to be
	configured with features that are defined for RIPv2 only, like
	authentication or using multicast sockets. The default is RIPv2 for IPv4
	RIP, the option is not supported for RIPng, as no further versions are
	defined.

	<tag><label id="rip-iface-version-only">version only <m/switch/</tag>
	Regardless of RIP version configured for the interface, BIRD accepts
	incoming packets of any RIP version. This option restrict accepted
	packets to the configured version. Default: no.

	<tag><label id="rip-iface-split-horizon">split horizon <m/switch/</tag>
	Split horizon is a scheme for preventing routing loops. When split
	horizon is active, routes are not regularly propagated back to the
	interface from which they were received. They are either not propagated
	back at all (plain split horizon) or propagated back with an infinity
	metric (split horizon with poisoned reverse). Therefore, other routers
	on the interface will not consider the router as a part of an
	independent path to the destination of the route. Default: yes.

	<tag><label id="rip-iface-poison-reverse">poison reverse <m/switch/</tag>
	When split horizon is active, this option specifies whether the poisoned
	reverse variant (propagating routes back with an infinity metric) is
	used. The poisoned reverse has some advantages in faster convergence,
	but uses more network traffic. Default: yes.

	<tag><label id="rip-iface-demand-circuit">demand circuit <m/switch/</tag>
	Regular RIP sends periodic full updates on an interface. There is the
	Triggered RIP extension for demand circuits (<rfc id="2091">), which
	removes periodic updates and introduces update acknowledgments. When
	enabled, there is no RIP communication in steady-state network. Note
	that in order to work, it must be enabled on both sides. As there are
	no hello packets, it depends on hardware link state to detect neighbor
	failures. Also, it is designed for PtP links and it does not work
	properly with multiple RIP neighbors on an interface. Default: no.

	<tag><label id="rip-iface-check-zero">check zero <m/switch/</tag>
	Received RIPv1 packets with non-zero values in reserved fields should
	be discarded. This option specifies whether the check is performed or
	such packets are just processed as usual. Default: yes.

	<tag><label id="rip-iface-update-time">update time <m/number/</tag>
	Specifies the number of seconds between periodic updates. A lower number
	will mean faster convergence but bigger network load. Default: 30.

	<tag><label id="rip-iface-timeout-time">timeout time <m/number/</tag>
	Specifies the time interval (in seconds) between the last received route
	announcement and the route expiration. After that, the network is
	considered unreachable, but still is propagated with infinity distance.
	Default: 180.

	<tag><label id="rip-iface-garbage-time">garbage time <m/number/</tag>
	Specifies the time interval (in seconds) between the route expiration
	and the removal of the unreachable network entry. The garbage interval,
	when a route with infinity metric is propagated, is used for both
	internal (after expiration) and external (after withdrawal) routes.
	Default: 120.

	<tag><label id="rip-iface-ecmp-weight">ecmp weight <m/number/</tag>
	When ECMP (multipath) routes are allowed, this value specifies a
	relative weight used for nexthops going through the iface. Valid
	values are 1-256. Default value is 1.

	<tag><label id="rip-iface-auth">authentication none|plaintext|cryptographic</tag>
	Selects authentication method to be used. <cf/none/ means that packets
	are not authenticated at all, <cf/plaintext/ means that a plaintext
	password is embedded into each packet, and <cf/cryptographic/ means that
	packets are authenticated using some cryptographic hash function
	selected by option <cf/algorithm/ for each key. The default
	cryptographic algorithm for RIP keys is Keyed-MD5. If you set
	authentication to not-none, it is a good idea to add <cf>password</cf>
	section. Default: none.

	<tag><label id="rip-iface-pass">password "<m/text/"</tag>
	Specifies a password used for authentication. See <ref id="proto-pass"
	name="password"> common option for detailed description.

	<tag><label id="rip-iface-ttl-security">ttl security [<m/switch/ | tx only]</tag>
	TTL security is a feature that protects routing protocols from remote
	spoofed packets by using TTL 255 instead of TTL 1 for protocol packets
	destined to neighbors. Because TTL is decremented when packets are
	forwarded, it is non-trivial to spoof packets with TTL 255 from remote
	locations.

	If this option is enabled, the router will send RIP packets with TTL 255
	and drop received packets with TTL less than 255. If this option si set
	to <cf/tx only/, TTL 255 is used for sent packets, but is not checked
	for received packets. Such setting does not offer protection, but offers
	compatibility with neighbors regardless of whether they use ttl
	security.

