[e2e] ICMP & TCP segments with IP ID = 0?
raj at cup.hp.com
Wed May 16 11:44:30 PDT 2001
i have to say that I agree with the conservative in what you send camp,
even though the IP ID has become essentially useless (not _completely_
but essentially :) since for nearly 10 years it has been possible to
wrap the IP ID space with fragmented traffic well within most reasonable
values for a fragment reassembly timeout.
if we really want to charge off, i'd point-out that the classic 32-bit
TCP sequence space can also be fairly easily wrapped within MSL or 2MSL
now. we have systems and NICs capable of 100MB/s transfer, which means
that a 4096 MB sequence space is wrapped in less than a minute, with 10
Gbit nics just around the corner as it were. within a couple years it
will be possible to wrap the classic TCP space in less than four
so, perhaps we should be _requiring_ timespamp options now? and starting
to work on 128 bit sequence numbers perhaps ?-)
these opinions are mine, all mine; HP might not want them anyway... :)
feel free to email, OR post, but please do NOT do BOTH...
my email address is raj in the cup.hp.com domain...
More information about the end2end-interest