Incorrect manual next hop attribute

[Ondrej Zajicek]

On Wed, Jan 23, 2013 at 01:33:16PM +0200, Javor Kliachev wrote:
>    Hello,
> 
>    We're using BIRD on Linux Debian 6 as a BGP route server with two subnets.
>    We would also like to manually change the next hop of certain prefixes and
>    peers.
>    When both subnets are configured at the single physical interface or in
>    alias
>    mode everything works fine. As soon as we configure any of the subnet at a
>    vlan
>    interface the next hop manipulation fails and the ip address of the RS is
>    set instead.
> 
>    Any ideas what is the reason for this behaviour?

Hello

The reason is BGP standard (RFC 4271 5.1.3). The next hop modification
in the export filter is processed and after that standard BGP processing
of next hop is done, which, in this case, replace the next hop with
local address because conditions that would allow to send supplied one
are not met (which is essentially matching iface based on subnet check
for single-hop eBGP).

-- 
Elen sila lumenn' omentielvo

Ondrej 'SanTiago' Zajicek (email: santiago at crfreenet.org)
OpenPGP encrypted e-mails preferred (KeyID 0x11DEADC3, wwwkeys.pgp.net)
"To err is human -- to blame it on a computer is even more so."
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: Digital signature
URL: <http://trubka.network.cz/pipermail/bird-users/attachments/20130124/0b27d67b/attachment-0001.asc>