FRR-BGPD(8)

📦
version 10.7.1 · 23 versions
, in ubuntu stonking

✨ FRR-BGPD ✨ (that's frr-bgpd(8) bestie, section 8)

okay so like. this is the manual page for frr-bgpd, part of the FRR (FRRouting) family, and we're about to go THROUGH it. no skipping. every flag, every port, every ✨detail✨. grab a snack. 🍓

💖 NAME 💖

frr-bgpd — a BGPv4, BGPv4+, BGPv4- routing engine for use with FRRouting.

(yes it says BGPv4 three times with different lil suffix thingies. that's not a typo, that's just the vibe. 🦋)

🌸 SYNOPSIS 🌸

bgpd [-h] [-v]

bgpd [-d|-t|-dt] [-C] [-f config-file] [-i pid-file] [-z zclient-path]
[-u user] [-g group] [-A vty-addr] [-P vty-port] [-M module[:options]]
[-N pathspace] [--vty_socket vty-path] [--moduledir module-path]

👑 DESCRIPTION 👑

bgpd is a routing component that works with the FRRouting routing engine.

that's it. that's the whole description. short and iconic, we love a main character who doesn't over-explain. 💅

✨ OPTIONS ✨

here's the OPTIONS situation, all available for the bgpd command:

🫶 HELP AND VERSION

  • -h, --help — print a short lil description of the daemon's command line options. it's giving "help me help you."
  • -v, --version — print version and build info for the daemon. iykyk.
both of these options inhibit normal operation and will immediately exit. like they literally cannot be stopped. bgpd sees -h and it's out. 🚪

🌈 PROCESS CONTROL

these options control background operation:

  • -d, --daemon — launches the process in background/daemon mode, forking and detaching from the terminal.

the parent process will delay its exit until the daemon/child has finished its initialization and has entered its main loop. this is important for zebra startup because the other daemons will attempt to connect to zebra. a return from zebra -d guarantees its readiness to accept these connections. (so like, -d is zebra's way of saying "im ready bestie, come thru." 🎀)

  • -t, --terminal — opens an interactive VTY session on the terminal, allowing for both state and configuration operations. note that the terminal starts operating after startup has completed and the configuration file has been loaded.

the process will exit when end of file is detected on the terminal. it is possible to daemonize a process started with -t (but without -d) by sending SIGQUIT to the process (normally mapped to a ^ keypress.)

the combination of --daemon and --terminal will delay the daemon from going into background until the terminal session ends (by end of file.)

if the process receives SIGINT (e.g. a ^C keypress) in this mode, it will exit instead of daemonizing. (respectfully, it said no. 💀)

it is safe to suspend (SIGTSTP / ^Z) the terminal session opened by the previous two options — this will only stop the terminal but not the protocol daemon itself (which runs in a separate second process.) so like, the terminal takes a nap, the daemon keeps slaying. 😴✨

🎀 CONFIGURATION AND PATHS

these options control configuration and file system locations for frr processes:

  • -f, --config_file config-file — specify a configuration file to be used instead of the default /etc/frr/<daemon>.conf file.

note that the daemon will attempt to write to this file if the write file command is issued on its VTY interface or through vtysh. (she writes back. 💖)

  • -C, --dryrun — load the configuration file and check its validity, then exit. a little vibe check, if you will. ✅
  • -i, --pid_file pid-file — output a pid file to a location other than the default /var/run/frr/<daemon>.pid.
  • -z, --socket zclient-path — override the path of the ZAPI socket used to communicate between zebra and the various protocol daemons. the default is /var/run/frr/zserv.api. the value of this option must be the same across all daemons. 🫂
  • -N, --pathspace pathspace — insert pathspace into all default paths, changing the defaults to:
/etc/frr/pathspace/<daemon>.conf
/var/run/frr/pathspace/<daemon>.pid
/var/run/frr/pathspace/<daemon>.vty
/var/run/frr/pathspace/zserv.api

´.´ and ´/´ characters will not be accepted in pathspace, but the empty string will be accepted. (yes the empty string is fine. the dot and slash are not. make it make sense. 🙈)

note that this only changes the respective defaults — it has no effect on the respective path if the -f, -i, -z or --vty_socket options are used.

the purpose of this option is to easily group all file system related bits together for running multiple fully-separate "logical routers" on a system, particularly with Linux network namespaces. groups of daemons running with distinct pathspace values will be completely unaware of each other and not interact in any way. (they don't know each other, they've never met, no cap. 💅)

this option does not do any system setup (like network namespaces.) this must be done by the user, for example by running:

ip netns exec namespace <daemon> -N namespace

💜 PROCESS CREDENTIALS

  • -u, --user user — (default: frr)
  • -g, --group group — (default: frr)

change the user/group which the daemon will switch to.

  • -S, --skip_runas — skip setting the process effective user and group.
note that there is an additional group, frrvty, which controls group ownership of the VTY sockets. the name of this group cannot currently be changed, and user must be a member of this group. 🎫

🦋 VTY SETUP 🦋

these following options control the daemon's VTY (interactive command line) interface. the interface is available over TCP, using the telnet protocol, as well as through the vtysh frontend. (two ways in, both iconic.)

