Re: Congestion problems

[Paul E. McKenney <mckenney at itstd.sri.com>]

>By the way, Paul Tsuchiya, Joe Lawrence and I had quite a debate
>over the relative importance of trying to squelch traffic at
>the source when congestion is being experienced in the network
>supporting SMDS.  They have some pretty convincing arguments 
>that suggest that if you cannot MAKE the traffic sources behave,
>there is little benefit to attempting to localize the penalty
>(i.e., discarding) to the noisiest sources, and that the overhead
>and complexity required to put such a system in place may be
>greater than its worth.  So, it may be worthwhile revisiting our
>recent thinking.  I guess I made some assumptions that relied on
>some transport layer behaviors that wouldn't take effect without
>explicit feedback mechanisms, sorry.

If by ``localize the penalty to the noisiest sources'', you mean
that excess packets must always be dropped at or near the sources,
I don't necessarily disagree (but of course I reserve the right . . . :-).
On the other hand, if you mean that it is not worthwhile to penalize
uncooperative (possibly even malicious) users, I strongly disagree.
I would be happy to USMail you a preprint of a paper (``Stochastic
Fairness Queuing'') that I will be presenting at INFOCOM'90 which
describes an efficient method of allocating network resources in
a fair manner.  The paper assumes that all users have equal rights
to the resources, however, the method can be easily extended to
handle priorities.

				Thanx, Paul