On Mon, Mar 25, 2013 at 7:35 AM, Mohan Sundaram <[email protected]> wrote: > 50Gbps by aggregating interfaces on a desktop and seamless handover > across interfaces participating in the bonding. > http://www.multipath-tcp.org/ > > Has anyone gone thro' this in detail? If so, could you educate this group? > > I wonder how this will play with firewalls when different interfaces > go thro' different firewalls? Are there multiple TCP sessions > established across all the firewalls so that data can come in thro' > all the firewalls? >
Mohan, Without going into details I can tell you this much. There are a few ways to do channel bonding. But the best way seems to be aggregating multiple VPN endpoints like a German company pioneered many years ago(10 years). TCP is a higher layer protocol. This project may have applications but TCP is already a very complicated higher level layer. Adding this on top does not make much sense according to me. Instead the IP and data link layers are best to do this. Without playing with routing we cannot achieve good bonding. I have done a rudimentary aggregation with Tenet's CORDECT wireless devices long ago when I was working in Banyan in Teynampet. That was Nucleus OS. And all I did in the outgoing router choosing code is just do a round robin with the two wireless links. Then I also remember testing it in another Tenet company in Adyar, forget the name now. But I think such half baked incomplete solutions are not good. In the UNIX world, you can achieve this with BGP AS policy. You can do this with ML PPP. You can do this with ECMP which is what I normally use.ECMP is equal cost multipath routing. Then for simple link priority and failover I can use the trunk(4) idea in OpenBSD(not VLAN trunk). I feel that VPN approach is best. I am yet to test this idea. -Girish -- Gayatri Hitech http://gayatri-hitech.com _______________________________________________ ILUGC Mailing List: http://www.ae.iitm.ac.in/mailman/listinfo/ilugc
