On 1/6/2011 2:17 PM, TJ wrote:
Again, off the top of my head, maybe - when under duress - age out the incomplete ND table entries faster.
Given that the incomplete age is to protect the L2 network from excessive broadcast/multicast, I agree that aging them out fast would be a wiser solution, if you must have it to begin with. It is better to increase traffic loads. I'm still a proponent for removing as needed requests like this, though. It would have been better to send a global "everyone update me" request periodically, even if triggered by an unknown entry, yet limited to only broadcasting once every 10-30 seconds. Given that all requests for an unknown arp/ND entry results in all hosts on the network checking, it only makes sense for all hosts to respond. There may be other concerns, but I'm actually not against all hosts responding via multicast to all other hosts, so that a full mesh can be established ahead of time. The idea of minimizing the table to an as-needed basis should not have continued with IPv6. Special provisions could be handled when dealing with proxy-ND, but I'm not sure that is needed either. Jack