  • -A, --vty_addr vty-addr — specify an IP/IPv6 address to bind the TCP VTY interface to. it is generally recommended to specify ::1 or 127.0.0.1. for reasons of backwards compatibility, the default is to listen on all interfaces. (which, bestie, we do not love, but it is what it is.)
  • -P, --vty_port vty-port — override the daemon's default TCP VTY port (each daemon has a different default value upwards of 2600, listed below.) specifying 0 disables the TCP VTY interface.

default ports are:

zebra           2601
ripd            2602
ripngd          2603
ospfd           2604
bgpd            2605
ospf6d          2606
isisd           2608
babeld          2609
nhrpd           2610
pimd            2611
ldpd            2612
eigrpd          2613
pbrd            2615
staticd         2616
bfdd            2617
fabricd         2618
vrrpd           2619

port 2607 is used for ospfd's Opaque LSA API. (it's special. it's doing its own thing. ⭐)

  • --vty_socket vty-path — overrides the directory used for the <daemon>.vty sockets. vtysh connects to these sockets in order to access each daemon's VTY. default: /var/run/frr[/<pathspace>]
NB: unlike the other options, this option specifies a directory, not a full path. this option is primarily used by the SNAP packaging system, its semantics may change. it should not be necessary in most other scenarios. (fr fr, you probably don't need this one.)

🧚 MODULE LOADING

frr supports optional dynamically loadable modules, although these can only be loaded at startup. the set of available modules may vary across distributions and packages, and modules may be available for installation as separate packages.

  • -M, --module module[:options] — load a module named module, optionally passing options to it.

if there is a ´/´ character in module, the value is assumed to be a pathname to a module.

if there is no ´/´ character, the module directory (see next option) is searched first for a module named <daemon>_<module>.so, then for <module>.so. this allows for a module to exist in variations appropriate for particular daemons, e.g. zebra_snmp and bgp_snmp, with the correct one selected by -M snmp. (the algorithm knows which one is yours. 🔍)

the meaning of options is specific to the module being loaded. most modules currently ignore it. (they're like "lol okay" and move on. 💀)

modules are loaded in the order as listed on the command line. this is not generally relevant.

  • --moduledir module-path — look for modules in the module-path directory instead of the default /usr/lib/frr/modules. (this path is not affected by the -N option.)

the list of loaded modules can be inspected at runtime with the show modules VTY command. 👀

  • -p, --bgp_port <port> — set the bgp protocol's port number. when port number is 0, that means do not listen bgp port.
  • -l, --listenon — specify a specific IP address for bgpd to listen on, rather than its default of 0.0.0.0 / ::. this can be useful to constrain bgpd to an internal address, or to run multiple bgpd processes on one host. (multiple bgp besties, one machine. obsessed.)
  • -n, --no_kernel — do not install learned routes into the linux kernel. this option is useful for a route-reflector environment or if you are running multiple bgp processes in the same namespace. this option is different than the --no_zebra option in that a ZAPI connection is made. ✨
  • -e, --ecmp — run BGP with a limited ecmp capability, that is different than what BGP was compiled with. the value specified must be greater than 0 and less than or equal to the MULTIPATH_NUM specified on compilation.
  • -Z, --no_zebra — do not communicate with zebra at all. this is different than the --no_kernel option in that we do not even open a ZAPI connection to the zebra process. (fully ghosting zebra. no calls, no texts, nothing. 💔)
  • -s, --socket_size — when opening tcp connections to our peers, set the socket send buffer size that the kernel will use for the peers socket. this option is only really useful at a very large scale. experimentation should be done to see if this is helping or not at the scale you are running at. (screenshot the numbers, run it twice, compare. science. 🔬)

🎀 LABEL MANAGER

  • -I, --int_num — set zclient id. this is required when using Zebra label manager in proxy mode. 🏷️

🫶 FILES 🫶

  • /usr/lib/frr/bgpd — the default location of the bgpd binary. 🏠
  • /etc/frr/bgpd.conf — the default location of the bgpd config file. 📄
  • $(PWD)/bgpd.log — if the bgpd process is configured to output logs to a file, then you will find this file in the directory where you started bgpd. (wherever you launched her, she leaves a lil diary. 📓)

💀 WARNING 💀

this man page is intended to be a quick reference for command line options. the definitive document is the info file frr latest or the documentation available on the project website at https://frrouting.org/.

(so like, this is the sparknotes. the real one is the info file. 💖)

🌟 DIAGNOSTICS 🌟

the daemon may log to standard output, to a VTY, to a log file, or through syslog to the system logs. FRR supports many debugging options — see the Info file, web docs or source for details. 🔍

🌈 SEE ALSO 🌈

frr-zebra(8), vtysh(1), frr-ripd(8), frr-ripngd(8), frr-ospfd(8), frr-ospf6d(8), frr-bgpd(8), frr-isisd(8), frr-babeld(8), frr-nhrpd(8), frr-pimd(8), frr-pbrd(8), frr-ldpd(8), frr-eigrpd(8), frr-staticd(8), frr-fabricd(8), frr-vrrpd(8), mtracebis(8) — and also https://frrouting.org/ 💅

🍓 BUGS 🍓

FRR eats bugs for breakfast. if you have food for the maintainers, please email <dev@lists.frrouting.org>. 🐛🍽️

okay so the legal side of things, and it's literally one line bestie: this page is Copyright 2017, FRR. that's it, that's the whole notice 💅 the FRR girlies hold the copyright from 2017 and we respect the paperwork periodt 📄✨

and that's a wrap on FRR-BGPD(8) 🦋 this whole vibe came out of the FRR manual, source FRR-BGPD(8), version latest, dated Sep 22, 2026 💖 go configure your daemon, bestie, you're literally the main character of your own routing table now 👑🌸

📦 where this page comes from