June 2026 - Bird-users - Mailing Lists

Re: Releases on bird.network.cz no longer updated
by babydr DBA James W. Laferriere 13 Jun '26

13 Jun '26

Hello Ondrej , On Sat, 13 Jun 2026, Ondrej Filip wrote: > On 11/06/2026 09:33, Leon Grasmeijer wrote: >> Hi there, >> >> Are https://bird.network.cz/?download and ftp://bird.network.cz/pub/bird/ >> no longer being updated? >> I see there are some versions missing that have been announced and are >> present on https://bird.nic.cz/get-bird/. > > Hi! > Yes, we are migrating everything to bird.nic.cz. Do you still see a need for > FTP? > > Ondrej While I admint to have not looked very hard . rsync might be a decent addition ! ... Tia , JimL -- +---------------------------------------------------------------------+ | James W. Laferriere | System Techniques | Give me VMS | | Network & System Engineer | 3237 Holden Road | Give me Linux | | jiml(a)system-techniques.com | Fairbanks, AK. 99709 | only on AXP | +---------------------------------------------------------------------+

1 0

Re: CentOS 7 repo not work
by Robert Scheck 13 Jun '26

13 Jun '26

On Sat, 13 Jun 2026, zx--- via Bird-users wrote: > Some of my servers are still using CentOS 7 and I'm following https://bird.nic.cz/get-bird/ to install bird, but it looks like CentOS 7 repo is broken? CentOS Linux 7 is EOL since 2024-06-30, thus nearly for two years. And personally, I highly appreciate that this repository is no longer working (sorry) as it really urges admins to get rid of their legacy systems. https://phoenixnap.com/kb/migrate-centos-to-rocky-linux and others provide steps how to inline upgrade existing CentOS Linux 7 systems, if it helps. Regards, Robert

1 0

None
by None 13 Jun '26

13 Jun '26

not be turned off in iBGP configurations, but this is also true on=20 othe platforms (cisco, juniper, ...). On the other hand having gateway direct (and thus multihop turned off) = might=20 be needed when using with sorted tables. > >=20 > > Removing 'gateway direct' fixes the error (but obviously not the > > behaviour). > >=20 > > This is running in an network container if that's of any interest. > >=20 > > Any ideas? > >=20 > > (apologies for a potential dupe - I sent this to what I think must = have > > been an old list email address) --=20 SP5474-RIPE Sergey Popovich

1 0

None
by None 13 Jun '26

13 Jun '26

bgp_community.add( [] ) bgp_ext_community.add( [] ) should treat arguments like (e)clist. It would be more logical. > But it should report error during filter: > bird: filters, line 169: Can't add set It does not reports anything. Moreover, if change [] brackets with () brackets: cmn = ( (ro,MyASN,10856), (ro,MyASN,2) ); ... there is no syntax error, but it should be? -- Alexander Shikov Technical Staff, Digital Telecom IX Tel.: +380 44 201 14 07 http://dtel-ix.net/

1 0

None
by None 13 Jun '26

13 Jun '26

ecurity=20 reasons): OSPF: Bad packet from 10.1.2.3 - authentication failed As the LOG messages increases, we try to debug this and enable=20 debug protocols { packets } so we see this as the case of the LOG messages: 2014-03-07 10:59:20 <TRACE> aclctrans: OSPF_auth: lower sequence number= (rcv=20 13229786, old 13229787) 2014-03-07 10:59:20 <ERR> OSPF: Bad packet from 10.1.2.3 - authenticati= on=20 failed 2014-03-07 10:59:20 <TRACE> aclctrans: OSPF_auth: lower sequence number= (rcv=20 13229786, old 13229787) 2014-03-07 10:59:20 <ERR> OSPF: Bad packet from 10.1.2.3 - authenticati= on=20 failed 2014-03-07 10:59:20 <TRACE> aclctrans: OSPF_auth: lower sequence number= (rcv=20 13229786, old 13229787) 2014-03-07 10:59:20 <ERR> OSPF: Bad packet from 10.1.2.3 - authenticati= on=20 failed 2014-03-07 10:59:20 <TRACE> aclctrans: OSPF_auth: lower sequence number= (rcv=20 13229786, old 13229787) 2014-03-07 10:59:20 <ERR> OSPF: Bad packet from 10.1.2.3 - authenticati= on=20 failed 2014-03-07 10:59:20 <TRACE> aclctrans: OSPF_auth: lower sequence number= (rcv=20 13229786, old 13229787) 2014-03-07 10:59:20 <ERR> OSPF: Bad packet from 10.1.2.3 - authenticati= on=20 failed 2014-03-07 10:59:20 <TRACE> aclctrans: OSPF_auth: lower sequence number= (rcv=20 13229786, old 13229787) 2014-03-07 10:59:20 <ERR> OSPF: Bad packet from 10.1.2.3 - authenticati= on=20 failed We do not have any clue about where the decremented sequence number cam= e from. As this is an internal network and we have an helo interval of 5 second= we can=20 not imagin that this could be an "late" or "old" paket. At the Moment we use mainly bird 1.3.11 with some exceptions where we h= ave=20 bird 1.4.0. The problem seems not only bird generated as we also have f= rom IPs=20 which are router in our network. --=20 Mit freundlichen Gr=FC=DFen --------------------------------- Daniel Wendler

