- Bird-users - Mailing Lists

Problem with BIRD importing routes from neighbor to master table
by Lyubomir Millanov 22 May '17

22 May '17

Hello, I'm having this problem with BIRD refusing to import a specific route into the routing table. This is strange because this is the session with the neighbor is not new and all of his other prefixes work fine. This is a newly added prefix. What I see: bird> show route all 185.200.156.0/22 185.200.156.0/22 via 185.1.40.2 on eth1 [as41059 2017-05-11] * (100) [AS61437i Type: BGP unicast univ BGP.origin: IGP BGP.as_path: 41313 16154 61437 BGP.next_hop: 185.1.40.2 BGP.med: 100 BGP.local_pref: 70 BGP.community: (0,42) (0,3856) (0,5483) (0,6939) (0,8400) (0,13335) (0,15169) (0,15958) (0,34772) (15169,13200) (16154,10000) (41059,2) (41059,65007) (41313,2) (49401,8877) (49401,49401) BGP.ext_community: (ro, 0, 197216) bird> sh route protocol as16154 all 185.200.156.0/22 via 185.1.40.12 on eth0 [as16154 2017-05-12] * (100) [AS61437i Type: BGP unicast univ BGP.origin: IGP BGP.as_path: 16154 61437 BGP.next_hop: 185.1.40.12 BGP.local_pref: 100 BGP.community: (15169,13300) (16154,10000) (41313,6699) (65000,208) For some reason the prefix is not installed in the master routing table even though it's accepted from the neighbor. We use IN-filtering based on queries with irr tool and those queries are correct - the prefix is correctly announced. >From the above I can see that the neighbor prefix is learned after the first one (same prefix, learned from another path). I've cleared the bgp session with the neighbor without any change of the behavior. Is there a way to clear the route from the master routing table? I'd appreciate if someone can give me advice on the matter. Best regards, Lyubo

1 0

Re: Please help, local interfaces not being recognized as local
by Jerry Scharf 18 May '17

18 May '17

Ondrej, Thanks for the response. That fixed it. jerry On 5/18/17 3:10 AM, Ondrej Zajicek wrote: > On Wed, May 17, 2017 at 06:28:59PM -0700, Jerry Scharf wrote: >> Hi, >> >> So I have narrowed things down but still don't understand what is going on. >> The symptoms are as follows: >> >> I have local interfaces with either simple addresses on net or on vlans. The >> peer is on the local subnet of the interface. Some are ibgp and some are >> ebgp. If I set direct, they all fail and I don't see any listener on port >> 179. As soon as I change them to multihop, everything works fine. > Hi > > You are missing 'protocol device' which is necessary for BIRD to find out > network interfaces. > -- Soundhound Devops "What could possibly go wrong?"

1 0

Please help, local interfaces not being recognized as local
by Jerry Scharf 18 May '17

