[e2e] traffic engineering considered harmful
DJamel H. Sadok
jamel at cin.ufpe.br
Thu Jun 14 05:41:23 PDT 2001
On Wed, 13 Jun 2001, Bob Braden wrote:
> *> One basic type of control we would like to see is a way for a custumer to
> *> measure network performance without having to deploy tons of application
> *> and/or network level processes at varios points in the network. Most of
> *> the time you only have control/access over your access point and don't
> *> even know what the egress node of your traffic is!
> *> I had a look at the RON approach and it is based on some
> *> kind of overlay network to get performance/routing information. Is there
> *> a simpler way of doing these things?
> A simpler way? How about if the "customer" (what we used to call a
> host, no doubt, but the world changes) sends a probe packet to the
> destination and measures its RTT. Or send a packet pair, etc., etc.
> You might even send some useful data on that packet. Unnnh, maybe
> we could invent TCP.
> Bob Braden
Ok. this maybe good for measuring simple things such as RTT, how can I
measure packet loss? actual bandwidth the ISP makes available to my
traffic? how much it is dropping, delay (and "NOT" RTT), etc? I don't
think that TCP does that!
More information about the end2end-interest