1 0

None
by None 13 Jun '26

13 Jun '26

as routing daemon using single socket 4 core Intel Xeon X3470 (2.93 Ghz= ), with 4GB RAM and 4 multiqueue NIC (like Intel 82576, 82580, ...) routes= 1,5 - 1,8 Gbit/sec at PPS ~300 000 with ~ 98% of CPU time spent at the lowest= frequency (1.2 Ghz). Software part is custom Linux distribution based o= n Debian GNU/Linux 7.x (wheezy), and custom Linux kernel based on 3.2 sta= ble=20 branch, build for x86_64 architecture. So Intel Xeon X3470 is much likely very powerful for such setups (pure forwarding with no traffic policy). On other deployments I have similar results: CPU: Intel Xeon X3470, 2.93 RAM: 2GB-6GB Task: Dual stack IPv4/IPv6, shaping with HTB (U32 hash classifier), 802.1ad (QinQ) to customer end (this turns off hardware offloading task= s on NICs), uRPF (Reverse Path filtering from customer side to prevent IP= spooofing) for both IPv4/IPv6, few iptables rules (connection tracking = is turned off) and aggressive ipset usage (billing hooks, DDoS, etc). Traffic (In:Out): 2Gbit/sec:1.5Gbit/sec, PPS <450 000. Peak load: <80% on all cores. Also having threads (Hyper-Threading) has no good effect on traffic for= warding workloads, as CPU cache is shared between physical cores, causing more = cache misses and thus lower performance. Amount of RAM seems not wery important, hardware with starting from 2GB= RAM=20 sizes is enough to store routing cache information. >=20 > Thanks to the dev team, and kind regards. >=20 > Tigran Zakoyan. --=20 SP5474-RIPE Sergey Popovich

1 0

None
by None 13 Jun '26

13 Jun '26

A iperf seemed to pick a single eth0 or eth1 but not both interfaces and use it to send out traffic on all 4 difference processes. Ideally how would one allow a single application (in this case iperf) with multiple threads/processes to send 2Gbit worth of traffic? Ideally to a logical interface and it's hash automatically to two difference eth0 and eth1? I've looked at setting (krt_prefsrc) to source the address as the dummy0 interface on each host. Would that be the answer? thanks, -bn 0216331C --047d7b6d7d5a9e0f7a04dc1709ba Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable <div dir=3D"ltr">I'm trying out OSPF ECMP on BIRD (with ECMP set and en= abled on kernel) and on hosts A and host B are configured as below and conn= ected via a switch, the dummy0 interface is used as the loopback address fo= r two physical nics.<div> <br></div><div style>host A</div><div style>eth1 - /30 - OSPF</div><div sty= le>eth0 - /30 - OSPF</div><div style>dummy0 - /32</div><div style><br></div= ><div style><div>host B</div><div>eth1 - /30 - OSPF</div><div>eth0 - /30 - = OSPF</div> <div>dummy0 - /32</div></div><div><br></div><div>From host A I do "ipe= rf -c x.x.x.x -P 4" to host B's dummy0 interface, host A iperf see= med to pick a single eth0 or eth1 but not both interfaces and use it to sen= d out traffic on all 4 difference processes. Ideally how would one allow a = single application (in this case iperf) with multiple threads/processes to = send 2Gbit worth of traffic? Ideally to a logical interface and it's ha= sh automatically to two difference eth0 and eth1?</div> <div><br></div><div>I've looked at setting (<span style=3D"line-height:= 19px;text-align:justify">krt_prefsrc) to source the address as the dummy0 i= nterface on each host. Would that be the answer?</span></div><div><br><div> thanks,</div><div><div>-bn<br>0216331C</div> </div></div></div> --047d7b6d7d5a9e0f7a04dc1709ba--

1 0

None
by None 13 Jun '26

13 Jun '26

How to prevent unneeded routing update to kernel?

