Re: Hardware requirements for BIRD

2017-03-09 Thread Clément Guivy
Thanks everyone for your answers, that's very helpful to me :-) On 07/03/2017 19:46, Matthew Walster wrote: On 7 March 2017 at 05:57, Clément Guivy mailto:clem...@guivy.fr>>wrote: Hello, I am considering the setup of BIRD as a router to handle our internet traffic. One information I fa

Re: [PATCH,RFC 1/3] BGP: Implement secondary remote AS number support.

2017-03-09 Thread Ondrej Zajicek
On Thu, Mar 09, 2017 at 04:29:40PM +0200, Lennert Buytenhek wrote: > This patch implements the latter option, and this functionality is > enabled by specifying two different remote AS numbers for a peer in > the neighbor statement. Instead of specifying: > > neighbor 1.2.3.4 as 12345; > >

Re: [PATCH, RFC 3/3] BGP: Don't unconditionally bounce connection for remote AS changes.

2017-03-09 Thread Ondrej Zajicek
On Thu, Mar 09, 2017 at 05:36:54PM +0200, Lennert Buytenhek wrote: > > Well, the simple way to avoid reconfiguration in this case is just to move > > 'remote_as2' field at the end of bgp_config, like 'check_link' or 'bfd'. > > Well, e.g. if remote_as is 12345 and the currently established peer AS

Re: Foster one BGP peer in case of AS path equality

2017-03-09 Thread Alarig Le Lay
On jeu. 2 mars 19:23:10 2017, Keenan Tims wrote: > You can accomplish this by setting `med metric on` in your BGP protocols, > and set bgp_med in the import policy to affect which route is selected when > localpref and as-path-length are the same. Hi, I tried this, but forgotten to set med metri

Re: [PATCH, RFC 3/3] BGP: Don't unconditionally bounce connection for remote AS changes.

2017-03-09 Thread Lennert Buytenhek
On Thu, Mar 09, 2017 at 04:23:05PM +0100, Ondrej Zajicek wrote: > > If we are using the secondary remote AS mechanism, we don't necessarily > > need to restart an already established BGP connection if the remote AS > > for a BGP session changes, as long as the AS number that the peer is > > curren

Re: [PATCH, RFC 2/3] BGP: Correctly handle changes to c->{check_link, bfd} when reconfiguring.

2017-03-09 Thread Lennert Buytenhek
On Thu, Mar 09, 2017 at 04:28:43PM +0100, Ondrej Zajicek wrote: > > Since the area between 'struct proto_config c' and 'char *password' > > in 'struct bgp_config' is compared using memcmp() in bgp_reconfigure(), > > move ->check_link and ->bfd so that they are part of that area. Also > > add a no

Re: [PATCH, RFC 2/3] BGP: Correctly handle changes to c->{check_link, bfd} when reconfiguring.

2017-03-09 Thread Ondrej Zajicek
On Thu, Mar 09, 2017 at 04:30:24PM +0200, Lennert Buytenhek wrote: > Since the area between 'struct proto_config c' and 'char *password' > in 'struct bgp_config' is compared using memcmp() in bgp_reconfigure(), > move ->check_link and ->bfd so that they are part of that area. Also > add a note to

Re: [PATCH, RFC 3/3] BGP: Don't unconditionally bounce connection for remote AS changes.

2017-03-09 Thread Ondrej Zajicek
On Thu, Mar 09, 2017 at 04:30:55PM +0200, Lennert Buytenhek wrote: > If we are using the secondary remote AS mechanism, we don't necessarily > need to restart an already established BGP connection if the remote AS > for a BGP session changes, as long as the AS number that the peer is > currently co

[PATCH, RFC 2/3] BGP: Correctly handle changes to c->{check_link, bfd} when reconfiguring.

2017-03-09 Thread Lennert Buytenhek
Since the area between 'struct proto_config c' and 'char *password' in 'struct bgp_config' is compared using memcmp() in bgp_reconfigure(), move ->check_link and ->bfd so that they are part of that area. Also add a note to 'struct bgp_config' to avoid this tripping up people in the future. --- pr

[PATCH,RFC 3/3] BGP: Don't unconditionally bounce connection for remote AS changes.

2017-03-09 Thread Lennert Buytenhek
If we are using the secondary remote AS mechanism, we don't necessarily need to restart an already established BGP connection if the remote AS for a BGP session changes, as long as the AS number that the peer is currently connected with is still in the list of AS numbers that we accept from the pee

[PATCH,RFC 1/3] BGP: Implement secondary remote AS number support.

