April 2024 - Bird-users - Mailing Lists

"<ERR> bfd1: Socket error: Destination address required"
by Fran 30 Apr '24

30 Apr '24

Hello there, I get the error message: "<ERR> bfd1: Socket error: Destination address required" (ubuntu 22.04, bird via ppa 2.15.1) running BFD and MP-BGP via IPv6 over wireguard interfaces. Not all BGP neighbors are reachable. Relevant (I hope) config: log "/var/log/bird/bird.log" all; protocol device { } protocol bfd { accept ipv6 direct; interface "wg_blue*"; interface "wg_green" { multiplier 3; interval 500 ms; }; neighbor 2001:db8:f000::2 dev "wg_green" local 2001:db8:f000::1; neighbor 2001:db8:f000:ffff::2 dev "wg_blue2" local 2001:db8:f000:ffff::1; neighbor 2001:db8:f000:fffc::5 dev "wg_blue1" local 2001:db8:f000:fffc::1; neighbor 2001:db8:f000:fffc::4 dev "wg_blue1" local 2001:db8:f000:fffc::1; neighbor 2001:db8:f000:fffc::3 dev "wg_blue1" local 2001:db8:f000:fffc::1; neighbor 2001:db8:f000:fffc::2 dev "wg_blue1" local 2001:db8:f000:fffc::1; } protocol bgp EXAMPLE { local as 65001; neighbor 2001:db8:f000::2 as 65044; direct; med metric yes; ipv4 { ... }; ipv6 { ... }; bfd on; } If I remove the BFD config for the non-established BGP neighbors, the error message disappears (although there are still non-established BFD sessions for established BGP neighbors (BFD not yet confed on the other end)). At first I did not create neighbor statements for BFD, then I added "dev" and "local" options - no improvement. No revelations with "debug protocols all". Any ideas? Thanks a lot! Best, fran

3 4

Large communities indicating RPKI VALID status
by Nigel Kukard 30 Apr '24

30 Apr '24

Hi all, I was busy reading https://bgpfilterguide.nlnog.net/guides/reject_invalids/ and noticed the following text... Note: REALLY DONT store the validation state inside a bgp_community or bgp_large_community or bgp_ext_community variables. It can cause CPU & memory overload resulting in convergence performance issues. I was wondering if this is still an issue and if it would still be a bad idea to indicate that RPKI was VALID using communities on multiple full BGP feeds? Is anyone doing this at present? are you seeing significant load? Kind Regards Nigel

7 10

`default cost` setting is not applied
by Benoit Chesneau 29 Apr '24

29 Apr '24

Hi, I have the following configuration: ``` define ospf_v4_routes = [ 198.19.0.0/16 ]; filter ospf_export { if (net.type = NET_IP4 && ! (net ~ [ 0.0.0.0/0 ])) then reject; accept; } filter ospf_import { if (net.type = NET_IP4 && net ~ ospf_v4_routes) then accept; reject; } protocol ospf v2 ospfv4 { debug all; ipv4 { import filter ospf_import; export filter ospf_export; }; area 0.0.0.0 { interface "lo1" { stub yes; }; interface "vlan600" { type ptp; cost 15; bfd on; }; };} ``` I would expect the metric around 300 + cost of the interface but I get this route metric "E2 0.0.0.0/0 [15/10000] via 198.19.4.65, vlan600" in the switch. When I set the metric manually in the export function: ``` filter ospf_export { if (net.type = NET_IP4 && ! (net ~ [ 0.0.0.0/0 ])) then reject; ospf_metric1 = 300; unset(ospf_metric2); accept;} ``` I get the correct metric: ``` E1 0.0.0.0/0 [315] via 198.19.4.65, vlan600 ``` Any idea what could be the issue ? Did I misunderstood the usage of the `default cost` setting? Benoît

1 0

How to set ospf as external route as ext1?
by chan alfie 29 Apr '24

29 Apr '24

The default ospf ASE will be set as rts_ospf_ext2, how to change it, and set the ospf ASE as rts_ospf_ext1?