1 0

None
by None 13 Jun '26

13 Jun '26

destination host is administratively prohibited. Have u guys encountered this issue? All other BGP with other peering are working ok. Below is the log from ASR.. Logs from ASR ------------- RP/0/RP0/CPU0:Mar 29 07:48:19.078 UTC: bgp[1044]: %ROUTING-BGP-5-ADJCHANGE : neighbor 119.27.63.253 Up (VRF: default) RP/0/RP0/CPU0:Mar 29 07:49:53.328 UTC: tcp[355]: %IP-TCP_NSR-5-DISABLED : 119.27.63.38:28514 <-> 119.27.63.253:179:: NSR disabled for TCP connection because Retransmission threshold exceeded RP/0/RP0/CPU0:Mar 29 07:49:53.343 UTC: bgp[1044]: %ROUTING-BGP-3-NBR_NSR_DISABLED : NSR disabled on neighbor 119.27.63.253 due to TCP retransmissions RP/0/RP1/CPU0:Mar 29 07:49:53.357 UTC: bgp[1044]: %ROUTING-BGP-5-NBR_NSR_DISABLED_STANDBY : NSR disabled on neighbor 119.27.63.253 on standby due to Peer closing down the session (VRF: default) Thanks, Jimmy

1 0

None
by None 13 Jun '26

13 Jun '26

I want to announce 2001:db8:2177::/48, which also needs to be in the kernel routing table as a blackhole, and also 2001:db8:30a9:ffff::/64, which is a subnet of 2001:db8:30a9::/48. Static protocol is used to also blackhole the 2001:db8:30a9::/48. Traffic for this 2001:db8:30a9::/48 besides the /64 we announce at this point will flow into AS64515 via another router that is connected to AS65517 and which is announcing the whole 2001:db8:30a9::/48 Anyway... router id 1.3.3.7; debug protocols { states, routes, filters, interfaces } protocol kernel { persist; scan time 60; import none; export all; } protocol device { scan time 60; } protocol static { import all; route 2001:db8:30a9::/48 via "dummy0"; route 2001:db8:2177::/48 via "dummy0"; } template bgp template_bgp { local as 64515; import all; export all; route limit 1000; } template pipe template_pipe { table master; mode transparent; } table T65517; filter bgp_pipe_T65517_to_master { if net = ::/0 then accept; reject; } filter bgp_pipe_master_to_T65517 { if net = 2001:db8:2177::/48 then accept; reject; } protocol pipe P65517 from template_pipe { peer table T65517; import filter bgp_pipe_T65517_to_master; export filter bgp_pipe_master_to_T65517; } protocol static static_bgp_T65517 { table T65517; import all; route 2001:db8:30a9:ffff::/64 via "dummy0"; } protocol bgp bgp_example_peer from template_bgp { table T65517; source address 2001:db8:2000:1::202; neighbor 2001:db8:2000:1::201 as 65517; } Note that the /64 route is directy injected into T65517, because it only has to be used inside the BGP announcement. I don't know if this is the best way, as I'm obviously still learning to find out, but it works (tm). Now when the context is clear... let's start bird with this config file... bird6: kernel1: Initializing bird6: device1: Initializing bird6: static1: Initializing bird6: P65517: Initializing bird6: static_bgp_T65517: Initializing bird6: bgp_example_peer: Initializing bird6: device1: Starting bird6: device1: Connected to table master bird6: device1: State changed to feed bird6: kernel1: Starting bird6: kernel1: Connected to table master bird6: kernel1: State changed to feed bird6: static1: Starting bird6: static1: Connected to table master bird6: static1: State changed to feed bird6: P65517: Starting bird6: P65517: Connected to table master bird6: P65517: Connected to table T65517 bird6: P65517: State changed to feed bird6: static_bgp_T65517: Starting bird6: static_bgp_T65517: Connected to table T65517 bird6: static_bgp_T65517: State changed to feed bird6: bgp_example_peer: Starting bird6: Started bird6: device1: State changed to up bird6: kernel1: State changed to up bird6: static1 < interface lo goes up bird6: static1 < interface eth0 goes up bird6: static1 < interface eth1 goes up