On Dec 7, 2021, at 3:47 PM, Dino Farinacci <farina...@gmail.com> wrote:
>
> And note if you get rid of data link MTUs, your head-of-line-blocking issue
> gets worse. Also note 1280 is not 53, and hence we have an international
> large scale network running, unlike ATM.
It’s more like I want to get rid of the reduction of user payload space to make
room for headers.
I don’t think link MTUs should be unlimited per se, but rather that we COULD
treat the innermost user payload as the thing we limit, where everything else
is “overhead” that makes the segment longer, rather than the payload shorter.
Schedulers would take that into account, e.g., scheduling to be fair “per
innermost payload”.
That gets around the HOL blocking but doesn’t restrict us so badly as we
currently are - again, this is basically how Ethernet *already* handles things.
We COULD allow that to propagate up the stack, if we wanted to.
Joe
_______________________________________________
Int-area mailing list
Int-area@ietf.org
https://www.ietf.org/mailman/listinfo/int-area
