Sebastien Roy wrote:
...
Path-MTU discovery might be tricky without some sort of
per-destination cache, however...
Yes...an idea that tumbled through my brain earlier was to do
something akin to copy-on-write when we receive notification
that the MTU to a particular host is different to the MTU in
the current IRE cache.
I need to go and look at the PMTU RFCs to see whether it is
appropriate for us to do per-connection or per-host PMTU
discovery. Doing it per-host does save us a couple of packets
on each connection but otherwise...?
Darren
_______________________________________________
networking-discuss mailing list
[email protected]
