- Bird-users - Mailing Lists

Re: Pointers please - OSPFv3 vs Cisco IOS XR
by Elmar K. Bins 07 Jul '20

07 Jul '20

Hi Ondrej, sorry, I missed your response (T., thank you for forwarding): >> I am very certain somebody has experienced similar issues and solved them; I >> did not find any pointers - please throw them at me. If this is new or unsolved, >> please help me forward, I can get more output, naturally. > > Hi > > It might be useful if you enable 'debug all' for OSPFv3 in BIRD. Alright...sorry, I thought I had "all" on, but apparently didn't. Here's the output from that, shortened and edited to protect the innocent. (If you require the full output, I'll email it to you) Thank you for taking a look at this, Elmar. ======================================================================================= 2020-06-03 12:30:13.801 <TRACE> ospf2: HELLO packet received from nbr <router-1> on igb0 2020-06-03 12:30:16.899 <TRACE> ospf2: DBDES packet received from nbr <router-1> on igb0 2020-06-03 12:30:16.899 <TRACE> ospf2: length 28 2020-06-03 12:30:16.899 <TRACE> ospf2: router <router-1> 2020-06-03 12:30:16.899 <TRACE> ospf2: mtu 1500 2020-06-03 12:30:16.899 <TRACE> ospf2: imms I M MS 2020-06-03 12:30:16.899 <TRACE> ospf2: ddseq 147850133 2020-06-03 12:30:16.899 <TRACE> ospf2: Neighbor <router-1> on igb0 changed state from ExStart to Exchange 2020-06-03 12:30:16.899 <TRACE> ospf2: DBDES packet sent to nbr <router-1> on igb0 2020-06-03 12:30:16.899 <TRACE> ospf2: length 108 2020-06-03 12:30:16.899 <TRACE> ospf2: router <bird-ip> 2020-06-03 12:30:16.899 <TRACE> ospf2: mtu 1500 2020-06-03 12:30:16.899 <TRACE> ospf2: imms 2020-06-03 12:30:16.899 <TRACE> ospf2: ddseq 147850133 (a handful of LSAs) 2020-06-03 12:30:16.901 <TRACE> ospf2: DBDES packet received from nbr <router-1> on igb0 2020-06-03 12:30:16.901 <TRACE> ospf2: length 1228 2020-06-03 12:30:16.901 <TRACE> ospf2: router <router-1> 2020-06-03 12:30:16.901 <TRACE> ospf2: mtu 1500 2020-06-03 12:30:16.901 <TRACE> ospf2: imms M MS 2020-06-03 12:30:16.901 <TRACE> ospf2: ddseq 147850134 (more LSAs) 2020-06-03 12:30:16.901 <TRACE> ospf2: DBDES packet sent to nbr <router-1> on igb0 2020-06-03 12:30:16.901 <TRACE> ospf2: length 28 2020-06-03 12:30:16.901 <TRACE> ospf2: router <bird-ip> 2020-06-03 12:30:16.901 <TRACE> ospf2: mtu 1500 2020-06-03 12:30:16.901 <TRACE> ospf2: imms 2020-06-03 12:30:16.901 <TRACE> ospf2: ddseq 147850134 2020-06-03 12:30:16.903 <TRACE> ospf2: LSREQ packet sent to nbr <router-1> on igb0 2020-06-03 12:30:16.903 <TRACE> ospf2: length 736 2020-06-03 12:30:16.903 <TRACE> ospf2: router <bird-ip> (some LSRs) 2020-06-03 12:30:16.906 <TRACE> ospf2: LSREQ packet received from nbr <router-1> on igb0 2020-06-03 12:30:16.906 <TRACE> ospf2: length 52 2020-06-03 12:30:16.906 <TRACE> ospf2: router <router-1> (more LSRs) 2020-06-03 12:30:16.906 <TRACE> ospf2: LSUPD packet sent to nbr <router-1> on igb0 2020-06-03 12:30:16.906 <TRACE> ospf2: length 144 2020-06-03 12:30:16.906 <TRACE> ospf2: router <bird-ip> (some LSAs) 2020-06-03 12:30:16.906 <TRACE> ospf2: DBDES packet received from nbr <router-1> on igb0 2020-06-03 12:30:16.906 <TRACE> ospf2: length 28 2020-06-03 12:30:16.906 <TRACE> ospf2: router <router-1> 2020-06-03 12:30:16.906 <TRACE> ospf2: mtu 1500 2020-06-03 12:30:16.906 <TRACE> ospf2: imms I M MS 2020-06-03 12:30:16.906 <TRACE> ospf2: ddseq 147850135 2020-06-03 12:30:16.906 <RMT> ospf2: Bad DBDES packet from nbr <router-1> on igb0 - I-bit mismatch (7) =======================================================================================

3 5

Minor usability enhancement for birdc - show hostname
by Tim Small 04 Jul '20

04 Jul '20

Hi, I'm finding Bird very useful, thanks! I wondered if it might be useful to show the hostname in the bird prompt? e.g. instead of: bird> Something like: testlab1 bird> ... would spare users the potential frustration or downtime of operating on the wrong bird instance. I couldn't see a way of doing this, but would be happy to submit a patch (to make this configurable or otherwise) if that would be useful? Thanks! Tim.