2 1

How to make ospf advertise stubnet base on the interface/subnet state?
by chan alfie 29 Apr '24

29 Apr '24

Hi, community I have create a loopback interface on linux and assign 4 subnets on it. ``` ip link add dev lo235 type dummy ip link set lo235 up ip addr add 10.23.5.1/30 dev lo235 ip addr add 10.23.5.5/30 dev lo235 ip addr add 10.23.5.9/30 dev lo235 ip addr add 10.23.5.13/30 dev lo235 ``` stubnet <prefix> { summary;}; is used on an originator routers(r235) ospf area, and it indeed propagted the summary prefix instead of subnets, here is the config: r235: ``` area 10.23.0.0 { ... stubnet 10.23.5.0/28 { summary; } ... } ``` r235: ``` show ospf state ... router 10.23.0.5 distance 0 ... stubnet 10.23.5.0/28 metric 10 ... ``` but when I shutdown the lo235 or delete all of the subnets, the 10.23.5.0/28 is still advertised by r235. r235 ``` ip link set lo235 down or ip addr del 10.23.5.1/30 dev lo235 ip addr del 10.23.5.5/30 dev lo235 ip addr del 10.23.5.9/30 dev lo235 ip addr del 10.23.5.13/30 dev lo235 ``` r235 ``` bird> show ospf state router 10.23.0.5 distance 0 ... stubnet 10.23.5.0/28 metric 10 ``` How to make ospf advertise stubnet base on the interface/subnet state, (interface up/down, subnets existed or not)? By the way, networks{} clause works as expected. When the subnet of originator router is dismiss, the ABR's networks will take effect(disappear of appear).

2 1

IPv4 next-hops over IPv6
by Jay Hanke 28 Apr '24

28 Apr '24

Has anyone implemented lab or production IPv4 next hops over an IPv6 only IX vlan with BIRD as the route server? I'm interested to hear your experiences. Specifically router vendor interop. Jay

7 12

Current bird via bookworm-backports?
by Oliver Seufer 26 Apr '24

26 Apr '24

Hello, The version 2.0.12 included in Debian Bookworm unfortunately lacks a few current features. I would therefore greatly appreciate it if a Debian package with the latest version could be provided. Either through bookworm-backports or directly through bird.network.cz Is there a possibility that this will be provided by the bird project? Like it was done for Debian Bullseye back then? I greatly appreciate your work on bird, and I also enjoy trying out new features. Best regards, Oliver Seufer

7 11

Re: Current bird via bookworm-backports?
by Chriztoffer Hansen 24 Apr '24

24 Apr '24

Kees, On Wed, 24 Apr 2024 at 14:28, Kees Meijs | Nefos <keesm(a)nefos.nl> wrote: > Would you like a mirror in the Netherlands as well? > > If so, what resources would you need? This is the current size of the individual repo's 679M knot-resolver 3.9G knot-dns 277M bird2 Access via http and https are usually the minimum I see with public repos. ftp, rsync, TOR endpoints are optionally provided. ********************** DotSrc publicises the repo's via http, https, ftp, rsync, and TOR. http://mirrors.dotsrc.org/knot-resolver http://mirrors.dotsrc.org/knot-dns http://mirrors.dotsrc.org/bird2 https://mirrors.dotsrc.org/knot-resolver https://mirrors.dotsrc.org/knot-dns https://mirrors.dotsrc.org/bird2 ftp://mirrors.dotsrc.org/knot-resolver ftp://mirrors.dotsrc.org/knot-dns ftp://mirrors.dotsrc.org/bird2 rsync://mirrors.dotsrc.org/knot-resolver rsync://mirrors.dotsrc.org/knot-dns rsync://mirrors.dotsrc.org/bird2 http://dotsrccccbidkzg7oc7oj4ugxrlfbt64qebyunxbrgqhxiwj3nl6vcad.onion/knot-… http://dotsrccccbidkzg7oc7oj4ugxrlfbt64qebyunxbrgqhxiwj3nl6vcad.onion/knot-… http://dotsrccccbidkzg7oc7oj4ugxrlfbt64qebyunxbrgqhxiwj3nl6vcad.onion/bird2 Cheers, Chriztoffer - Volunteer Foreningen for DotSrc - https://dotsrc.org/