	For RIPng, TTL security is a standard behavior (required by <rfc
	id="2080">) and therefore default value is yes. For IPv4 RIP, default
	value is no.

	<tag><label id="rip-iface-tx-class">tx class|dscp|priority <m/number/</tag>
	These options specify the ToS/DiffServ/Traffic class/Priority of the
	outgoing RIP packets. See <ref id="proto-tx-class" name="tx class"> common
	option for detailed description.

	<tag><label id="rip-iface-rx-buffer">rx buffer <m/number/</tag>
	This option specifies the size of buffers used for packet processing.
	The buffer size should be bigger than maximal size of received packets.
	The default value is 532 for IPv4 RIP and interface MTU value for RIPng.

	<tag><label id="rip-iface-tx-length">tx length <m/number/</tag>
	This option specifies the maximum length of generated RIP packets. To
	avoid IP fragmentation, it should not exceed the interface MTU value.
	The default value is 532 for IPv4 RIP and interface MTU value for RIPng.

	<tag><label id="rip-iface-check-link">check link <m/switch/</tag>
	If set, the hardware link state (as reported by OS) is taken into
	consideration. When the link disappears (e.g. an ethernet cable is
	unplugged), neighbors are immediately considered unreachable and all
	routes received from them are withdrawn. It is possible that some
	hardware drivers or platforms do not implement this feature.
	Default: yes.
</descrip>

<sect1>Attributes
<label id="rip-attr">

<p>RIP defines two route attributes:

<descrip>
	<tag><label id="rta-rip-metric">int rip_metric</tag>
	RIP metric of the route (ranging from 0 to <cf/infinity/). When routes
	from different RIP instances are available and all of them have the same
	preference, BIRD prefers the route with lowest <cf/rip_metric/. When a
	non-RIP route is exported to RIP, the default metric is 1.

	<tag><label id="rta-rip-tag">int rip_tag</tag>
	RIP route tag: a 16-bit number which can be used to carry additional
	information with the route (for example, an originating AS number in
	case of external routes). When a non-RIP route is exported to RIP, the
	default tag is 0.
</descrip>

<sect1>Example
<label id="rip-exam">

<p><code>
protocol rip {
	ipv4 {
		import all;
		export all;
	};
	interface "eth*" {
		metric 2;
		port 1520;
		mode multicast;
		update time 12;
		timeout time 60;
		authentication cryptographic;
		password "secret" { algorithm hmac sha256; };
	};
}
</code>


<sect>RPKI
<label id="rpki">

<sect1>Introduction

<p>The Resource Public Key Infrastructure (RPKI) is mechanism for origin
validation of BGP routes (<rfc id="6480">). BIRD supports only so-called
RPKI-based origin validation. There is implemented RPKI to Router (RPKI-RTR)
protocol (<rfc id="6810">). It uses some of the RPKI data to allow a router to
verify that the autonomous system announcing an IP address prefix is in fact
authorized to do so. This is not crypto checked so can be violated. But it
should prevent the vast majority of accidental hijackings on the Internet today,
e.g. the famous Pakistani accidental announcement of YouTube's address space.

<p>The RPKI-RTR protocol receives and maintains a set of ROAs from a cache
server (also called validator). You can validate routes (<rfc id="6483">,
<rfc id="6811">) using function <cf/roa_check()/ in filter and set it as import
filter at the BGP protocol. BIRD offers crude automatic re-validating of
affected routes after RPKI update, see option <ref id="proto-rpki-reload"
name="rpki reload">. Or you can use a BIRD client command <cf>reload in
<m/bgp_protocol_name/</cf> for manual call of revalidation of all routes.

<sect1>Supported transports
<p>
<itemize>
        <item>Unprotected transport over TCP uses a port 323. The cache server
        and BIRD router should be on the same trusted and controlled network
        for security reasons.
        <item>SSHv2 encrypted transport connection uses the normal SSH port
        22.
</itemize>

<sect1>Configuration

<p>We currently support just one cache server per protocol. However you can
define more RPKI protocols generally.

