January 2013 - Bird-users

IS-IS
by kveri new 01 Jun '15

01 Jun '15

Hello, I would like to implement IS-IS as a part of my master thesis. Are there any apparent problems which would prevent successful implementation, political and/or technical? Also I would like to hear opinions of the authors. Implementation timeframe is 2013 (if we live past 21th). Regards, Martin Čechvala Faculty of Informatics and Information Technologies Slovak Technical University

7 8

print BGP AGGREGATOR attribute?
by Job Snijders 25 Feb '13

25 Feb '13

Dear all, I want to inject information into syslog when certain magic prefixes are announced or withdrawn. So far I have this snippet of configuration, but I'd also like to print the BGP AGGREGATOR ip address, and the AGGREGATOR AS number. How can I do this? filter collector_in { if (net ~ [ 84.205.64.0/24, 84.205.65.0/24 ]) then { print "MAGIC PREFIX SIGNALLED: " , net, "' from AS", bgp_path.last, " via AS", bgp_path.first, " [", proto, "]"; accept; } else accept; } How can I get BIRD to print something when a prefix is withdrawn? Kind regards, Job

2 2

Re: iBGP
by Miłosz Oller 21 Feb '13

21 Feb '13

After add comment next hop self and add gateway direct: protocol static { route 91.231.205.0/24 reject; route 91.237.68.0/22 reject; } [..] # next hop self; gateway direct; [..] bird> show route 91.231.205.0/24 all 91.231.205.0/24 unreachable [static1 15:27] * (200) Type: static unicast univ via 91.231.205.193 on eth0 [bgp_syscom 15:27] (100) [i] Type: BGP unicast univ BGP.origin: IGP BGP.as_path: BGP.next_hop: 91.231.205.193 BGP.local_pref: 100 Why static route is unreachable? On both routers are two prefixes in static protocol. On second router is the same problem. -- Miłosz Oller Dnia 2012-12-30, nie o godzinie 14:14 +0200, Roman Skopenko pisze: > On 30.12.2012 14:02, Miłosz Oller wrote: > > Session on router no.2: > > protocol bgp bgp_r1_ibgp { > > description "iBGP"; > > import all; > > export all; > > next hop self; > > local as 198611; > > neighbor 91.231.205.193 as 198611; > > source address 91.231.205.194; > > } > > > > bird> show route 91.231.205.0/24 all > > 91.231.205.0/24 unreachable [static1 14:36] * (200) > > Type: static unicast univ > > unreachable [bgp_r2_ibgp 14:37 from 91.231.205.193] > > (100/-) [i] > > Type: BGP unicast univ > > BGP.origin: IGP > > BGP.as_path: > > BGP.next_hop: 91.231.205.193 > > BGP.local_pref: 100 > > > > Any ideas? > > > > I am not sure... > try (if neigbour unreach) > gateway direct; > or/and (if prefixes do not pass thru) > rr client; >

2 7

Meaning of "show protocols" output
by Arnold Nipper 07 Feb '13

07 Feb '13

Does anyone know what "preferred" in the context of rs6-l:~> birdc show protocols all R192_y BIRD 1.3.7 ready. name proto table state since info R192_y BGP Txxxxx up 2013-01-15 11:19:19 Established Description: xxxxx Preference: 100 Input filter: (unnamed) Output filter: (unnamed) Import limit: 150000 Action: restart Routes: 14303 imported, 59164 exported, 22150 preferred means. R192_y is a BGP IPv4 neighbour. TIA, Arnold -- Arnold Nipper / nIPper consulting, Sandhausen, Germany email: arnold(a)nipper.de phone: +49 6224 5593407 2 mobile: +49 172 2650958 fax: +49 6224 5593407 9

2 3

Incorrect manual next hop attribute
by Javor Kliachev 05 Feb '13

05 Feb '13

2 2

Sstatic protocol routes not loaded until extra reload is done on unchanged config.
by Hans van Kranenburg 04 Feb '13

04 Feb '13

Hi all, Loosely based on https://redmine.labs.nic.cz/projects/bird/wiki/Route_server_with_community_… I set up a bird test configuration for an eBGP router which is located in AS64515 and talks to another router in AS65517, using an extra routing table, T65517 and pipe protocol to get fine-grained, and easily debuggable control about which routes I want to announce to the peer, and which received routes I want to import into the local AS: (AS64515) Tmaster - P65517 - T65517 <-> eBGP peer AS65517

