On Tue, Apr 24, 2001 at 10:11:15AM +0200, Paolo Castagna wrote:
> Hi,
> I'm an Italian student and I'm doing a Master Thesis on TCP rate
> control.
You have already posted this very same message to
linux-net@xxxxxxxxxxxxxxx
and:
netdev@xxxxxxxxxxx
lists. If you don't get reply from those (netdev mainly), the
linux-kernel is not going to yield cheers either.
People that know the networking code intimately are presently
"somewhat" busy, Be patient, repeat this topic at netdev
list after about a week.
....
> ---------------------------------------------------------------------
> About this algorithm, I've a problem... how can I measure the rate
> Ri for each TCP flow? I've found NeTraMet on this URL:
> http://www.auckland.ac.nz/net/Accounting/ntm.Release.note.html
> ... and I've also read the discussion about that on linux-kernel
> mailing list. Where is the best place to make a such thing?
> I've thought in /net/core/dev.c in function dev_queue_xmit.
> And, again, how can I associate the rate Ri to each TCP flow?
TCP Timestamps ?
(Which Linux does use if the other end supports them too.)
Of course, what is "rate" ? Units of something per units
of time ? Packets ? Payload bytes ? How does the size
of payload data in the packets affect the "rate" ?
...
> Greetings,
> Paolo Castagna.
/Matti Aarnio
|