3 2

Kernel protocol and multiple routes
by Magnus Löfqvist 04 Jul '20

04 Jul '20

Hi, Im running bird 1.6.6 on a openwrt device, having two internet connections (one over 4G and one over ethernet). The main routing table : default via 10.70.69.193 dev wwan0 proto static src 10.70.69.192 default via 10.3.5.1 dev eth1 proto static src 10.3.5.129 metric 10 10.3.5.0/24 dev eth1 proto static scope link metric 10 10.3.5.1 dev eth1 proto static scope link src 10.3.5.129 metric 10 10.70.69.128/25 dev wwan0 proto kernel scope link src 10.70.69.192 10.70.69.193 dev wwan0 proto static scope link src 10.70.69.192 As you see I get the default gateway over 4G as primary (lowest metric), and that are by design. I want to use bird to import the default gateways and pipe them to two different tables. Relevant Config: filter export_wan { if proto = "krn_main" then { if ifname = "wwan0" then { if (net = 0.0.0.0/0) then { accept; } } } reject; } filter export_wan2 { if proto = "krn_main" then { if ifname = "eth1" then { if (net = 0.0.0.0/0) then { accept; } } } reject; } protocol kernel krn_main { learn; import all; scan time 5; kernel table 254; merge path yes; } protocol kernel krn_wan { learn; import none; export all; kernel table 200; scan time 5; table wan; } protocol kernel krn_wan2 { learn; import none; export all; kernel table 200; scan time 5; table wan2; } protocol pipe pipe_wan { peer table wan; export filter export_wan; import all; } protocol pipe pipe_wan2 { peer table wan2; export filter export_wan2; import all; } # birdc4 show route 0.0.0.0/0 all protocol krn_main BIRD 1.6.6 ready. 0.0.0.0/0 via 10.70.69.193 on wwan0 [krn_main 2020-07-03] (10) Type: inherit unicast univ Kernel.source: 4 Kernel.metric: 0 Kernel.prefsrc: 10.70.69.192 As you see Im only get the route for 4G interface (wwan0) and not for the ethernet (eth1). Why is that, and what can be done to fix it? / Magnus Sent from Nine<http://www.9folders.com/>

1 0

Bird 2 for CentOS/EPEL 6?
by Tim Meusel 02 Jul '20

02 Jul '20

Hi people, thanks for the great software! I'm happily using Bird 2 on multiple CentOS 7 nodes. Sadly I now need to rollout BGP to some CentOS 6 boxes :sadface:. I noticed that ftp://bird.network.cz/pub/bird/centos/ only contains CentOS 7 packages, and EPEL 6 only has Bird 1.6.8. However I wanted to ask if Bird 2 would work on CentOS 6? Would Bird itself compile but dependencies are outdated/missing? Cheers, Tim

3 3

help bgp peer issue.
by seunghwan hwang 29 Jun '20

29 Jun '20

Hello. I'm using BIRD 1.6.3. Configurations(Peer is Juniper of Mellanox Switch): protocol 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 merge paths on; # Enable ECMP Multi path. } protocol direct { interface "lo"; } protocol device { scan time 10; # Scan interfaces every 10 seconds } ... protocol bgp peer1 { import all; export all; ... hold time 3; ... } protocol bgp peer2 { import all; export all; ... hold time 3; ... } In this situation, sometimes the BGP "Hold Time Expire" log remains and peers destroyed. But when I capture tcpdump, I can check the kernel sent an TCP ACK message for BGP keepalive packet. I think BGP Keepalive packets can't be transmitted to the BIRD(from the kernel). Is there an issue related to this in this version? And I would like to know if there are issues related to "scan time" and netlink. Please check it. Regards.