18 May '17

Hi, So I have narrowed things down but still don't understand what is going on. The symptoms are as follows: I have local interfaces with either simple addresses on net or on vlans. The peer is on the local subnet of the interface. Some are ibgp and some are ebgp. If I set direct, they all fail and I don't see any listener on port 179. As soon as I change them to multihop, everything works fine. Also, I have defined a bunch of local aliases on lo. I then create a static block that lists all the addresses as via "lo". When I show static, they all show as dormant and they are not being redistributed by the bgp speaker. This is on a centos 7 system currently stock configured runing 1.4.5 from epel. I had been doing things with systemd-networkd and alternate kernels, but I did a clean install now. I have also tried 1.6.0 bird/centos7 build with the same issues. If someone can help me debug this, I would really appreciate it. If not, I may have to abandon using bird in favor of another routing daemon. That is certainly not my first choice. thanks, jerry here's my current config (with multihop specified and bfd disabled. There are currently no other files being included.) # this is automatically generated from borderbird4.j2, version 0.3 protocol kernel { # learn; # Learn all alien routes from the kernel persist; # Don't remove routes on bird shutdown scan time 20; # Scan kernel routing table every 20 seconds # import none; # Default is import all export all; # Default is export none # kernel table 5; # Kernel table to synchronize with (default: mai n) } protocol direct { interface "lo", "*"; preference 32000; # make sure bgp routes don't overwrite direct } log "/var/log/birdlog" all; router id 10.100.16.10; protocol static { preference 16000; route 198.49.100.45/32 via "lo"; route 8.25.217.18/32 via "lo"; route 198.49.100.18/32 via "lo"; route 8.25.217.45/32 via "lo"; route 8.6.161.18/32 via "lo"; route 8.5.161.45/32 via "lo"; } function issecondary() { return net ~ [198.49.100.45,8.25.217.18,198.49.100.18,8.25.217.45,8.6.161.18 ,8.5.161.45]; } function isprimary() { return net ~ [0.0.0.4/30]; } function net_outside() { return net ~ [172.18.10.0/24+,8.5.161.0/24+,198.49.100.0/24+]; } function is_internal_local() { return net ~ [10.100.32.10/32,10.100.32.0/20,10.100.16.10/32,10.100.16.0/20] ; } filter inside_export { if ((net = 0.0.0.0/0) || (net ~ 10.0.0.0/8)) then accept; else reject; } filter inside_import { if is_internal_local() then reject; if net ~ [10.0.0.0/8+, 172.16.0.0/12+] then accept; else reject; } filter load_default { if net = 0.0.0.0/0 then accept; else reject; } filter outside_only { if isprimary() then bgp_local_pref=100; else if issecondary() then bgp_local_pref=80; else if net_outside() then bgp_local_pref=25; if net_outside() then accept; else reject; } template bgp tocore { description "bgp to data center as"; local as 65302; multihop; export filter inside_export; import filter inside_import; # bfd on; } protocol bgp inside_netA from tocore { neighbor 10.100.16.3 as 65302; source address 10.100.16.10; } protocol bgp inside_netB from tocore { neighbor 10.100.32.3 as 65302; source address 10.100.32.10; } template bgp toedge { description "bgp to edge/ISP as"; local as 64102; multihop; export filter outside_only; import filter load_default; allow local as; # bfd on; } protocol bgp outside_edge_A from toedge { neighbor 172.18.10.1 as 18741; source address 172.18.10.4; } protocol bgp outside_edge_B from toedge { neighbor 172.18.10.2 as 18741; source address 172.18.10.5; } protocol bfd { interface "eth*" { interval 300 ms; multiplier 3; }; } include "/soundhound/etc/bird/*4.part.cfg"; -- Jerry Scharf, Soundhound DevOps "What could possibly go wrong?"

2 1

Redistribute loopback in OSPF
by Quentin Ritoul 18 May '17

18 May '17

Hello, I'm starting to use Bird, and I try to redistribute loopback in ospf. (So it seems easy...) rt01 : interface ens18 > 10.0.4.5/24 interface loopback > 10.99.99.1/32 (and 127.0.0.1/8) rt02 : interface ens18 > 10.0.4.6/24 interface loopback > 10.99.99.100/32 (and 127.0.0.1/8) Below the configuration of rt01 and rt02 : rt01 : filter loopbacks { if (net = 10.99.99.1/32 ) then accept; else reject; } protocol ospf { import all; export filter loopbacks; area 0 { interface "ens18" { cost 5; type broadcast; hello 5; retransmit 2; wait 10; dead 20; authentication cryptographic; password "toto"; }; }; } rt02 : filter loopbacks { if (net = 10.99.99.100/32 ) then accept; else reject; } protocol ospf { import all; export filter loopbacks; area 0 { interface "ens18" { cost 5; type broadcast; hello 5; retransmit 2; wait 10; dead 20; authentication cryptographic; password "toto"; }; }; } But it doesn't work. Each router don't receive the route 10.99.99.X bird> sh route 0.0.0.0/0 via 10.0.4.1 on ens18 [kernel1 22:01:40] * (10) 10.0.4.0/24 dev ens18 [direct1 22:01:40] * (240) dev ens18 [ospf1 22:01:48] I (150/5) [10.0.4.5] Could you help me please ? Regards, Quentin