<code>
protocol rpki [&lt;name&gt;] {
        roa4 { table &lt;tab&gt;; };
        roa6 { table &lt;tab&gt;; };
        remote &lt;ip&gt; | "&lt;domain&gt;" [port &lt;num&gt;];
        port &lt;num&gt;;
        refresh [keep] &lt;num&gt;;
        retry [keep] &lt;num&gt;;
        expire [keep] &lt;num&gt;;
        transport tcp;
        transport ssh {
                bird private key "&lt;/path/to/id_rsa&gt;";
                remote public key "&lt;/path/to/known_host&gt;";
                user "&lt;name&gt;";
        };
}
</code>

<p>Alse note that you have to specify the ROA channel. If you want to import
only IPv4 prefixes you have to specify only roa4 channel. Similarly with IPv6
prefixes only. If you want to fetch both IPv4 and even IPv6 ROAs you have to
specify both channels.

<sect2>RPKI protocol options
<p>
<descrip>
        <tag>remote <m/ip/ | "<m/hostname/" [port <m/num/]</tag> Specifies
        a destination address of the cache server.  Can be specified by an IP
        address or by full domain name string.  Only one cache can be specified
        per protocol. This option is required.

        <tag>port <m/num/</tag> Specifies the port number. The default port
        number is 323 for transport without any encryption and 22 for transport
        with SSH encryption.

        <tag>refresh [keep] <m/num/</tag> Time period in seconds. Tells how
        long to wait before next attempting to poll the cache using a Serial
        Query or a Reset Query packet. Must be lower than 86400 seconds (one
        day). Too low value can caused a false positive detection of
        network connection problems.  A keyword <cf/keep/ suppresses updating
        this value by a cache server.
        Default: 3600 seconds

        <tag>retry [keep] <m/num/</tag> Time period in seconds between a failed
        Serial/Reset Query and a next attempt.  Maximum allowed value is 7200
        seconds (two hours). Too low value can caused a false positive
        detection of network connection problems.  A keyword <cf/keep/
        suppresses updating this value by a cache server.
        Default: 600 seconds

        <tag>expire [keep] <m/num/</tag> Time period in seconds. Received
        records are deleted if the client was unable to successfully refresh
        data for this time period.  Must be in range from 600 seconds (ten
        minutes) to 172800 seconds (two days).  A keyword <cf/keep/
        suppresses updating this value by a cache server.
        Default: 7200 seconds

	<tag>ignore max length <m/switch/</tag>
	Ignore received max length in ROA records and use max value (32 or 128)
	instead. This may be useful for implementing loose RPKI check for
	blackholes. Default: disabled.

        <tag>transport tcp</tag> Unprotected transport over TCP. It's a default
        transport. Should be used only on secure private networks.
        Default: tcp

        <tag>transport ssh { <m/SSH transport options.../ }</tag> It enables a
        SSHv2 transport encryption. Cannot be combined with a TCP transport.
        Default: off
</descrip>

<sect3>SSH transport options
<p>
<descrip>
	<tag>bird private key "<m>/path/to/id_rsa</m>"</tag>
	A path to the BIRD's private SSH key for authentication.
	It can be a <cf><m>id_rsa</m></cf> file.

	<tag>remote public key "<m>/path/to/known_host</m>"</tag>
	A path to the cache's public SSH key for verification identity
	of the cache server. It could be a path to <cf><m>known_host</m></cf> file.

	<tag>user "<m/name/"</tag>
	A SSH user name for authentication. This option is a required.
</descrip>

<sect1>Examples
<sect2>BGP origin validation
<p>Policy: Don't import <cf/ROA_INVALID/ routes.
<code>
roa4 table r4;
roa6 table r6;

protocol rpki {
	debug all;

	roa4 { table r4; };
	roa6 { table r6; };

	# Please, do not use rpki-validator.realmv6.org in production
	remote "rpki-validator.realmv6.org" port 8282;

	retry keep 5;
	refresh keep 30;
	expire 600;
}

filter peer_in_v4 {
	if (roa_check(r4, net, bgp_path.last) = ROA_INVALID) then
	{
		print "Ignore RPKI invalid ", net, " for ASN ", bgp_path.last;
		reject;
	}
	accept;
}

protocol bgp {
	debug all;
	local as 65000;
	neighbor 192.168.2.1 as 65001;
	ipv4 {
		import filter peer_in_v4;
		export none;
	};
}
</code>