2 2

[PATCH 1/2] Filter: Fix comparison of BGP path mask
by Kazuki Yamaguchi 28 Jun '20

28 Jun '20

Add a missing return statement. Path masks with the same length were all considered the same. Comparing two with different length would cause out-of-bounds memory access. This bug was introduced by version 2.0.4, commit 4c553c5a5b40 ("Filter refactoring: dropped the recursion from the interpreter", 2018-12-27). --- filter/data.c | 1 + 1 file changed, 1 insertion(+) diff --git a/filter/data.c b/filter/data.c index 402202554c2f..9547a4c84833 100644 --- a/filter/data.c +++ b/filter/data.c @@ -230,6 +230,7 @@ static int pm_same(const struct f_path_mask *m1, const struct f_path_mask *m2) { if (m1->len != m2->len) + return 0; for (uint i=0; i<m1->len; i++) if (!pmi_same(&(m1->item[i]), &(m2->item[i]))) -- 2.27.0

2 2

Another newbie question: Creating BGP prefixes in a filter
by Elmar K. Bins 22 Jun '20

22 Jun '20

Hi guys, (status update: no solution for the OSPFv3 <--> IOS XR yet, no idea where to look, both sides pretend to be happy but are not) So.... I need to create a prefix to be BGP-advertised to peers and transits; this would look like: if (net = xxx/32) { remove it and replace by yyy/zz; } (more or less standard prefix aggregation on the border) I could not find a hint of documentation in regard of doing a thing like this, where should I look? (We are currently forced to configure the wrong netmasks on loopback addresses, so BIRD has something to export) Yours, Elmar.

2 2

OSPF RFC 6860 Support
by Jake Mercer 21 Jun '20

21 Jun '20

Hi all, I was thinking of adding RFC 6860 support to the OSPF implementation (hiding transit-only networks); is there any appetite for this? Or has someone already looked at it? I have a hacky modification to make it work for PtP networks for my specific purpose, but I’m happy to implement it properly in all cases and submit a patch. Is this mailing list the correct place to submit code? Best regards, Jake Jake Mercer Senior Embedded Software Engineer [cid:logo_796e1f33-efd0-4657-bc9a-9b3e989e6bb3.png]<https://www.civica.com> [cid:instagram_99d052c0-ba99-4d14-9d03-2f8ac9ed66bd.png]<https://www.instagram.com/lifeatcivica> [cid:twitter_f5b6718e-d9bd-4350-a74c-f8d9b1774b3e.png] <https://twitter.com/CivicaUK> [cid:youtube_e9775c8c-7e04-4ef9-b8ca-4b19b46c1325.png] <https://www.youtube.com/user/CivicaUK> [cid:linkedin_92fd785b-4baa-4769-aba2-200d7d2cd7a8.png] <https://www.linkedin.com/company/civica> ------------------------------------------------------------------------------------------- This email is sent for and on behalf of Civica UK Ltd, Civica NI Ltd, Civica Group Ltd or Civica Election Services Ltd. Civica Group Ltd (no: 04968437), Civica UK Ltd (no: 01628868) and Civica Election Services Ltd (no: 02263092) are registered in England and have their registered office at Southbank Central, 30 Stamford Street, London, SE1 9LQ. Civica NI Ltd (no: NI043987) is registered in Northern Ireland and has its registered office at 10 Weavers Court, Belfast, BT12 5GH. See https://www.civica.com/en-gb/policies-and-statements/privacy-notice for our privacy notice. Confidentiality: This email and its attachments are intended for the above named only and may be confidential. If they have come to you in error you must take no action based on them, nor must you copy or show them to anyone. If you have received this in error, please advise the sender by replying to this email immediately highlighting the error and deleting it from your system. -------------------------------------------------------------------------------------------

2 1

[radv] Backport solicited ra unicast to 1.6.x?
by Boris Lytochkin 17 Jun '20

17 Jun '20

Hi. I am curious if commit 70a4320bdd44122d7a93bc71c77a9d684b3c9adc could be cherry-pick'ed into 1.6.x branch? I've tested it and it's pretty straightforward: the only thing that needs to be changed is "current_time()" with "now". Aside from that, it just works out-of-box. -- Boris Lytochkin Yandex NOC +7 (495) 739 70 00 ext 7671

1 0

Static routes BGP
by Fabiano D'Agostino 15 Jun '20

15 Jun '20

Hi everyone, I have this configuration: protocol static { ipv4 { preference 110;}; route 10.0.0.0/8 via 192.168.1.1; } protocol bgp { local as 1; neighbor x.x.x.x as x ipv4 { import all; export filter {if (net ~ [10.0.0.0/8] ) then accept; reject;}; #also tried with export all; }; } On route-server I have: ipv4 table test_4; protocol pipe Ptest_4 { table master4; peer table test_4; import all; export all; } protocol bgp Btest_4 from rsclient_4 { neighbor x.x.x.x as x multihop 15; ipv4 { table test_4; import all; export all; }; } I don’t understand why 10.0.0.0/8 is not in master4 and test_4 tables of the route-server. How could I announce a test prefix from a client to the route-server? Thanks in advance, Fabiano D'Agostino

1 0