On Sat, Aug 25, 2018 at 08:32:41AM +0200, Mikael Abrahamsson wrote:
> > IMHO, we (network layer) should accept defeat on network layer
> > fragmentation and agree that we should make it easier for the
> > transport layer to resolve the problem.
>
> I want to keep the fragmentation requirement for the network.
Why ? Whats the biggest benefit with IPv6 ?
Took us decades to figure out that in-network
fragmentation (as mandaory in IPv4) is not a good thing, and
we eliminated it for IPv6. Why do we hang on to fragmentation
from the host when tranport layers would be better doing it than the IP
layer ?
Cheers
Toerless
> > Aka: I would lvoe to see a new ICMPv4/ICMPv6 reply and/or PTB reply option
> > indicating "Fragmented Packets Not Permitted". Any network device which
> > for whatever reason does not like Fragemnts would simply drop
> > fragmented packets and send this as a reply. Allows then the
> > transport layer to automatically use packetization (such as TCP MSS)
> > to get packets through.
>
> I am not opposed to this option being created, but you still need PLPMTUD.
> This option might trigger faster PLPMTUD, but it doesn't make the problem go
> away. If the application still keeps sending packets that needs to be
> fragmented, what should the stack do, just send an error to the application?
> Yes, this will mean we will fail faster, but apart from that?
>
> --
> Mikael Abrahamsson email: swm...@swm.pp.se
--
---
t...@cs.fau.de
_______________________________________________
Int-area mailing list
Int-area@ietf.org
https://www.ietf.org/mailman/listinfo/int-area
