I think I would agree that the "no ip-directed broadcast" command  is the
way to go here. It has the additional advantage of preventing the Fraggle
attack also as the command would drop packets based on the IP address not
the protocol so both ICMP and UDP (and TCP etc for that matter) would be
dropped on the interface if trying to target a broadcast address behind the
router.

Rate limiting I don't think would have any effect as the damage of the
attack is caused by the response traffic to the spoofed host not the amount
of inbound ICMP traffic from the attacker. With a large enough network, one
would not need many inbound ICMP packets to the broadcast address to create
a large number of ICMP reply packets back to the spoofed host.

uRPF would not solve the problem because even if the Smurf or Fraggle
attack came from a spoofed source address (which it likely would) that
doesn't mean that the Echo Request packets would still not enter the router
on the correct interface and pass any uRPF checks, especially if it was an
internet connected router in which the Echo packets came in the interface
via which the default route is reached.

I am not sure about the Selective Directed Broadcast option Eugene
mentions. I hadn't heard of the option before but after a very brief read
of the RFC it seems that it is used more for providing a UDP packet with
multiple destination addresses (one of which I suppose could be a broadcast
address) but it would seem that one does not need to use this IP Option to
specifically target a broadcast address as this can be done with a standard
packet without using the SDB IP Option.

Thanks
Ben
_______________________________________________
For more information regarding industry leading CCIE Lab training, please visit 
www.ipexpert.com

Are you a CCNP or CCIE and looking for a job? Check out 
www.PlatinumPlacement.com

Reply via email to