June 2022 - Bird-users - Mailing Lists

Debian armel test fail blocking transition to testing
by Jakub Ružička 28 Jun '22

28 Jun '22

Hello, bird-2.0.10 builds fine on most of Debian archs, but on armel, hppa, and riscv64 there is a failing test: obj/filter/trie_test: t_trie_walk: Timeout expired obj/filter/trie_test: t_trie_walk 5.010453063s [FAIL] obj/filter/trie_test 12.572917503s [FAIL] make[1]: *** [Makefile:180: obj/filter/trie_test.ok] Error 1 make[1]: *** Waiting for unfinished jobs.... make[1]: Leaving directory '/<<PKGBUILDDIR>>' dh_auto_test: error: make -j4 test "TESTSUITEFLAGS=-j4 --verbose" VERBOSE=1 returned exit code 2 make: *** [debian/rules:21: build-arch] Error 25 Nice overview including log tails here: https://buildd.debian.org/status/package.php?p=bird2 As these tests are run as part of build, failing tests result in FTBFS and armel FTBFS is blocking transition to testing as seen here: https://tracker.debian.org/pkg/bird2 armel refers to older ARM 32-bit HW - if you have access to the arch, could you please test/fix this? Cheers, Jakub Ružička

4 4

BIRD version 2.0.10
by Ondrej Filip 27 Jun '22

27 Jun '22

Dear users, We just released a second stable version of BIRD this year. Many thanks to my great colleagues and namely Santiago. Here is a list of the main improvements: Version 2.0.10 o BGP performance improvements o BFD: New 'strict bind' option o RPKI: VRF support o Allow use of 240.0.0.0/4 as a private range o BIRD client uses exit status to report errors o Important bugfixes Enjoy! Ondrej

