May 2017 - Bird-users - Mailing Lists

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

default-originate from quagga
by Milen Bogunski 11 May '17

11 May '17

Hello, Looks like bird handles "neighbor x.x.x.x default-originate" (0.0.0.0/0) from quagga in different way than "network 0.0.0.0/0" from quagga. I'm currently trying to change local-preference of default route, If I originate with "network 0.0.0.0/0" from quagga, everything is fine. using this filter: filter in_gw2 { if ( net_martian() ) then { reject; } else { bgp_local_pref=110; accept; } } if change local preference from 110 to 130, configure and reload all - local preference is changed but if I receive 0.0.0.0/0 from quagga with neighbor x.x.x.x. default-originate, you cannot manipulate local-preference of 0.0.0.0/0 route without restart of BGP protocol. Any suggestion ? Thanks in advance Milen Bogunski

1 0

Re: Import directly connected routes
by Damien Clabaut 10 May '17

10 May '17

Thank you for your answer. This is really useful and helped me solve a problem that seemed impossible in Quagga. Thanks :) On 09/05/17 08:47 AM, Ondrej Zajicek wrote: > On Tue, May 09, 2017 at 08:28:26AM -0400, Damien Clabaut wrote: >> Hello, >> >> Thank you for your answer. >> >> Are there some reserved variable names in function arguments ? > Yes, unfortunately each keyword is reserved and cannot be used as symbol > (variable, argument, protocol, table, ... name). 's' is keyword used for > seconds in some time interval options. > -- Damien Clabaut R&D vRouter ovh.qc.ca

1 0

Shell script to monitor BGP sessions thought NRPE
by Alarig Le Lay 10 May '17

10 May '17

Hi, I didn’t find any script to be executed on the router to monitor BGP sessions on https://gitlab.labs.nic.cz/labs/bird/wikis/Related so I wrote my own: https://www.swordarmor.fr/monitoring-des-sessions-bgp-de-bird-via-nrpe.html The explanations are in french, mainly because I’m french and lazy, but the comments inside the script are in english, so feel free to share an modify it :) -- alarig

1 0

Re: Unexpected behavior of static routes
by Alexander Demenshin 10 May '17

10 May '17

On 2017-05-10 03:01, Ondrej Zajicek wrote: > It is expected behavior. It is not optimal, but it is how it works. Is it limitation by design or just "not implemented"? > If you have a recursive route that is resolved to a device route, > then the original gateway is kept. Yes, indeed, this works (to my surprise), though it starts to produce "Received route with strange next-hop" errors on every scan (harmless, but too noisy and incorrect - there is nothing "strange" with that next-hop). It does not look clean though - while it works, it looks not "natural" in configuration ("recursive" instead of "via"). Would it not be an option to implement "onlink" static routes, i.e. including both device and gateway addresses, or there are arguments against this? > If you do not want to assign a real address to the iface, > then you could use ptp address like: This is even less clean, to be honest. --- With best regards, Alexander.

2 1

Unexpected behavior of static routes
by Alexander Demenshin 10 May '17

10 May '17

Hi, I have something like this in my config (bird 1.6.3 @linux): protocol static { route 10.1.0.0/24 via "eth1"; route 10.2.0.0/24 via 10.1.0.1; } Route to 10.1.0.0/24 is installed as expected, but, route to 10.2.0.0/24 is not showing up anywhere, while it seems logical that once its covering network is routed it should be installed too. Even if I add route to 10.1.0.0/24 manually (outside of bird) and it is visible to bird, 10.2/24 is still ignored, it is only installed if (and only if) I add an IP within 10.2.0.0/24 range to the interface. Is it a bug or intentional behavior? Is there any way to workaround this? Yes, I know that I can use recursive routes but this is not an option in my setup - the gateway has to be exact, routing through interface will not work. The aim is to maintain a list of routes which will be activated if: - device is up - gateway becomes available through any other active route (like in my example) Thank you! -- With best regards, Alexander.

3 4

Re: Import directly connected routes
by Damien Clabaut 09 May '17

09 May '17

Hello, Thank you for your answer. Are there some reserved variable names in function arguments ? If I do (as per example in filter.conf in git repo): function onef(string s) { return 1; } I get syntax error in the first line of the function. However, function onef(string a) { return 1; } works. Regards, On 09/05/17 07:32 AM, Ondrej Zajicek wrote: > On Mon, May 08, 2017 at 04:43:32PM -0400, Damien Clabaut wrote: >> Hello again, >> >> Is there any way to declare a list of interfaces in a string set or >> something ? > No, string sets are not implemented. Although it is something that makes > sense to add. > > You could define functions that match a set of string: > > function xxx(string s) > { > return s = "dev1" || s = "dev2" || s = "dev3"; > } > > Then you could use such functions in filters of direct and pipe protocol. > > Instead of: > > protocol direct direct_numa0 { > table numa0; > interface "eth4.800", "eth4.801"; > } > > You could use: > > protocol direct direct_numa0 { > table numa0; > import where xxx(ifname); > } > > -- Damien Clabaut R&D vRouter ovh.qc.ca

2 1