3 2

OSPFv3 flooding unknown LSAs
by Dean 17 May '17

17 May '17

Hello, I am using bird 2.0.0 (from the int-new branch) and I noticed that OSPF is flooding unknown LSAs. According to RFC 5340, unknown LSAs shouldn't be flooded if the U-bit isn't set. But the lsa_flooding_allowed function doesn't check the U-bit or whether the LSA has an unknown type. Is that a mistake or did I miss something? Regards, Dean Luga.

1 1

are my posts working?
by Jerry Scharf 17 May '17

17 May '17

Hi, Sorry to spam the list and people, but I have tried to send a couple messages to the list with nothing showing up. jerry -- Jerry Scharf, Soundhound DevOps "What could possibly go wrong?"

2 1

not seeing statics being advertised to bgp neighbors
by Jerry Scharf 16 May '17

16 May '17

Hi, I have the following setup and am not seeing what I expect. I had this working before, so I am sure it is something I have changed in the configurations when automating them. I just can't see what it is. Situation is that that have a bunch of host aliases (/32) defined on the lo interface. For each of these, I create a static in bird so there is a route. Then I filter and forward those via BGP to the edge routers that talk to ISPs. A single machine is presenting a number of addresses in different routing blocks and addresses. here is the list of loopback aliases [root@pnpborder1 bird]# ip addr show lo 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN qlen 1 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 scope host lo valid_lft forever preferred_lft forever inet 198.49.100.18/32 brd 198.49.100.18 scope global lo valid_lft forever preferred_lft forever inet 8.25.217.45/32 brd 8.25.217.45 scope global lo valid_lft forever preferred_lft forever inet 8.6.161.18/32 brd 8.6.161.18 scope global lo valid_lft forever preferred_lft forever inet 8.5.161.45/32 brd 8.5.161.45 scope global lo valid_lft forever preferred_lft forever inet 198.49.100.45/32 brd 198.49.100.45 scope global lo valid_lft forever preferred_lft forever inet 8.25.217.18/32 brd 8.25.217.18 scope global lo valid_lft forever preferred_lft forever inet 4.7.20.4/32 brd 4.7.20.4 scope global lo valid_lft forever preferred_lft forever inet 4.7.20.5/32 brd 4.7.20.5 scope global lo valid_lft forever preferred_lft forever inet 173.1.2.36/32 brd 173.1.2.36 scope global lo valid_lft forever preferred_lft forever inet 173.1.2.37/32 brd 173.1.2.37 scope global lo valid_lft forever preferred_lft forever inet6 dead:beef::1/32 scope global valid_lft forever preferred_lft forever inet6 ::1/128 scope host valid_lft forever preferred_lft forever the routing table doesn't show the routes and the neighbors don't get them (did a tcpdump to make sure it wasn't filtering on the far side.) [root@pnpborder1 bird]# netstat -nr Kernel IP routing table Destination Gateway Genmask Flags MSS Window irtt Iface 0.0.0.0 - 0.0.0.0 ! - - - - 4.7.20.0 0.0.0.0 255.255.255.240 U 0 0 0 edge10Bvl801 4.7.20.0 0.0.0.0 255.255.255.240 U 0 0 0 edge10Avl801 10.100.16.0 0.0.0.0 255.255.240.0 U 0 0 0 eth4 10.100.32.0 0.0.0.0 255.255.240.0 U 0 0 0 eth5 10.201.0.0 0.0.0.0 255.255.240.0 U 0 0 0 eth0 169.254.0.0 0.0.0.0 255.255.0.0 U 0 0 0 eth0 172.18.10.0 0.0.0.0 255.255.255.0 U 0 0 0 edge10Bvl800 172.18.10.0 0.0.0.0 255.255.255.0 U 0 0 0 edge10Avl800 173.1.2.32 0.0.0.0 255.255.255.240 U 0 0 0 edge10Bvl803 173.1.2.32 0.0.0.0 