- Bird-users - Mailing Lists

bird on Xenial doesn't restart after reboot
by Neil Jerram 12 Sep '16

12 Sep '16

Is this the right place to report and discuss an apparent bug with packaging in ppa:cz.nic-labs/bird ? If not, please advise where would be better. If it is the right place... I'd like to report that bird on Xenial doesn't restart after a reboot, using bird version 1.6.0-1+xenial+1 from http://ppa.launchpad.net/cz.nic-labs/bird/ubuntu xenial/main amd64. It appears the package is a bit inconsistent in its init system handling: - It installs init configuration files for SysV, upstart and systemd. - Its postinst (install-time) script calls 'update-rc.d bird defaults' and 'invoke-rc.d bird start'. On Xenial (which uses systemd), the invoke-rc.d call goes through to systemd - or at least, systemctl status bird indicates running after an 'invoke-rc.d bird start', and inactive after an 'invoke-rc.d bird stop' - but the update-rc.d call does not call on to 'systemctl enable bird'. Hence the observation that bird is not restarted on reboot. Thoughts? From analogy with other packages (that are systemd-enabled when installed), I think the fix may be as simple as: - adding '--with systemd' to the dh invocation - adding dh-systemd as a build dep. Thanks, Neil

2 2

bgp : rejected by protocol
by Patrick Lamaiziere 05 Sep '16

05 Sep '16

Hello, I'm not sure to understand what "rejected by protocol" means. When a (bgp) route is updated on a bgp peer, bird learns the route and the route is then always rejected by protocol (bgp protocol). log: bgp_renater: Got UPDATE bgp_renater > added [best] 0.0.0.0/0 via 193.XX.XX.XX on ix0 bgp_renater < rejected by protocol 0.0.0.0/0 via 193.XX.XX.XX on ix0 config: protocol static routes_renater { table Trenater; route 129.20.0.0/16 via 129.20.252.1; ... } # peer BGP RENATER protocol bgp bgp_renater from bgptemplate { table Trenater; description "RenaterV4"; neighbor 193.XX.XX.XX as 2200; export filter { if proto != "routes_renater" then reject; accept; }; import filter fbgp_renater_in; } Thanks, regards.

1 0

[PATCH] birdc: wider columns in show protocol output
by Simon Ruderich 01 Sep '16

01 Sep '16

-- Hello, The show protocol output gets more difficult to read with longer protocol names. This patch increases the column width to 15 characters. Regards Simon Index: bird-1.6.0/nest/proto.c =================================================================== --- bird-1.6.0.orig/nest/proto.c +++ bird-1.6.0/nest/proto.c @@ -1498,13 +1498,13 @@ proto_cmd_show(struct proto *p, uint ver /* First protocol - show header */ if (!cnt) - cli_msg(-2002, "name proto table state since info"); + cli_msg(-2002, "name proto table state since info"); buf[0] = 0; if (p->proto->get_status) p->proto->get_status(p, buf); tm_format_datetime(tbuf, &config->tf_proto, p->last_state_change); - cli_msg(-1002, "%-8s %-8s %-8s %-5s %-10s %s", + cli_msg(-1002, "%-15s %-8s %-8s %-5s %-10s %s", p->name, p->proto->name, p->table->name, -- + privacy is necessary + using gnupg http://gnupg.org + public key id: 0x92FEFDB7E44C32F9

5 6

How to make a aggregate function in bird ?
by Xavier Beaudouin 01 Sep '16

01 Sep '16

Hello, I am attempting to make some intelligent stuff on one lab bird. I am sorry if this has been already asked since I do not have very mutch experience about bird, and some of my google:// didn't find what I am looking for. My idea is to find which aggregate I have within a bunch of subnet I have. For example : get_aggregate( 192.0.2.1/32 ) will return 192.0.2.0/24 Is there any nice way to make such function ? Also in the same work I would like to find how old is the route received in BGP. Any good pointers or examples are really appreciated. Kind regards, Xavier

1 0

Generate Defaul Route
by André Carlim 29 Aug '16

29 Aug '16

Hello, first, sorry for my bad English, but I need help, I am trying to make the ospf the bird send the route default same does not being present in table main kernel, I have multiple tables, ex to the protocol ospf have the table ospf to BGP table BGP, and so on, I even have a route standard in table default kernel, I tried various forms configuration, but not got success, someone already explained that? []'s André Carlim via CloudMagic Email [https://cloudmagic.com/k/d/mailapp?ct=pa&cv=8.6.38&pv=6.0.1&source=email_fo…]