1 0

OSPF & merge external on
by Mikhail Mayorov 26 Jun '22

26 Jun '22

Hello! I would like use ECMP for 0.0.0.0/0. This route announced on 10.10.97.33, 10.10.97.72, 10.10.97.73 as LSA5. Below I have given the output of the command on the router 10.10.97.67. I expected to see a default route through 2 interfaces C3 to D1, but next hops not merge. Maybe I misunderstand the work "merge external on". I expected that the bird would combine routes from different routers for the same destination although OSPF.metric1 or OSPF.metric2. bird>show route for10.10.97.73 table igpcoret4 all Table igpcoret4: 10.10.97.73/32unicast [igpcore4 20:05:45]*E2 (150/310/10000)[10.10.97.73] via 10.112.0.14 on C3 weight 40 via 10.112.0.28 on D1 weight 40 Type:OSPF-E2 univ OSPF.metric1:310 OSPF.metric2:10000 OSPF.tag:0x00000000 OSPF.router_id:10.10.97.73 bird>show route for10.10.97.72 table igpcoret4 all Table igpcoret4: 10.10.97.72/32unicast [igpcore4 20:05:45]*E2 (150/2300/10000)[10.10.97.72] via 10.112.0.14 on C3 Type:OSPF-E2 univ OSPF.metric1:2300 OSPF.metric2:10000 OSPF.tag:0x00000000 OSPF.router_id:10.10.97.72 bird>show route for10.10.97.33 table igpcoret4 all Table igpcoret4: 10.10.97.33/32unicast [igpcore4 20:05:45]*E2 (150/300/10000)[10.10.97.33] via 10.112.0.14 on C3 Type:OSPF-E2 univ OSPF.metric1:300 OSPF.metric2:10000 OSPF.tag:0x00000000 OSPF.router_id:10.10.97.33 bird>show route for0.0.0.0/24table igpcoret4 all Table igpcoret4: 0.0.0.0/0unicast [igpcore4 20:05:45]*E2 (150/300/10000)[10.10.97.33] via 10.112.0.14 on C3 Type:OSPF-E2 univ OSPF.metric1:300 OSPF.metric2:10000 OSPF.tag:0x00000000 OSPF.router_id:10.10.97.33 bird> >showospf state all igpcore4 area 0.0.0.0 router 10.10.97.33 distance 300 router 10.10.97.72metric 2000 router 10.10.97.72metric 2000 router 10.10.97.72metric 2000 router 10.10.97.72metric 2000 router 10.10.97.73metric 10 router 10.10.97.67metric 300 router 10.10.97.66metric 300 router 10.10.97.68metric 65535 stubnet 10.112.0.0/31metric 2000 stubnet 10.112.0.2/31metric 2000 stubnet 10.112.0.4/31metric 2000 stubnet 10.112.0.6/31metric 2000 stubnet 10.112.0.12/31metric 10 stubnet 10.112.0.14/31metric 300 stubnet 10.112.0.18/31metric 300 stubnet 10.112.0.22/31metric 65535 external 0.0.0.0/0metric2 10000 external 10.112.0.0/31metric2 10000 external 10.112.0.2/31metric2 10000 external 10.112.0.4/31metric2 10000 external 10.112.0.6/31metric2 10000 external 10.112.0.12/31metric2 10000 external 10.112.0.14/31metric2 10000 external 10.112.0.18/31metric2 10000 external 10.112.0.22/31metric2 10000 external 87.245.238.204/31metric2 10000 external 10.10.8.0/21metric2 10000 external 10.10.97.33/32metric2 10000 router 10.10.97.66 distance 10 router 10.10.97.73metric 300 router 10.10.97.33metric 300 router 10.10.97.69metric 65535 router 10.10.97.67metric 10 stubnet 10.112.0.16/31metric 300 stubnet 10.112.0.18/31metric 300 stubnet 10.112.0.24/31metric 65535 stubnet 10.112.0.28/31metric 10 external 10.112.0.16/31metric2 10000 external 10.112.0.18/31metric2 10000 external 10.112.0.24/31metric2 10000 external 10.112.0.28/31metric2 10000 external 10.10.0.0/21metric2 10000 external 10.10.97.66/32metric2 10000 router 10.10.97.67 distance 0 router 10.10.97.33metric 300 router 10.10.97.66metric 10 router 10.10.97.69metric 65535 stubnet 10.112.0.14/31metric 300 stubnet 10.112.0.28/31metric 10 stubnet 10.112.0.26/31metric 65535 external 10.112.0.14/31metric2 10000 external 10.112.0.26/31metric2 10000 external 10.112.0.28/31metric2 10000 external 10.10.0.0/21metric2 10000 external 10.10.97.67/32metric2 10000 router 10.10.97.68 distance 65835 router 10.10.97.73metric 65535 router 10.10.97.33metric 65535 stubnet 10.112.0.20/31metric 65535 stubnet 10.112.0.22/31metric 65535 external 10.112.0.20/31metric2 10000 external 10.112.0.22/31metric2 10000 external 10.10.97.68/32metric2 10000 router 10.10.97.69 distance 65535 router 10.10.97.66metric 65535 router 10.10.97.67metric 65535 stubnet 10.112.0.24/31metric 65535 stubnet 10.112.0.26/31metric 65535 external 10.112.0.24/31metric2 10000 external 10.112.0.26/31metric2 10000 external 10.10.97.69/32metric2 10000 router 10.10.97.72 distance 2300 router 10.10.97.33metric 2000 router 10.10.97.33metric 2000 router 10.10.97.33metric 2000 router 10.10.97.33metric 2000 router 10.10.97.73metric 2000 router 10.10.97.73metric 2000 stubnet 10.112.0.0/31metric 2000 stubnet 10.112.0.2/31metric 2000 stubnet 10.112.0.4/31metric 2000 stubnet 10.112.0.6/31metric 2000 stubnet 10.112.0.8/31metric 2000 stubnet 10.112.0.10/31metric 2000 external 0.0.0.0/0metric2 10000 external 10.112.0.0/31metric2 10000 external 10.112.0.2/31metric2 10000 external 10.112.0.4/31metric2 10000 external 10.112.0.6/31metric2 10000 external 10.112.0.8/31metric2 10000 external 10.112.0.10/31metric2 10000 external 10.10.97.72/32metric2 10000 external 100.105.99.0/30metric2 10000 router 10.10.97.73 distance 310 router 10.10.97.72metric 2000 router 10.10.97.72metric 2000 router 10.10.97.33metric 10 router 10.10.97.66metric 300 router 10.10.97.68metric 65535 stubnet 10.112.0.8/31metric 2000 stubnet 10.112.0.10/31metric 2000 stubnet 10.112.0.12/31metric 10 stubnet 10.112.0.16/31metric 300 stubnet 10.112.0.20/31metric 65535 external 0.0.0.0/0metric2 10000 external 10.112.0.8/31metric2 10000 external 10.112.0.10/31metric2 10000 external 10.112.0.12/31metric2 10000 external 10.112.0.16/31metric2 10000 external 10.112.0.20/31metric2 10000 external 10.10.8.0/21metric2 10000 external 10.10.97.73/32metric2 10000 external 188.170.160.52/30metric2 10000 -- Regards, Mikhail V. Majorov