255.255.255.240 U 0 0 0 edge10Avl803 can anyone spot what I did wrong? thanks, jerry config: # this is automatically generated from borderbird4.j2, version 0.3 protocol kernel { # learn; # Learn all alien routes from the kernel persist; # Don't remove routes on bird shutdown scan time 20; # Scan kernel routing table every 20 seconds # import none; # Default is import all export all; # Default is export none # kernel table 5; # Kernel table to synchronize with (default: main) } protocol direct { interface "lo", "*"; preference 32000; # make sure bgp routes don't overwrite direct } log "/var/log/birdlog" all; router id 10.100.16.10; protocol static { preference 16000; route 198.49.100.45/32 via "lo"; route 8.25.217.18/32 via "lo"; route 198.49.100.18/32 via "lo"; route 8.25.217.45/32 via "lo"; route 8.6.161.18/32 via "lo"; route 8.5.161.45/32 via "lo"; } function issecondary() { return net ~ [198.49.100.45,8.25.217.18,198.49.100.18,8.25.217.45,8.6.161.18,8.5.161.45]; } function isprimary() { return net ~ [0.0.0.4/30]; } function net_outside() { return net ~ [172.18.10.0/24+,8.5.161.0/24+,198.49.100.0/24+]; } function is_internal_local() { return net ~ [10.100.32.10/32,10.100.32.0/20,10.100.16.10/32,10.100.16.0/20]; } filter inside_export { if ((net = 0.0.0.0/0) || (net ~ 10.0.0.0/8)) then accept; else reject; } filter inside_import { if is_internal_local() then reject; if net ~ [10.0.0.0/8+, 172.16.0.0/12+] then accept; else reject; } filter load_default { if net = 0.0.0.0/0 then accept; else reject; } filter outside_only { if isprimary() then bgp_local_pref=100; else if issecondary() then bgp_local_pref=80; else if net_outside() then bgp_local_pref=25; if net_outside() then accept; else reject; } template bgp tocore { description "bgp to data center as"; local as 65302; direct; export filter inside_export; import filter inside_import; bfd on; } protocol bgp inside_netA from tocore { neighbor 10.100.16.3 as 65302; source address 10.100.16.10; } protocol bgp inside_netB from tocore { neighbor 10.100.32.3 as 65302; source address 10.100.32.10; } template bgp toedge { description "bgp to edge/ISP as"; local as 18741; # direct; export filter outside_only; import filter load_default; allow local as; bfd on; } protocol bgp outside_edge_A from toedge { neighbor 172.18.10.1 as 18741; source address 172.18.10.4; } protocol bgp outside_edge_B from toedge { neighbor 172.18.10.2 as 18741; source address 172.18.10.5; } protocol bfd { interface "eth*" { interval 300 ms; multiplier 3; }; } include "/soundhound/etc/bird/*4.part.cfg"; -- Jerry Scharf, Soundhound DevOps "What could possibly go wrong?"

1 0

[ERROR]: BIRD client errors: 1: Socket not found
by Ajai Kumar 16 May '17

16 May '17

Dear All, I am using bird-lg downloaded from subnet.ru and it is giving output of all the commands for IPv4. bird.client.php is in one server and route server is in remote server. I have enabled IPv6 on bird.lg.conf.php and route server is running properly for IPv6 too as I am able to see the output after entering into the bird6. I am able to see ping6 report only and for other commands,I am getting appended error [ERROR]: BIRD client errors: 1: Socket not found Can anyone help on this. Thanks in advance for your support. Regards, Ajai Kumar -- Please don't print this email unless you really need to. This will preserve trees on our planet.