2 3

BGPSEC implementation
by Michael Baer 30 Jan '13

30 Jan '13

Hi, I sent mail awhile ago mentioning work being done to implement BGPSEC within the BIRD code and I wanted to give an update. The current code supports the newest version of BGPSEC protocol document as far as capability negotiation and update messages. This includes the cryptographic signing/checking using the openssl libraries. We are continuing to work on it. There are a couple things missing from the update processing (e.g. proper confed handling) and one of the major next steps would be to integrate RPKI data, hopefully using the BIRD ROA tables and pulling from a cache. I can make a patch available of the current code base, if anyone is interested in playing with it (GPL licensed to match the BIRD). Feedback on bugs and improvements would be welcome. I noticed that Ondrej Filip is giving a BIRD presentation at NANOG 57. Several people associated with our project, including Sam cc'd above, will be at that meeting if you want to talk to one of us about it. Let me or Sam know, and one of us can send other contact info privately. Thanks, Mike -- Michael Baer baerm(a)tislabs.com

2 1

Fwd: cloning direct routes into another routing table
by Gustav Koller 29 Jan '13

29 Jan '13

Hello, I'm trying to clone some routes bird leaned via the direct protocol into another kernel routing table. A simple example: log syslog all; debug protocols all; table t_test; protocol device {} # export t_test table to kernel table 100 protocol kernel k_test { table t_test; kernel table 100; export all; } # static route - shows up in table 100 protocol static static_test { table t_test; route 10.1.1.1/32 drop; } # direct routes # doesn't show up in table 100 protocol direct d_test { table t_test; } bird learns the routes perfectly fine but doesn't export them to the kernel table again: # birdc show route table t_test BIRD 1.3.9 ready. 10.1.1.1/32 blackhole [static_test 19:40] * (200) 192.168.0.0/24 dev eth0 [d_test 19:40] * (240) # ip route show table 100 blackhole 10.1.1.1 proto bird This log extract shows the reason for this: k_test: Connected to table t_test k_test < added 10.1.1.1/32 blackhole k_test < rejected by protocol 192.168.0.0/24 dev eth0 I suspect that this is intentional behavior (kernel protocol rejecting routes from direct protocol) as it doesn't make much sense adding direct routes to the kernel in scenarios with a single default/main routing table. But in my case that's what I need. Is there a workaround to make this work? Best regards, Gustav

2 1

Problem with bird 1.3.9 on mips/linux
by Robert Hass 26 Jan '13

26 Jan '13

Hi I compiled bird 1.3.9 on Linux/MIPS (Debian 6.0.6) and I have problem that my bird is not listening on port TCP/179 ? Any comments or recommendations ? Rob # ps auxw | grep bird root 16334 0.0 0.2 2612 1056 pts/0 S 23:46 0:00 ./bird # netstat -a -l | grep bird unix 2 [ ACC ] STREAM LISTENING 84588 bird.ctl # birdc BIRD 1.3.9 ready. bird> show protocols all name proto table state since info device1 Device master up 23:46 Preference: 240 Input filter: ACCEPT Output filter: REJECT Routes: 0 imported, 0 exported, 0 preferred Route change stats: received rejected filtered ignored accepted Import updates: 0 0 0 0 0 Import withdraws: 0 0 --- 0 0 Export updates: 0 0 0 --- 0 Export withdraws: 0 --- --- --- 0 OLIVE4 BGP master start 23:46 Idle Description: JunOS Preference: 100 Input filter: bgp_olive4_in Output filter: bgp_olive4_out Routes: 0 imported, 0 exported, 0 preferred Route change stats: received rejected filtered ignored accepted Import updates: 0 0 0 0 0 Import withdraws: 0 0 --- 0 0 Export updates: 0 0 0 --- 0 Export withdraws: 0 --- --- --- 0 BGP state: Idle Neighbor address: 10.17.0.139 Neighbor AS: 60000 My bird.conf configuration file: log syslog { debug, trace, info, remote, warning, error, auth, fatal, bug }; router id 10.17.0.254; listen bgp address 10.17.0.254 port 179; protocol device { scan time 60; } filter bgp_olive4_in { bgp_local_pref=100; accept; } filter bgp_olive4_out { reject; } protocol bgp OLIVE4 { description "JunOS"; local as 60000; source address 10.17.0.254; neighbor 10.17.0.139 as 60000; import filter bgp_olive4_in; export filter bgp_olive4_out; }

4 10

rs client and bgp soft reload out
by Jimmy Halim 25 Jan '13

25 Jan '13

Hi guys, I have 2 BGP related questions that would like to confirm with you guys: 1. Can we run the route server without setting "rs client" BGP configuration in each client? I can bring up the eBGP as well without that setting in place. 2. Birdc reload out "protocol_name" command will do soft reload on the exported routes to that particular neighbour. However this will reset the uptime. Any idea whether we can keep the uptime? It is very worrying while doing the troubleshooting if the uptime is not reflecting the correct uptime. Thanks, Jimmy

2 2