<sect2>SSHv2 transport encryption
<p>
<code>
roa4 table r4;
roa6 table r6;

protocol rpki {
	debug all;

	roa4 { table r4; };
	roa6 { table r6; };

	remote 127.0.0.1 port 2345;
	transport ssh {
		bird private key "/home/birdgeek/.ssh/id_rsa";
		remote public key "/home/birdgeek/.ssh/known_hosts";
		user "birdgeek";
	};

	# Default interval values
}
</code>


<sect>Static
<label id="static">

<p>The Static protocol doesn't communicate with other routers in the network,
but instead it allows you to define routes manually. This is often used for
specifying how to forward packets to parts of the network which don't use
dynamic routing at all and also for defining sink routes (i.e., those telling to
return packets as undeliverable if they are in your IP block, you don't have any
specific destination for them and you don't want to send them out through the
default route to prevent routing loops).

<p>There are three classes of definitions in Static protocol configuration --
global options, static route definitions, and per-route options. Usually, the
definition of the protocol contains mainly a list of static routes. Static
routes have no specific attributes, but <ref id="rta-igp-metric" name="igp_metric">
attribute is used to compare static routes with the same preference.

<p>The list of static routes may contain multiple routes for the same network
(usually, but not necessary, distinquished by <cf/preference/ or <cf/igp_metric/),
but only routes of the same network type are allowed, as the static protocol
has just one channel. E.g., to have both IPv4 and IPv6 static routes, define two
static protocols, each with appropriate routes and channel.

<p>The Static protocol can be configured as MPLS-aware (by defining both the
primary channel and MPLS channel). In that case the Static protocol assigns
labels to IP routes and automatically announces corresponding MPLS route for
each labeled route.

<p>Global options:

<descrip>
	<tag><label id="static-check-link">check link <m/switch/</tag>
	If set, hardware link states of network interfaces are taken into
	consideration.  When link disappears (e.g. ethernet cable is unplugged),
	static routes directing to that interface are removed. It is possible
	that some hardware drivers or platforms do not implement this feature.
	Default: off.

	<tag><label id="static-igp-table">igp table <m/name/</tag>
	Specifies a table that is used for route table lookups of recursive
	routes. Default: the same table as the protocol is connected to.
</descrip>

<p>Route definitions (each may also contain a block of per-route options):

<sect1>Regular routes; MPLS switching rules

<p>There exist several types of routes; keep in mind that <m/prefix/ syntax is
<ref id="type-prefix" name="dependent on network type">.

<descrip>
	<tag>route <m/prefix/ [mpls <m/number>] via <m/ip/|<m/"interface"/ [<m/per-nexthop options/] [via ...]</tag>
	Regular routes may bear one or more <ref id="route-next-hop" name="next
	hops">. Every next hop is preceded by <cf/via/ and configured as shown.

	When the Static protocol is MPLS-aware, the optional <cf/mpls/ statement
	after <m/prefix/ specifies a static label for the labeled route, instead
	of using dynamically allocated label.

	<tag>route <m/prefix/ [mpls <m/number>] recursive <m/ip/ [mpls <m/num/[/<m/num/[/<m/num/[...]]]]</tag>
	Recursive nexthop resolves the given IP in the configured IGP table and
	uses that route's next hop. The MPLS stacks are concatenated; on top is
	the IGP's nexthop stack and on bottom is this route's stack.

	<tag>route <m/prefix/ [mpls <m/number>] blackhole|unreachable|prohibit</tag>
	Special routes specifying to silently drop the packet, return it as
	unreachable or return it as administratively prohibited. First two
	targets are also known as <cf/drop/ and <cf/reject/.
</descrip>

<p>When the particular destination is not available (the interface is down or
the next hop of the route is not a neighbor at the moment), Static just
uninstalls the route from the table it is connected to and adds it again as soon
as the destination becomes adjacent again.

<sect2>Per-nexthop options

<p>There are several options that in a case of multipath route are per-nexthop
(i.e., they can be used multiple times for a route, one time for each nexthop).
Syntactically, they are not separate options but just parts of <cf/route/
statement after each <cf/via/ statement, not separated by semicolons. E.g.,
statement <cf>route 10.0.0.0/8 via 192.0.2.1 bfd weight 1 via 192.0.2.2 weight
2;</cf> describes a route with two nexthops, the first nexthop has two per-nexthop
options (<cf/bfd/ and <cf/weight 1/), the second nexthop has just <cf/weight 2/.