bird6: static1 < interface eth2 goes up bird6: static1 < interface eth3 goes up bird6: static1 < interface dummy0 created bird6: static1: State changed to up bird6: P65517: State changed to up bird6: static_bgp_T65517 < interface lo goes up bird6: static_bgp_T65517 < interface eth0 goes up bird6: static_bgp_T65517 < interface eth1 goes up bird6: static_bgp_T65517 < interface eth2 goes up bird6: static_bgp_T65517 < interface eth3 goes up bird6: static_bgp_T65517 < interface dummy0 created bird6: static_bgp_T65517: State changed to up bird6: bgp_example_peer: Connected to table T65517 bird6: bgp_example_peer: State changed to feed bird6: bgp_example_peer: State changed to up bird6: bgp_example_peer > added [best] ::/0 via 2001:db8:2000:1::201 on eth0 bird6: P65517 < added ::/0 via 2001:db8:2000:1::201 on eth0 bird6: P65517 > added [best] ::/0 via 2001:db8:2000:1::201 on eth0 bird6: P65517 < rejected by protocol ::/0 via 2001:db8:2000:1::201 on eth0 bird6: kernel1 < added ::/0 via 2001:db8:2000:1::201 on eth0 bird6: bgp_example_peer < rejected by protocol ::/0 via 2001:db8:2000:1::201 on eth0 BGP session is established, we get the predicted ::/0 route from the remote AS. bird> show route table T65517 ::/0 via 2001:db8:2000:1::201 on eth0 [bgp_example_peer 00:33] * (100) [AS3265i] bird> show route table master ::/0 via 2001:db8:2000:1::201 on eth0 [bgp_example_peer 00:41] * (100) [AS3265i] But!... the static routes are not in the system. I don't see the /48's and /64 anywhere... Let's gently reload the same (!) configuration file again... bird> configure soft Reading configuration from /etc/bird6.conf Reconfigured bird6: Reconfiguring bird6: kernel1: Reconfigured bird6: device1: Reconfigured bird6: static1 > added [best] 2001:db8:30a9::/48 dev dummy0 bird6: P65517 < filtered out 2001:db8:30a9::/48 dev dummy0 bird6: kernel1 < added 2001:db8:30a9::/48 dev dummy0 bird6: static1 > added [best] 2001:db8:2177::/48 dev dummy0 bird6: P65517 < added 2001:db8:2177::/48 dev dummy0 bird6: P65517 > added [best] 2001:db8:2177::/48 dev dummy0 bird6: P65517 < rejected by protocol 2001:db8:2177::/48 dev dummy0 bird6: bgp_example_peer < added 2001:db8:2177::/48 dev dummy0 bird6: kernel1 < added 2001:db8:2177::/48 dev dummy0 bird6: static1: Reconfigured bird6: P65517: Reconfigured bird6: static_bgp_T65517 > added [best] 2001:db8:30a9:ffff::/64 dev dummy0 bird6: P65517 < filtered out 2001:db8:30a9:ffff::/64 dev dummy0 bird6: bgp_example_peer < added 2001:db8:30a9:ffff::/64 dev dummy0 bird6: static_bgp_T65517: Reconfigured bird6: bgp_example_peer: Reconfigured bird6: Reconfigured bird> show route table T65517 ::/0 via 2001:db8:2000:1::201 on eth0 [bgp_example_peer 00:41] * (100) [AS3265i] 2001:db8:30a9:ffff::/64 dev dummy0 [static_bgp_T65517 00:48] * (200) 2001:db8:2177::/48 dev dummy0 [static1 00:48] * (200) bird> show route table master ::/0 via 2001:db8:2000:1::201 on eth0 [bgp_example_peer 00:41] * (100) [AS3265i] 2001:db8:2177::/48 dev dummy0 [static1 00:48] * (200) 2001:db8:30a9::/48 dev dummy0 [static1 00:48] * (200) And tadaa.wav: # ip -6 r list proto bird 2001:db8:2177::/48 dev dummy0 metric 1024 mtu 1500 advmss 1440 hoplimit 4294967295 2001:db8:30a9::/48 dev dummy0 metric 1024 mtu 1500 advmss 1440 hoplimit 4294967295 default via 2001:db8:2000:1::201 dev eth0 metric 1024 mtu 1500 advmss 1440 hoplimit 4294967295 My question is: what's going on here? Why are the static routes not fed into bird at startup, but only after I reload the unchanged configuration? Have I overlooked a configuration detail, or is there something else 'fishy' going on? Thanks, -- Hans van Kranenburg - System / Network Engineer T +31 (0)10 2760434 | hans.van.kranenburg(a)mendix.com | www.mendix.com

1 0