Mikael is correct;

IPv6 mechanisms are different.

SLAAC adds broadcasts that are not present in IPv4, MLD report then NS DAD 
then, sometimes and though it is not required by the spec, NA(O).

IPv6 nodes tend to create multiple addresses, many of which are temporary for 
privacy reasons. So the above happens again and again.

Finally the emergence of smaller hosts like phones have changed the average 
mobility profile of devices, making things even worse. Devices may for instance 
broadcast an RS, form new addresses and/or DAD again after a movement or a 
sleep period.

With classical ND Wireless devices shoot themselves in the foot, and the more 
they move and the larger the subnet, the worst they make the situation for one 
another.

The association process that Wi-Fi uses at layer 2 is more appropriate. It is 
also appropriate for sleeping devices as Pat discussed, since the registration 
can request sleeping proxy operations from the router.

Cheers,

Pascal

Le 8 août 2015 à 16:01, Juliusz Chroboczek <[email protected]> a 
écrit :

>> I'd imagine the amount of multicast for IPv6 is more than 10x (just
>> guessing) larger unless mitigation is put in place,
> 
> I'm perhaps confused, but I don't see why.  ARP is somewhat loosely
> specified, so it's difficult to say for sure without checking the
> implementations, but I'd expect ND to be just as efficient as ARP.
> 
>> If you miss a couple of consecutive RAs in IPv6, you abandon your prefix
>> and your default route.
> 
> I'm confused again.  PIO lifetimes are on the order of hours, or even
> days, while unsolicited RAs are sent every 60s.  Plus there's nothing
> preventing you from sending them more often.
> 
> -- Juliusz
> 
> _______________________________________________
> homenet mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/homenet

_______________________________________________
homenet mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/homenet

Reply via email to