<descrip>
	<tag><label id="static-route-bfd">bfd <m/switch/</tag>
	The Static protocol could use BFD protocol for next hop liveness
	detection. If enabled, a BFD session to the route next hop is created
	and the static route is BFD-controlled -- the static route is announced
	only if the next hop liveness is confirmed by BFD. If the BFD session
	fails, the static route (or just the affected nexthop from multiple
	ones) is removed. Note that this is a bit different compared to other
	protocols, which may use BFD as an advisory mechanism for fast failure
	detection but ignore it if a BFD session is not even established. Note
	that BFD protocol also has to be configured, see <ref id="bfd" name="BFD">
	section for details. Default value is no.

	<tag><label id="static-route-mpls">mpls <m/num/[/<m/num/[/<m/num/[...]]]</tag>
	MPLS labels that should be pushed to packets forwarded by the route.
	The option could be used for both IP routes (on MPLS ingress routers)
	and MPLS switching rules (on MPLS transit routers). Default value is
	no labels.

	<tag><label id="static-route-onlink">onlink <m/switch/</tag>
	Onlink flag means that the specified nexthop is accessible on the
	(specified) interface regardless of IP prefixes of the interface. The
	interface must be attached to nexthop IP address using link-local-scope
	format (e.g. <cf/192.0.2.1%eth0/). Default value is no.

	<tag><label id="static-route-weight">weight <m/switch/</tag>
	For multipath routes, this value specifies a relative weight of the
	nexthop. Allowed values are 1-256. Default value is 1.
</descrip>

<sect1>Route Origin Authorization

<p>The ROA config is just <cf>route <m/prefix/ max <m/int/ as <m/int/</cf> with no nexthop.

<sect1>Flowspec
<label id="flowspec-network-type">

<p>The flow specification are rules for routers and firewalls for filtering
purpose. It is described by <rfc id="5575">. There are 3 types of arguments:
<m/inet4/ or <m/inet6/ prefixes, numeric matching expressions and bitmask
matching expressions.

Numeric matching is a matching sequence of numbers and ranges separeted by a
commas (<cf/,/) (e.g. <cf/10,20,30/). Ranges can be written using double dots
<cf/../ notation (e.g. <cf/80..90,120..124/). An alternative notation are
sequence of one or more pairs of relational operators and values separated by
logical operators <cf/&&/ or <cf/||/. Allowed relational operators are <cf/=/,
<cf/!=/, <cf/</, <cf/<=/, <cf/>/, <cf/>=/, <cf/true/ and <cf/false/.

Bitmask matching is written using <m/value/<cf>/</cf><m/mask/ or
<cf/!/<m/value/<cf>/</cf><m/mask/ pairs. It means that <cf/(/<m/data/ <cf/&/
<m/mask/<cf/)/ is or is not equal to <m/value/. It is also possible to use
multiple value/mask pairs connected by logical operators <cf/&&/ or <cf/||/.
Note that for negated matches, value must be either zero or equal to bitmask
(e.g. !0x0/0xf or !0xf/0xf, but not !0x3/0xf).

<sect2>IPv4 Flowspec

<p><descrip>
	<tag><label id="flow-dst">dst <m/inet4/</tag>
	Set a matching destination prefix (e.g. <cf>dst 192.168.0.0/16</cf>).
	Only this option is mandatory in IPv4 Flowspec.

	<tag><label id="flow-src">src <m/inet4/</tag>
	Set a matching source prefix (e.g. <cf>src 10.0.0.0/8</cf>).

	<tag><label id="flow-proto">proto <m/numbers-match/</tag>
	Set a matching IP protocol numbers (e.g.  <cf/proto 6/).

	<tag><label id="flow-port">port <m/numbers-match/</tag>
	Set a matching source or destination TCP/UDP port numbers (e.g.
	<cf>port 1..1023,1194,3306</cf>).

	<tag><label id="flow-dport">dport <m/numbers-match/</tag>
	Set a matching destination port numbers (e.g. <cf>dport 49151</cf>).

	<tag><label id="flow-sport">sport <m/numbers-match/</tag>
	Set a matching source port numbers (e.g. <cf>sport = 0</cf>).

