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.
Dino > On Dec 7, 2021, at 2:48 PM, to...@strayalpha.com wrote: > > On Dec 7, 2021, at 12:15 PM, Brian E Carpenter <brian.e.carpen...@gmail.com> > wrote: >> >> On 08-Dec-21 05:30, to...@strayalpha.com wrote: >> ... >>>> But you make another point which is pretty fundamental and foundational. >>>> Should data links be MTU-less, so to speak? And can they really do that. I >>>> won't hold my breath. >>> I don’t know yet, but I do know that’s what I *want* and why it’s different >>> than simply assuming a smaller MTU anywhere in the system. >> >> I think you'll need to discuss this point with the inventors of packet >> switching and queueing theory. Since the Internet is still basically a great >> big statistical multiplexer, getting rid of MTU at every layer seems like an >> impossibility. > > Perhaps, or maybe MTUs are an artifact of limited memory and BW. > > I’m not suggesting infinite packets - though that’s another interesting > exercise in protocol design (what if packets were infinite, what if protocols > used only one packet, etc.). I’m suggesting an environment in which > applications always assume some “MTU” and the headers are just “eaten” as > overhead throughout the rest of the system, sort of the way that Ethernet has > a 1500 fixed payload but no strict packet length limit (arguably, you can add > Q-tags or MAC-in-MAC all day long). > > Joe _______________________________________________ Int-area mailing list Int-area@ietf.org https://www.ietf.org/mailman/listinfo/int-area
