Re: [tcpm] TCP tuning

[Joe Touch <[e-mail]>]

Mark Allman wrote:
> [These are re-sends from long ago... they never made it to the list.
>  But, that seems to have been worked out, now.  --allman]
> 
>> Making TCP smarter is fine; doing so at the expense of robustness is
>> not. 
> 
> I'll agree with this and push back at the same time.  I agree that
> robustness is important and should be protected.  However, I have
> increasingly come to view things like initial RTOs and initial cwnds,
> and etc. as not buying us much robustness.  What buys us robustness is
> the reaction---any reaction---to indications there might be problems.
> These gross notions are more important than any small parameter change
> it seems to me.

That may be true for initializations of parameters that have ongoing
tuning, but not all parameters have that property.

Further, the group behavior of changing these parameters is of concern
too. The following paper is a good example of "it won't hurt because we
tune", but it *can* hurt:

http://www.computer.org/portal/web/csdl/doi/10.1109/ICNP.1997.643715

Joe