Date: Wed, 21 Jan 2004 19:55:35 +0800 (MYT) From: Dinesh Nair <dinesh@alphaque.com> To: Adrian Pavlykevych <pam@polynet.lviv.ua> Cc: freebsd-net@freebsd.org Subject: Re: Two ISP lines Message-ID: <20040121195028.M532-100000@prophet.alphaque.com> In-Reply-To: <400E614F.4060804@polynet.lviv.ua>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, 21 Jan 2004, Adrian Pavlykevych wrote: > Hmm, have you accounted for the fact, that if packets of same connection > will be distributed in round-robin fashion between several outgoing > interfaces and get their src IP "fixed", destination host won't see that > as single connection and drop them? you're absolutely right, of course. i realised the same a couple of hours ago, but am still mulling over the idea of a netgraph node instead of the multipath patch. one does get tired eyes from reading too many screenfuls of tcpdump and ng_tee outputs. :) > On routing level, IMHO, patch operation, you've described, is exactly > following multipath functionality. Issue with return traffic could be > resolved with both ISPs announcing route to your IP to the Internet. In > this way, return packets will be routed from the destination according > to routing decisions take by routers on the path. true, but still, if you're behind nat boxes on both interfaces, multipath still wont help. multipath as it stands now only works if both interfaces have public ip addresses. oh well, back to the drawing board. thanx for being a sounding board, adrian. Regards, /\_/\ "All dogs go to heaven." dinesh@alphaque.com (0 0) http://www.alphaque.com/ +==========================----oOO--(_)--OOo----==========================+ | for a in past present future; do | | for b in clients employers associates relatives neighbours pets; do | | echo "The opinions here in no way reflect the opinions of my $a $b." | | done; done | +=========================================================================+
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040121195028.M532-100000>