On Wed, Jan 29, 2020 at 03:29:47PM +0100, Adam Kułagowski wrote:
Hi,
I've found that It is supported since 4.14:
https://github.com/torvalds/linux/commit/8917a777be3ba566377be05117f71b93a5f...
Great, so it could be easily fixed, will check that.
In FRR I hit the same issue (when using range 0.0.0.0/0) but I was able to resolve using: bgp listen range 0.0.0.0/1 peer-group quiz bgp listen range 128.0.0.0/1 peer-group quiz
I'll try to check if the same approach works on BIRD.
Best regards, Adam
On Wed, Jan 29, 2020 at 3:17 PM Ondrej Zajicek <santiago@crfreenet.org> wrote:
On Tue, Jan 28, 2020 at 07:57:42PM +0100, Adam Kułagowski wrote:
Hi,
Is the following configuration supported ?
Hi
Probably not. The neighbor address (in this case 0.0.0.0) is passed to kernel in TCP_MD5SIG socket option, but i guess that passing zero address just make it to fail silently. There is also (AFAIK) no way to pass a range to kernel. So probably we should just add configuration check and reject password together with range.
protocol bgp { debug { states, routes, filters, interfaces, events, packets }; debug all; ipv4 { import none; export filter export_bgp; }; local as 65001; neighbor range 0.0.0.0/0 as 65001 internal; #neighbor 10.255.254.3 as 65001 internal; password "12345678"; }
-- Elen sila lumenn' omentielvo
Ondrej 'Santiago' Zajicek (email: santiago@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."
-- Elen sila lumenn' omentielvo Ondrej 'Santiago' Zajicek (email: santiago@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."