Adrian,
In message <[email protected]>, "Adrian J. Hook
e" writes:
> Two concerns need to be voiced about the direction in which the TCPSAT
[...]
Here's an idea FWIW.
For the special case of NASA satellite work, might it be possible to
build on the TCP-ECN (I guess it is IP-ECN now) work of Sally Floyd.
The current I-D defines a two bit field where 00 is ECN clueless, 01
is ECN capable, and 11 is ECN capable and congestion experienced (CE).
If we grab the one unused value, we might be able to accomodate
satellite.  If you plan to ignore loss set the bits to 01 rather than
10.  The internet-draft would have to indicate that a router should
implement the ability to set CE or throw away all data with the field
set to 01.  This still doesn't do much for existing routers unless it
is possible to purge 01 packets at the edges of an IP provider.  That
might be sufficient to get agreement on putting this in the IP-ECN
draft despite the existing router problem.  Indicating CE would have
to be redefined as 01->11 as well as 10->11 (the former CE bit is
already set in the 01 case).
This would prevent traffic which was stating that it will ignore loss
from straying into parts of the Internet that didn't know how to deal
with it (ie: couldn't yet set the CE bit).
> Instead of pushing the problem off to the global end-user community,
> wouldn't a better solution be to handle the problem locally - where it
> occurs, i.e., over the satellite link - via a transparent gateway approach
> (at least in the near term)?
That's one of the recommendations - proxy servers aka application
gateways.  The IP-ECN bit value of 01 might be of use between such
servers.  Ignore loss, just use the IP-ECN CE indication.
> Furthermore, we have to deal with other sources of intermittent data outage
Isn't that why we have routing protocols?  :-)
Would grabbing the 01 bit value and defining it this way help?
Curtis
ps- Please excuse the bit reversal if the unused is 10.  The idea is
the same.
This archive was generated by hypermail 2b29 : Mon Feb 14 2000 - 16:14:41 EST