1 0

Re: Current bird via bookworm-backports?
by Chriztoffer Hansen 23 Apr '24

23 Apr '24

A copy to the list On Tue, 23 Apr 2024 at 20:00, <ch(a)dotsrc.org> wrote: > > Dear Jakub, > > Fantastic to hear about the RSYNC option is now available. > > I enabled an RSYNC mirror for our volunteer run MIRROR service - > dotsrc.org - in Denmark. > > https://mirrors.dotsrc.org/bird2/ > > https://mirrors.dotsrc.org/knot-dns/ > > https://mirrors.dotsrc.org/knot-dns/ > > Cheers, > > Chriztoffer - Volunteer > Foreningen for DotSrc - https://dotsrc.org/ > > On Tue, 23 Apr 2024 at 14:50, Jakub Ružička <jakub.ruzicka(a)nic.cz> wrote: > > > > Hello, > > > > I maintain pkg.labs.nic.cz and rsync wasn't supported until today, but > > you're a second user interested in mirroring the repos so I had a look > > and enabled the rsync daemon: > > > > $ rsync rsync://pkg.labs.nic.cz/ > > bird2 "BIRD 2.x Debian & Ubuntu packages repo" > > knot-resolver "Knot Resolver Debian & Ubuntu packages repo" > > knot-dns "Knot DNS Debian & Ubuntu packages repo" > > > > $ rsync -av rsync://pkg.labs.nic.cz/bird2/ bird2-repo > > > > I've also CC'd Arnaud who was interested in this. > > > > Let me know if this setup is working for you, or what needs to be > > changed in order for it to work as users (you) expect. > > > > > > Cheers, > > Jakub Ružička > > > > On 4/18/24 18:41, netravnen%2Bbirdlist(a)gmail.com wrote: > > > Dear Maria, > > > > > > Do you support mirrors (rsync) for your package site at > > > https://pkg.labs.nic.cz/doc/?project=bird ? > > > > > > I am part of a group operating a local mirror in DK. > > > > > > On Thu, 18 Apr 2024 at 14:31, Maria Matejka via Bird-users > > > <bird-users(a)network.cz> wrote: > > >> Hello Oliver, > > >> > > >> On Thu, Apr 18, 2024 at 11:24:31AM +0200, Oliver Seufer via Bird-users wrote: > > >> > > >> Hello, > > >> > > >> The version 2.0.12 included in Debian Bookworm unfortunately lacks a few current features. I would therefore greatly appreciate it if a Debian package with the latest version could be provided. Either through bookworm-backports or directly through bird.network.cz > > >> > > >> Please try this repository:https://pkg.labs.nic.cz/doc/?project=bird > > >> > > >> It should have everything you need. > > >> > > >> Maria > > >> > > >> – Maria Matejka (she/her) | BIRD Team Leader | CZ.NIC, z.s.p.o.

1 0

How to interpret `!` in show route when a route was successfully inserted into the kernel table?
by Ivan Kurnosov 22 Apr '24

22 Apr '24

Hi team, can somebody please help interpret the following output: ``` 10.50.0.1/32 unicast [static1 02:16:07.523] * (200) dev xfrm-40 10.240.224.2/32 unicast [hq 02:27:17.417] ! (100/?) [AS65020i] via 10.50.0.1 on xfrm-40 10.240.2.0/24 unicast [hq 02:27:17.417] ! (100/?) [AS65020i] via 10.50.0.1 on xfrm-40 ``` The current (birdc) machine establishes an ipsec tunnel and 10.50.0.1 is a firewall on the other side of the ipsec tunnel. And kernel routes look like ``` 10.50.0.1 dev xfrm-40 proto bird scope link src 10.80.1.76 metric 32 10.240.2.0/24 via 10.50.0.1 dev xfrm-40 proto bird metric 32 10.240.224.2 via 10.50.0.1 dev xfrm-40 proto bird metric 32 ``` So, technically all works: packets flow as expected. It's only the `!` exclamation mark and `(100/?)` question mark that I cannot interpret. Is it something I should worry about? -- With best regards, Ivan Kurnosov

1 1