2 1

Question about graceful restart timer
by Robert Wein 25 Aug '16

25 Aug '16

Hello bird users, one of our VIX customers (RS-Client) came up with a question about the graceful restart timer. At his Cisco he found: ------------ Graceful Restart Capability: advertised and received Remote Restart timer is 0 seconds ------------ We don't have anything related to graceful restart in our config, defaults should apply. At packet level, I found in BGP Open sent from bird: ------------ Capability: Route refresh capability Type: Route refresh capability (2) Length: 0 Capability: Graceful Restart capability Type: Graceful Restart capability (64) Length: 2 Restart Timers: 0x0000 0... .... .... .... = Restart: No .... 0000 0000 0000 = Time: 0 ------------ Next, I tried to set the timer explicitly, which ended up again with restart timer 0 in BGP Open. The only way to get the correct timer in the packet was to set graceful to yes in the protocol definition. ------ graceful restart time 10; graceful restart yes; ------ As far as I understand the default for graceful restart ("aware") should be enough to advertise capability and timer. I am grateful for every advice, regards Robert -- www.vix.at | www.aco.net robert.wein(a)univie.ac.at | RW2586-RIPE Vienna University Computer Center Tel: +43 1 4277-14046 | Fax: -14338

2 1

Patch for IPv6 ECMP support?
by Neil Jerram 23 Aug '16

23 Aug '16

Does anyone know what happened to this patch/discussion: http://bird.network.cz/pipermail/bird-users/2016-February/010192.html I believe it (or similar) is needed for IPv6 merge paths to work, and I'm interested in that working. Thanks, Neil

4 7

New in BIRD
by Ilias Pallikarakis 23 Aug '16

23 Aug '16

Hi, I recently started trying to learn BIRD. I should also mention I have no previous experience with any other Routing daemon either. I want to configure a bgp monitor for both ipv4 and ipv6. I am running BIRD in a Ubuntu 13.10 VM. I managed to install both bird and bird6 but I have some problems with the configurations. In ipv4 : I created a very simple configuration which I attach. The conf is compiled but in birdc the bgp session is considered idle. I tried to do a tcpdump and found that the router sends a TCP SYN on the VM's bgp port but the VM responds with a RST and never tries to set a BGP session itself. In ipv6 : No matter the configuration (my configuration or the sample coming with bird6 or examples found on the web) the configuration is not compiled giving the following error : BIRD ipv6 doesn't run (bird: /usr/local/etc/bird6.conf, line 1: Router ID must be configured manually on IPv6 routers) Could anyone give me some pointers on what I am doing wrong ? Thank you for your help, Ilias Pallikarakis

5 8

Less specific with community
by Max Tulyev 19 Aug '16

19 Aug '16

Hi All, is it possible in BIRD BGP incoming filter to check if there are less specific prefixes with certain community on it? Thank you!

1 0

multipath recursive issue
by micah anderson 09 Aug '16

09 Aug '16

I'm using bird 1.4.5 and I have a multihop bgp peer that I advertise a couple networks to, and they provide me with a full feed. My directly connected interface is configured like this: inet 208.99.192.142/30 brd 208.99.192.143 scope global eth3 but the neighbor config points to a different address (because its multihop): neighbor 208.99.192.121 as xxxx; it seems like I get this from them too: BGP.next_hop: 208.99.192.121 With this setup, I get these constant spammings in the logs: 2016-08-09 18:18:25 <WARN> Next hop address 208.99.192.121 resolvable through recursive route for 208.99.192.0/23 so did some reasearch and I managed to find this post[0] which suggested I add a static /32 so it can be resolved through a route that is not from BGP, so I did that: route 208.99.192.121/32 via 208.99.192.142%eth3; but then I get this flooding for every network I get from them: 2016-08-09 18:17:22 <ERR> KRT: Received route 223.255.210.192/26 with strange next-hop 208.99.192.142 so then I found this post[1], and so I tried: root@wren:/etc/bird# ip addr add 208.99.192.142 peer 208.99.192.121 dev eth3 but that didn't help... and bird seems to be eating a lot of CPU (100% of one CPU) when it wasn't doing that so much before. any ideas would be helpful! micah 0. http://bird.network.cz/pipermail/bird-users/2011-May/006842.html 1. http://bird.network.cz/pipermail/bird-users/2012-April/002990.html

1 0