Hi David,

Respones to some of your questions inline:

> 1)       What's a good initial value for the cwnd, whether starting from
scratch or after a RTO?  I know RFC2581 says one thing, and 3465 says
another, and others say other things.  I'm curious what you actually use
and like, and why.

Linux has a value of 2. According to RFC 2414, you can set it to
4 x MSS. For media applications, it is better to set it to this value.

MSS=maxium segment size and is typically 536 or 1460 bytes (the number
might be off by some bytes but they are in this range)

> 2)       To what do you set the ssthresh in response to a RTO?  To a
fixed-fraction of pipe (say, 50%) or some fraction of something else?

ssthresh is set to half of the number of packets in flight. See this
paper on how Linux does it.
http://www.cs.helsinki.fi/research/iwtcp/papers/linuxtcp.pdf


> 3)       How far do you "fall back" the cwnd in response to a RTO?  To a
fixed value (say, 1472 bytes) or to some fraction of something else?

one MSS.


> 5)       How far do you "fall back" the cwnd in response to a RTO?  To
'ssthresh' or something else?

cwnd is set to one MSS inresponse to a timeout. It is set to half of the
number of packets in flight after a TDACK loss (or ssthresh which was
updated at the time of the TDACK loss).


Regards,
salman



On Sat, 21 Oct 2006, Wes Felter wrote:

> On Oct 20, 2006, at 6:22 PM, David Barrett wrote:
>
> > Following up on my earlier question, TCP has a bunch of “magic
> > numbers” whose best values vary depending upon the source you consult.
> Use the source; specifically Linux or your favorite BSD. You can
> probably grep the code for those RFC numbers.
>
> Wes Felter - [EMAIL PROTECTED] - http://felter.org/wesley/
>
> _______________________________________________
> p2p-hackers mailing list
> [email protected]
> http://lists.zooko.com/mailman/listinfo/p2p-hackers
>
_______________________________________________
p2p-hackers mailing list
[email protected]
http://lists.zooko.com/mailman/listinfo/p2p-hackers

Reply via email to