Florian Weimer wrote:
* Ray Hunter:
Are 2^24 interface identifiers small enough that every implementation
could simply provide enough resources (for ND) to cope with all
addresses being in play simultaneously?
You'd need per-interface and per-VLAN tables. I don't think that's
feasible.
So where's the limit for ND and prefix length on today's implementations
and platforms?
What if SLAAC was limited to using 20 bits or even just 16 bits of MAC
BIA information or even in extremis dare I say 12 bits?
That would mean taking one or even two steps closer to Appletalk's node
ID auto configuration mechanism. [implying higher chance of collisions
and that DAD would almost certainly have to be adjusted to allow
generation of random node IDs, and then testing for uniqueness on
interface status transition. ].
IMHO There's no reason why this (temporary) limit on the amount of MAC
BIA information used for SLAAC couldn't change over time as the power of
platforms increases, and that this limit could also be dynamically
advertised via RA as a per prefix option.
Or are you saying that a brute force solution (a la ARP) is simply a no go?
regards,
RayH
--------------------------------------------------------------------
IETF IPv6 working group mailing list
[email protected]
Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------
