Re: [re-ECN] Comments on "congestion exposure problem" document

[<[e-mail]>]

Hi Ingemar,

There is a reasonably recent paper from Kenjiro Cho et al. in Japan that
includes data that shows a marked pattern of peak and off-peak traffic
in residential networks. It is called 
"Observing Slow Crustal Movement in Residential Traffic" 
and was presented at CoNEXT last year. 

Is this the sort of reference that would convince you that the slow
network is down to traffic volume as much as server load?

1 comment inline

Toby

> -----Original Message-----
> From: Ingemar Johansson S [mailto:[e-mail]]
> Sent: 30 September 2009 07:54
> To: [e-mail]
> Cc: Moncaster,T,Toby,DER3 R; Ingemar Johansson S
> Subject: RE: Comments on "congestion exposure problem" document
> 
> Hi
> 
> I read through the latest "congestion exposure problem" document
> 
> A few comments:
> 
> Section 5.1, last sentence in first para: "The main effect of this has
> been that users now routinely see their network connections running
> slow
> in the evenings"
> From a laymans perspective I can probably agree with this, I
experience
> it every now than then when I "surf the internet" from home. I have
> however not been able to determine for sure if the problem is in the
> network or in the webservers. I would believe that a reference to
> scientific paper or mailing list discussion or whatever that describes
> this problem would be beneficial.
> 
> Section 6 (Use cases):
> Maybe obvious (and maybe does not fit into the use cases) but as I see
> it the future results of this WG would be to provide mainly the
> operators with at least one of the tools necessary to give e.g content
> providers and application developers a
> sustainable "carrot and whip" offer. "Make your traffic behave
> according
> to these rules and you're granted access to our network". I have no
> opinion regarding what kinds of traffic this would include, I guess
> time
> will tell, but the main task for the WG should be to provide with one
> important item to the "carrot and whip" toolbox.

Whilst this is indeed one possible outcome of this work, I am keen at
present to show this in an entirely neutral light until more people have
a better understanding of the impact of this new paradigm...

> 
> PS.. Resent as the first effort looked like crap on the list
> 
> Regards
> /Ingemar