2 2

routes, "strange next-hop" and p2p interfaces
by Eugene M. Zheganin 25 Jun '22

25 Jun '22

Hello, I have a peculiar VPN client that I want to redistributes the routes from. Technically it creates a tun0 interface and attaches all the routes it's connected to as it's own on-interface routes pointing to it's own IP (yeah, I know it's la me, but seems like it's written by Windows 98 addicts, so it's still pretty normal for them). This confuses bird and it 's complaining with usual ===Cut=== Jun 7 15:10:43 vipnet bird[28752]: KRT: Received route 1.2.3.4/26 with strange next-hop 5.6.7.8 ===Cut=== However: ===Cut=== [root@host ~]# ip addr show tun0 4: tun0: <POINTOPOINT,MULTICAST,NOARP,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UNKNOWN group default qlen 500 link/none inet 5.6.7.8/32 scope global tun0 valid_lft forever preferred_lft forever inet6 fe80::6d6d:315d:3deb:6ae0/64 scope link stable-privacy valid_lft forever preferred_lft forever ===Cut=== So far I'm reexporting these routes as static, but is there any way to do thsi automatically ? Thanks.

2 1

Scaling BFD support
by Douglas Fischer 24 Jun '22

24 Jun '22

The question raised by colleague Irene reminded me of a topic that may or may not be the focus of BIRD's development. I imagine that the biggest supporters of SMP/Multi-Core/Thread-Safe evolution on BIRD are Operators of Route-Servers of large IXPs, and operators of large-scale Route-Reflectors. Although BFD has its greatest use in the transport network and Underlay, it is increasingly common to see the use of BFD in BGP Internet. I'm personally overly excited about what BIRD version 3 is demonstrating in terms of vertical scalability. But I keep imagining that, even having scalability in the BGP engine, it is almost prohibitive to use BFD in a scenario with a thousand BGP Peers. Is there any view from the IBRD development team for this matter? Or even... Is there any open project focused on BFD that can address this? And going a little further... In a quick Brainstorming... What would be the techniques that could help a Route-Server or Route-Reflector with a thousand BGP Peers to also support a thousand BFD sessions effectively and efficiently? Perhaps resort to some hardware-based off-loading method? Would eBPF help this in any way? -- Douglas Fernando Fischer Engº de Controle e Automação

6 8

ECMP and merge routes from different routers/protocols for the same destination
by Mikhail Mayorov 24 Jun '22

24 Jun '22

Hello! I'm trying to increase the efficiency of ECMP on links between cities without static routes. On the same physical network (routers and links) I'm trying to run multiple independent OSPF protocols area 0 with different IDs (instance id). Example config file: protocol ospf v2 igpcore4 { ipv4 { table igpcoret4; import filter { ospf_metric1 = 100; accept; }; export all; }; ecmp on limit 12; merge external on; area 0.0.0.0 { interface "A2.1" { type pointopoint; bfd on; ecmp weight 10; cost 1000; }; interface "A2.2" { type pointopoint; bfd on; ecmp weight 10; cost 1000; }; interface "A3.1" { type pointopoint; bfd on; ecmp weight 10; cost 1000; }; interface "A3.2" { type pointopoint; bfd on; ecmp weight 10; cost 1000; }; interface "A1.1" { type pointopoint; bfd on; ecmp weight 10; cost 500; }; interface "A1.2" { type pointopoint; bfd on; ecmp weight 10; cost 500; }; }; }; protocol ospf v2 igpcore4_1 { ipv4 { table igpcoret4; import filter { ospf_metric1 = 100; accept; }; export all; }; ecmp on limit 12; merge external on; instance id 1; area 0.0.0.0 { interface "A2.1" { type pointopoint; bfd on; ecmp weight 10; cost 500; }; interface "A2.2" { type pointopoint; bfd on; ecmp weight 10; cost 500; }; interface "A3.1" { type pointopoint; bfd on; ecmp weight 10; cost 500; }; interface "A3.2" { type pointopoint; bfd on; ecmp weight 10; cost 500; }; interface "A1.1" { type pointopoint; bfd on; ecmp weight 10; cost 1000; }; interface "A1.2" { type pointopoint; bfd on; ecmp weight 10; cost 1000; }; }; }; I got this results area 0 instance 0 in protocol igpcore4 and area 0 instance 1 in protocol igpcore4_1 BIRD 2.0.9 ready. bird> show route for 10.10.97.33/32 all Table igpcoret4: 10.10.97.33/32 unicast [igpcore4_1 11:58:31] * E2 (150/100/10000) [10.10.97.33] via 10.112.0.1 on A2.1 weight 10 via 10.112.0.3 on A2.2 weight 10 via 10.112.0.5 on A3.1 weight 10 via 10.112.0.7 on A3.2 weight 10 Type: OSPF-E2 univ OSPF.metric1: 100 OSPF.metric2: 10000 OSPF.tag: 0x00000000 OSPF.router_id: 10.10.97.33 unicast [igpcore4 11:58:31] E2 (150/100/10000) [10.10.97.33] via 10.112.0.1 on A2.1 weight 10 via 10.112.0.3 on A2.2 weight 10 via 10.112.0.5 on A3.1 weight 10 via 10.112.0.7 on A3.2 weight 10 via 10.112.0.9 on A1.1 weight 10 via 10.112.0.11 on A1.2 weight 10 Type: OSPF-E2 univ OSPF.metric1: 100 OSPF.metric2: 10000 OSPF.tag: 0x00000000 OSPF.router_id: 10.10.97.33 How to megre these routes and get a single route with 6 next hops interfaces.... Maybe via pipe or other variants? -- Regards, Mikhail V. Majorov