	<tag><label id="flow-icmp-type">icmp type <m/numbers-match/</tag>
	Set a matching type field number of an ICMP packet (e.g. <cf>icmp type
	3</cf>)

	<tag><label id="flow-icmp-code">icmp code <m/numbers-match/</tag>
	Set a matching code field number of an ICMP packet (e.g. <cf>icmp code
	1</cf>)

	<tag><label id="flow-tcp-flags">tcp flags <m/bitmask-match/</tag>
	Set a matching bitmask for TCP header flags (aka control bits) (e.g.
	<cf>tcp flags 0x03/0x0f;</cf>). The maximum length of mask is 12 bits
	(0xfff).

	<tag><label id="flow-length">length <m/numbers-match/</tag>
	Set a matching packet length (e.g. <cf>length > 1500</cf>)

	<tag><label id="flow-dscp">dscp <m/numbers-match/</tag>
	Set a matching DiffServ Code Point number (e.g. <cf>dscp 8..15</cf>).

	<tag><label id="flow-fragment">fragment <m/fragmentation-type/</tag>
	Set a matching type of packet fragmentation. Allowed fragmentation
	types are <cf/dont_fragment/, <cf/is_fragment/, <cf/first_fragment/,
	<cf/last_fragment/ (e.g. <cf>fragment is_fragment &&
	!dont_fragment</cf>).
</descrip>

<p><code>
protocol static {
	flow4;

	route flow4 {
		dst 10.0.0.0/8;
		port > 24 && < 30 || 40..50,60..70,80 && >= 90;
		tcp flags 0x03/0x0f;
		length > 1024;
		dscp = 63;
		fragment dont_fragment, is_fragment || !first_fragment;
	};
}
</code>

<sect2>Differences for IPv6 Flowspec

<p>Flowspec IPv6 are same as Flowspec IPv4 with a few exceptions.
<itemize>
	<item>Prefixes <m/inet6/ can be specified not only with prefix length,
	but with prefix <cf/offset/ <m/num/ too (e.g.
	<cf>::1234:5678:9800:0000/101 offset 64</cf>). Offset means to don't
	care of <m/num/ first bits.
	<item>IPv6 Flowspec hasn't mandatory any flowspec component.
	<item>In IPv6 packets, there is a matching the last next header value
	for a matching IP protocol number (e.g. <cf>next header 6</cf>).
	<item>It is not possible to set <cf>dont_fragment</cf> as a type of
	packet fragmentation.
</itemize>

<p><descrip>
	<tag><label id="flow6-dst">dst <m/inet6/ [offset <m/num/]</tag>
	Set a matching destination IPv6 prefix (e.g. <cf>dst
	::1c77:3769:27ad:a11a/128 offset 64</cf>).

	<tag><label id="flow6-src">src <m/inet6/ [offset <m/num/]</tag>
	Set a matching source IPv6 prefix (e.g. <cf>src fe80::/64</cf>).

	<tag><label id="flow6-next-header">next header <m/numbers-match/</tag>
	Set a matching IP protocol numbers (e.g. <cf>next header != 6</cf>).

	<tag><label id="flow6-label">label <m/bitmask-match/</tag>
	Set a 20-bit bitmask for matching Flow Label field in IPv6 packets
	(e.g. <cf>label 0x8e5/0x8e5</cf>).
</descrip>

<p><code>
protocol static {
	flow6 { table myflow6; };

	route flow6 {
		dst fec0:1122:3344:5566:7788:99aa:bbcc:ddee/128;
		src 0000:0000:0000:0001:1234:5678:9800:0000/101 offset 63;
		next header = 23;
		sport > 24 && < 30 || = 40 || 50,60,70..80;
		dport = 50;
		tcp flags 0x03/0x0f && !0/0xff || 0x33/0x33;
		fragment !is_fragment || !first_fragment;
		label 0xaaaa/0xaaaa && 0x33/0x33;
	};
}
</code>

<sect1>Per-route options
<p>
<descrip>
	<tag><label id="static-route-filter"><m/filter expression/</tag>
	This is a special option that allows filter expressions to be configured
	on per-route basis. Can be used multiple times. These expressions are
	evaluated when the route is originated, similarly to the import filter
	of the static protocol. This is especially useful for configuring route
	attributes, e.g., <cf/ospf_metric1 = 100;/ for a route that will be
	exported to the OSPF protocol.
</descrip>

