On Wed, 3 Mar 2004, Bob Hinden wrote: > I would agree with your concern if it worked that way. The load balancing > being proposed is not load balancing on a per packet basis. It is load > sharing when the host is about to pick a router when sending to a new > destination. [...]
Note that Changming was also raising an issue of "flows" (using intentionally vague terminology). Some classes of packets, which do not have the same destination address may belong together. For example SIP data and SIP signalling. Having them redistributed to other routers would be disadvantageous. The same also apply to (rather small) extent in certain cases where you check that an ICMP Destination Unreachable response is sent in response to a valid packet that the firewall has seen. This is hampered in a few corner cases at least. So, my point is that host-to-router load sharing works on the link, but is causing big problems further down in the network, e.g., in failure cases (some connections die and others don't: difficult to debug), firewalls, etc. -- without a good cause, if load sharing is not needed to well, share the load. -- Pekka Savola "You each name yourselves king, yet the Netcore Oy kingdom bleeds." Systems. Networks. Security. -- George R.R. Martin: A Clash of Kings -------------------------------------------------------------------- IETF IPv6 working group mailing list [EMAIL PROTECTED] Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