1 0

BGP MSG Updates counters - Sent and Received per peer
by Douglas Fischer 21 Jun '22

21 Jun '22

I know it sounds stupid was it here but I'm feeling a bit dumb and impatient with this. On and environment with several(1000+) peers, I need to collect on a very simples and with less performance impact of the following information each 5-to-5 ou 1-to-1 minute: - Peer status - Number of Routes Received by peer - Number of Routes Sent by peer - BGP Msg Updates Received by Peer - BGP Msg Sent Sent by peer And I don't know what is the best way to get those informations. I saw something about birdwatcher. That sounds good, but I heard some about a bit os instability. I saw bird_exporter, that maybe can be a good tool. And the other "ugly" idea is to use Zabbix agent that will collect the routes via birdc and parse it. Any news about that? Any update.

1 0

Memory leak after a burst of route changes (bird v2.0.8)
by bird bird 13 Jun '22

13 Jun '22

Dear experts, We have a few routers running bird with BGP with route-reflector (full mesh) which has approximately 2K routes. Recently, the RAM usage of bird daemon increased to a few GBs, while normally the RAM usage is only a few MBs. The output of "show memory" showed using only a few KBs of memory, so it looks like some leak happened in the daemon. I wasn't able to find the cause of the event which led to the memory increase. The only lead I had is seeing a large amount of route changes to the routing table in short time. The output of "birdc show proto all all": global_routing_0_kernel Kernel global_routing_0 up 2022-05-01 Export updates: 748274310 0 0 --- 748274310 After the event stopped, the RAM usage of bird did not reduce for a few days until restarting bird daemon. Unfortunately I don't have additional logs or outputs from the time of the issue. However I was able to reproduce a similar scenario by adding a large amount of BGP routes at one burst. The route changes caused the RAM usage to increase to 500MB, and after removing the routes the RAM usage stayed 500MB and didn't reduce automatically until restart of bird daemon. I see a connection between the reproduction and the event that happened in our environment, therefore I hope that it can be enough information to investigate the memory leak. Any assistance will be highly appreciated. Many thanks!

1 0

Is there a way to merge communities from the same prefixes?
by Valery Lutoshkin 07 Jun '22

07 Jun '22

Hi all! I use bird2 (2.0.9) as RS for BGP prefix aggregation - several peers send me their prefixes, I aggregate them and send the result to the system. But there is an obvious issue - when peers send me the same prefix, BGP selects the best one and sends only that one. In this case, if prefixes had different communities, the result loses all but one. For example: peer 1: 1.1.1.1/32 1:100 peer 2: 1.1.1.1/32 2:200 peer 3: 1.1.1.1/32 3:300 If the last one was chosen, the result is 1.1.1.1/32 3:300 If there is any way to collect all communities and attach them to the chosen prefix (like this: 1.1.1.1/32 1:100,2:200,3:300), would anyone please tell me how to do that? Thank you, Valery

4 5

Feature request: CARP awareness in BIRD?
by Thomas Steen Rasmussen 06 Jun '22

06 Jun '22

Hello :) I am using BIRD on FreeBSD and it is working okay, apart from missing CARP awareness. I am wondering if anyone knows if work has been done in this area already? As you may know, CARP is a patent-free VRRP alternative originally from OpenBSD. A network interface can have one or more VHIDs, each VHID can be either MASTER or BACKUP. The IPs tied to a VHID are only usable when the VHID is MASTER. Currently "birdc show interfaces" shows no difference between two nodes where one is CARP MASTER and the other is CARP BACKUP. The consequence of this is that BIRD thinks it has a working local direct route which it then tries to use, but the OS refuses, since it understands that VHID is CARP BACKUP. Ideally BIRD would understand CARP semantics so when it has a DIRECT protocol it would only consume the routes from CARP MASTER VHIDs, and consider the CARP BACKUP VHID IPs as unusable. Please consider this a feature request. I am interested getting some BIRD experts opinion on 1) is this a feature you would be willing to include in BIRD and 2) how big a job is it? I would be willing to pay to have it developed, although I am just myself so I don't have unlimited funds. Thanks in advance, and thank you for BIRD; even without CARP support it is already awesome. Best regards, Thomas Steen Rasmussen

7 8