<sect1>Example static configs
<label id="static-example">

<p><code>
protocol static {
	ipv4 { table testable; };	# Connect to a non-default routing table
	check link;			# Advertise routes only if link is up
	route 0.0.0.0/0 via 198.51.100.130; # Default route
	route 10.0.0.0/8		# Multipath route
		via 198.51.100.10 weight 2
		via 198.51.100.20 bfd	# BFD-controlled next hop
		via 192.0.2.1;
	route 203.0.113.0/24 blackhole; # Sink route
	route 10.2.0.0/24 via "arc0";	# Secondary network
	route 192.168.10.0/24 via 198.51.100.100 {
		ospf_metric1 = 20;	# Set extended attribute
	};
	route 192.168.11.0/24 via 198.51.100.100 {
		ospf_metric2 = 100;	# Set extended attribute
		ospf_tag = 2;		# Set extended attribute
	};
	route 192.168.12.0/24 via 198.51.100.100 {
		bgp_community.add((65535, 65281));	# Set extended BGP attribute
		bgp_large_community.add((64512, 1, 1));	# Set extended BGP attribute
	};
}

protocol static {
	ipv6;						# Channel is mandatory
	route 2001:db8:10::/48 via 2001:db8:1::1;	# Route with global nexthop
	route 2001:db8:20::/48 via fe80::10%eth0;	# Route with link-local nexthop
	route 2001:db8:30::/48 via fe80::20%'eth1.60';	# Iface with non-alphanumeric characters
	route 2001:db8:40::/48 via "eth2";		# Direct route to eth2
	route 2001:db8::/32 unreachable;		# Unreachable route
	route ::/0 via 2001:db8:1::1 bfd;		# BFD-controlled default route
}
</code>


<chapt>Conclusions
<label id="conclusion">

<sect>Future work
<label id="future-work">

<p>Although BIRD supports all the commonly used routing protocols, there are
still some features which would surely deserve to be implemented in future
versions of BIRD:

<itemize>
<item>Opaque LSA's
<item>Route aggregation and flap dampening
<item>Multicast routing protocols
<item>Ports to other systems
</itemize>


<sect>Getting more help
<label id="help">

<p>If you use BIRD, you're welcome to join the bird-users mailing list
(<HTMLURL URL="mailto:bird-users@network.cz" name="bird-users@network.cz">)
where you can share your experiences with the other users and consult
your problems with the authors. To subscribe to the list, visit
<HTMLURL URL="http://bird.network.cz/?m_list" name="http://bird.network.cz/?m_list">.
The home page of BIRD can be found at <HTMLURL URL="http://bird.network.cz/" name="http://bird.network.cz/">.

<p>BIRD is a relatively young system and it probably contains some bugs. You can
report any problems to the bird-users list and the authors will be glad to solve
them, but before you do so, please make sure you have read the available
documentation and that you are running the latest version (available at
<HTMLURL URL="ftp://bird.network.cz/pub/bird" name="bird.network.cz:/pub/bird">).
(Of course, a patch which fixes the bug is always welcome as an attachment.)

<p>If you want to understand what is going inside, Internet standards are a good
and interesting reading. You can get them from
<HTMLURL URL="ftp://ftp.rfc-editor.org/" name="ftp.rfc-editor.org"> (or a
nicely sorted version from <HTMLURL URL="ftp://atrey.karlin.mff.cuni.cz/pub/rfc"
name="atrey.karlin.mff.cuni.cz:/pub/rfc">).

<p><it/Good luck!/

</book>

<!--
LocalWords:  GPL IPv GateD BGPv RIPv OSPFv Linux sgml html dvi sgmltools
LocalWords:  linuxdoc dtd descrip config conf syslog stderr auth ospf bgp Mbps
LocalWords:  router's eval expr num birdc ctl UNIX if's enums bool int ip GCC
LocalWords:  len ipaddress pxlen netmask enum bgppath bgpmask clist gw md eth
LocalWords:  RTS printn quitbird iBGP AS'es eBGP RFC multiprotocol IGP
LocalWords:  EGP misconfigurations keepalive pref aggr aggregator BIRD's RTC
LocalWords:  OS'es AS's multicast nolisten misconfigured UID blackhole MRTD MTU
LocalWords:  uninstalls ethernets IP binutils ANYCAST anycast dest RTD ICMP rfc
LocalWords:  compat multicasts nonbroadcast pointopoint loopback sym stats
LocalWords:  Perl SIGHUP dd mm yy HH MM SS EXT IA UNICAST multihop Discriminator txt
LocalWords:  proto wildcard
-->