On 1-aug-2007, at 16:56, Matt Mathis wrote:
A further suggestion: the performance of jumbo ARP/padded ND can be
improved
if the router is configured in advance with the likely MTU sizes on
the
subnet. For the most common case, with one supported jumbo size
plus the
standard, this approach only requires one extra directed ARP/ND per
peer
(assuming that long term caching is ok).
This in not unreasonable, give that all >1500 Ethernet solutions
requires at
least some minimal configuration on each device.
In my opinion, it's extremely important to move away from any special
configuration for larger packets. We're only going to make good
progress if hardware that supports larger packets will automatically
use this capability where appropriate.
However, if, until the time that any new mechanisms that we define
are widely implemented, making configuration adjustments makes
everything work better, that's fine, of course.
BTW, does anyone have a description of jumbo ARP? I want to see if
there's the possibility to sneak in an MTU field so it's possible to
check for the maximum supported MTU quickly rather than probe
extensively, if both ends support this. Apparently padding ARP
packets works most of the time. I'm assuming that the content of the
padding probably doesn't matter, so putting the MTU somewhere in the
padding shouldn't lead to trouble...
_______________________________________________
Int-area mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/int-area
