[163194] in North American Network Operators' Group
Re: High throughput bgp links using gentoo + stipped kernel
daemon@ATHENA.MIT.EDU (Ryan Gard)
Fri May 24 15:18:12 2013
In-Reply-To: <CAHf3uWyGYrUQ0UP_vc8oB+_9eR_L01y3y66-n_LL6wuWJWAmLA@mail.gmail.com>
Date: Fri, 24 May 2013 15:17:39 -0400
From: Ryan Gard <ryangard@gmail.com>
To: NANOG <nanog@nanog.org>
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org
Do you have a source on this? Reason I ask is because any recent
documentation I've come across indicates that polling is recommended to
reduce chances of livelock on a running system.
On Mon, May 20, 2013 at 2:51 PM, Eduardo Schoedler <listas@esds.com.br>wrote:
> 2013/5/19 Andrew Jones <aj@jonesy.com.au>
>
> > As for migration to another OS, I find FreeBSD better as a matter of
> >> network performance. The last time I checked OpenBSD was either
> >> lacking or was in the early stages of multiple cores support.
> >>
> >
> > If you do decide to go the FreeBSD route (you can run openbgpd on FreeBSD
> > if you like), check out the POLLING option on ethernet NICs, it cuts down
> > on the number of interrupts and can increase performance, particularly
> when
> > dealing with smaller packets.
> >
>
> Polling on FreeBSD in modern NICs is discouraged.
>
> --
> Eduardo Schoedler
>
--
Ryan Gard
