Pekka Savola <[EMAIL PROTECTED]> wrote:
> On Tue, 13 Jan 2004, Fred Templin wrote:
> > 10Kbps seems the best value for LCD, given commonly-deployed
> > links that will see continued use into the future (see BCP
> > documents produced by the PILC working group).
>
> .. that is, restricting mechanisms based on the least common
> denominator of Internet does not cut it. As we have up-to 10 Gbit/s
> links in our network, why again we should limit our use of resources
> to 10 Kbit/s? :-)
> > 10Kbps seems the best value for LCD, given commonly-deployed
> > links that will see continued use into the future (see BCP
> > documents produced by the PILC working group).
>
> .. that is, restricting mechanisms based on the least common
> denominator of Internet does not cut it. As we have up-to 10 Gbit/s
> links in our network, why again we should limit our use of resources
> to 10 Kbit/s? :-)
No; not seeking to constrain high-speed links based on LCD at all.
But, the important point that you seem to be agreeing with is that
there will be a wide diversity of link bandwidths in the Internet both
now and into the future.
> One should note that such very constrained environments, like 10
> Kbit/s, have to restrict the traffic somehow in any case -- for
> example, provide the rate-limiting at the edge of such a link.
> One should note that such very constrained environments, like 10
> Kbit/s, have to restrict the traffic somehow in any case -- for
> example, provide the rate-limiting at the edge of such a link.
I must agree on this point as being well supported by the BCP
documents. The active queue management recommendation for
forwarding nodes in front of slow links seems particularly important.
Fred L. Templin
