Date: Mon, 11 Feb 2002 23:39:39 -0800 From: Alfred Perlstein <bright@mu.org> To: Ari Suutari <ari.suutari@syncrontech.com> Cc: Mike Silbersack <silby@silby.com>, Tariq Rashid <tariq@inty.net>, freebsd-net@FreeBSD.ORG Subject: Re: squeeze more performance out of natd? Message-ID: <20020211233939.M63886@elvis.mu.org> In-Reply-To: <200202120650.g1C6o7L40811@guinness.syncrontech.com>; from ari.suutari@syncrontech.com on Tue, Feb 12, 2002 at 08:50:07AM %2B0200 References: <20020211140933.Y84750-100000@patrocles.silby.com> <200202120650.g1C6o7L40811@guinness.syncrontech.com>
next in thread | previous in thread | raw e-mail | index | archive | help
* Ari Suutari <ari.suutari@syncrontech.com> [020211 22:50] wrote: > Hi, > > > On Monday 11 February 2002 16:15, Mike Silbersack wrote: > > On Mon, 11 Feb 2002, Alfred Perlstein wrote: > > > > > another way would be to loop doing recvfrom's until EAGAIN is returned, > > > I suspect this may give at least a 2 fold increase in performance and > > > is trivial to accomplish. > > > > Wow. Yeah, it sounds like that change would make more of a difference > > than moving away from select(). Tell us how it works, Tariq. :) > > > > Natd doesn't use select if you use alias address instead of interface > name. In this case there is just simple recvfrom - sendto loop. Yes, that looks right, still, that case probably doesn't perform all that well at all and could be improved. > The select is there currently mostly to get information from routing > socket which is needed on DHCP / PPP environments. There > was also second purpose - to check space on output socket - > but it has been found unnecessary recently and I think it > was removed from -current, at least. I guess then it looks like natd really either wants batch-recvfrom/sendto or to be moved into the kernel for higher performance. -- -Alfred Perlstein [alfred@freebsd.org] To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020211233939.M63886>