2017-03-09 Thread Lennert Buytenhek
When carrying out a BGP topology migration such as an AS renumbering or an AS merge, which involves changing the local AS being reported by one or both ends of a BGP session, being able to avoid the need to reconfigure both peers at the same time can significantly reduce the complexity of such a mi

[PATCH,RFC 0/3] Secondary remote AS support for protocol BGP.

2017-03-09 Thread Lennert Buytenhek
Hi! These patches implement a way of simplifying BGP topology migrations that involve changing AS numbers on one or both ends of BGP sessions. The first patch allows a BGP peer to connect to us using either of two different AS numbers specified in our local config. (See the commit message for th

Re: bgp proto bug with add paths in bird 1.6.3

2017-03-09 Thread Ondrej Zajicek
On Thu, Mar 09, 2017 at 02:20:28PM +0200, Lennert Buytenhek wrote: > On Mon, Mar 06, 2017 at 07:41:44PM +0200, Lennert Buytenhek wrote: > > > Hi! > > > > bgp_init() in proto/bgp/bgp.c does: > > > > P->accept_ra_types = c->secondary ? RA_ACCEPTED : RA_OPTIMAL; > > > > and then bgp_rx_open() in

Re: RIP with passive option

2017-03-09 Thread Alexander Festl
Thank you very much! You are right! Ubuntu 10.04 is too old. *ppa:cz.nic-labs/bird* provides only versions for Ubuntu 12.04 or higher. So I had the version from the official repositorie (bird 1.1.5-1). Since the update to version 1.6.3 everything works. Thank you very much! :-) kind regards Alexan

Re: bgp proto bug with add paths in bird 1.6.3

2017-03-09 Thread Lennert Buytenhek
On Mon, Mar 06, 2017 at 07:41:44PM +0200, Lennert Buytenhek wrote: > Hi! > > bgp_init() in proto/bgp/bgp.c does: > > P->accept_ra_types = c->secondary ? RA_ACCEPTED : RA_OPTIMAL; > > and then bgp_rx_open() in proto/bgp/packets.c does: > > if (p->add_path_tx) > p->p.accept_ra_types = RA

RE: RIP with passive option

2017-03-09 Thread DRUART Xavier
It worked with this version NS-BSD 3.0.2- #0: Fri Dec 16 15:49:47 CET 2016 bu...@buildmajtrunk32.labo.int:/usr/home/build/fw-3.0.2/ sys-10.3/work/sys/i386/compile/NETASQ.S.NOSMP.HW.RELEASE Xavier Druart GROUPE NEXEYA Centrale Parc - Bâtiment 2 Avenue Sully Prudhomme 92290 Châtenay-Malabry

Re: RIP with passive option

2017-03-09 Thread Ondrej Zajicek
On Thu, Mar 09, 2017 at 07:47:52AM +, Alexander Festl wrote: > Hi, this is my first post in a mailing list. I hope I do it fine. > > My server should receive routes via RIP. > I have tried it on Ubuntu 10.04 with bird from *ppa:cz.nic-labs/bird* . > The server should work with the interface op

RE: RIP with passive option

2017-03-09 Thread DRUART Xavier
Alexander, I did the test with your configuration and it works. If you have the version 3.0.2 in your SN please upgrade it in version 3.0.2 Xavier Druart GROUPE NEXEYA Centrale Parc - Bâtiment 2 Avenue Sully Prudhomme 92290 Châtenay-Malabry France Tel: 33 (0)1.41.87.25.34 xavier.druart@nexeya.

Re: RIP with passive option

2017-03-09 Thread Alexander Festl
Hey, thank you for your support! I tried "accept" and the other possibility's, with and without the ";" behind the "}" but I get every-time the same error. DRUART Xavier schrieb am Do., 9. März 2017 um 09:14 Uhr: Or may be no need the « ; » behind the »} » *Xavier Druart* *GROUPE NEXEY

RE: RIP with passive option

2017-03-09 Thread DRUART Xavier
Or may be no need the « ; » behind the »} » Xavier Druart GROUPE NEXEYA Centrale Parc – Bâtiment 2 Avenue Sully Prudhomme 92290 Châtenay-Malabry France Tel: 33 (0)1.41.87.25.34 xavier.druart@nexeya.com / www.nexeya.com Retrouvez-nous sur / F

RE: RIP with passive option

2017-03-09 Thread DRUART Xavier
Hello, Did you already try « passive accept ; » Regards Xavier Druart De : Bird-users [mailto:bird-users-boun...@network.cz] De la part de Alexander Festl Envoyé : jeudi 9 mars 2017 08:48 À : bird-users@network.cz Objet : RIP with passive option Hi, this is my first post in a mailing list. I