2 4

not seeing statics being advertised to bgp neighbors
by Jerry Scharf 16 May '17

16 May '17

Hi, I have the following setup and am not seeing what I expect. I had this working before, so I am sure it is something I have changed in the configurations when automating them. I just can't see what it is. Situation is that that have a bunch of host aliases (/32) defined on the lo interface. For each of these, I create a static in bird so there is a route. Then I filter and forward those via BGP to the edge routers that talk to ISPs. A single machine is presenting a number of addresses in different routing blocks and addresses. I have attached the bird config for this machine. here is the list of loopback aliases [root@pnpborder1 bird]# ip addr show lo 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN qlen 1 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 scope host lo valid_lft forever preferred_lft forever inet 198.49.100.18/32 brd 198.49.100.18 scope global lo valid_lft forever preferred_lft forever inet 8.25.217.45/32 brd 8.25.217.45 scope global lo valid_lft forever preferred_lft forever inet 8.6.161.18/32 brd 8.6.161.18 scope global lo valid_lft forever preferred_lft forever inet 8.5.161.45/32 brd 8.5.161.45 scope global lo valid_lft forever preferred_lft forever inet 198.49.100.45/32 brd 198.49.100.45 scope global lo valid_lft forever preferred_lft forever inet 8.25.217.18/32 brd 8.25.217.18 scope global lo valid_lft forever preferred_lft forever inet 4.7.20.4/32 brd 4.7.20.4 scope global lo valid_lft forever preferred_lft forever inet 4.7.20.5/32 brd 4.7.20.5 scope global lo valid_lft forever preferred_lft forever inet 173.1.2.36/32 brd 173.1.2.36 scope global lo valid_lft forever preferred_lft forever inet 173.1.2.37/32 brd 173.1.2.37 scope global lo valid_lft forever preferred_lft forever inet6 dead:beef::1/32 scope global valid_lft forever preferred_lft forever inet6 ::1/128 scope host valid_lft forever preferred_lft forever the routing table doesn't show the routes and the neighbors don't get them (did a tcpdump to make sure it wasn't filtering on the far side.) [root@pnpborder1 bird]# netstat -nr Kernel IP routing table Destination Gateway Genmask Flags MSS Window irtt Iface 0.0.0.0 - 0.0.0.0 ! - - - - 4.7.20.0 0.0.0.0 255.255.255.240 U 0 0 0 edge10Bvl801 4.7.20.0 0.0.0.0 255.255.255.240 U 0 0 0 edge10Avl801 10.100.16.0 0.0.0.0 255.255.240.0 U 0 0 0 eth4 10.100.32.0 0.0.0.0 255.255.240.0 U 0 0 0 eth5 10.201.0.0 0.0.0.0 255.255.240.0 U 0 0 0 eth0 169.254.0.0 0.0.0.0 255.255.0.0 U 0 0 0 eth0 172.18.10.0 0.0.0.0 255.255.255.0 U 0 0 0 edge10Bvl800 172.18.10.0 0.0.0.0 255.255.255.0 U 0 0 0 edge10Avl800 173.1.2.32 0.0.0.0 255.255.255.240 U 0 0 0 edge10Bvl803 173.1.2.32 0.0.0.0 255.255.255.240 U 0 0 0 edge10Avl803 can anyone spot what I did wrong? thanks, jerry -- Jerry Scharf, Soundhound DevOps "What could possibly go wrong?"

1 1

Bird & VRF on Linux
by Tim Weippert 15 May '17

15 May '17

HI all, currently i'm playing with Bird & VRF on Linux Kernel >4.9. I try to investigate if it is possible to build up an system which use Bird/BGP to populate the VRF/Tables. I tried to run on bird instance and use the described feature of running it in the default context and build up peerings within the sample VRF. As i can configure bird to try to build the Peerings with the correct interfaces/VRFs it seems no packet is send "through" the VRF or the listen socket is not working in this scenario (refused). My question is, is Bird capable to cope with the VRF Implementation in newer Linux Kernels? Has any one tried this? Regards, tim -- Tim Weippert http://weiti.org - weiti(a)weiti.org GPG Fingerprint - E704 7303 6FF0 8393 ADB1 398E 67F2 94AE 5995 7DD8

2 2