read Who Confirms The Accuracy (or is it precision) Of ISP Usage [sic] Meters?

[Dave Farber <dave () farber net>]

Begin forwarded message:

> From: "Kahn, Kevin" <kevin.kahn () intel com>
> Date: January 14, 2010 11:06:30 AM EST
> To: "dave () farber net" <dave () farber net>
> Subject: RE: [IP] re Who Confirms The Accuracy (or is it precision)  
> Of ISP Usage [sic] Meters?

> > The deeper issue is our willingness to accept the bad metaphor  
> that leads us to think we are using up the "Internet" as if we were  
> consuming electricity.
>
> Bob’s wrong – if you’re using bandwidth on a path, I can’t  
> use it on the same path. While we’re not talking about a finite natu 
> ral resource, like oil, or similar, we’re talking about what I’ll  
> characterize as financially limited finite resource: bandwidth.
>
>
>
>
>
> At least on this list inhabited by folks who actually do generally  
> know something about the technology could we manage to be precise in  
> our use of “bandwidth” (a measure of instantaneous transmission  
> rate) and “ transferred  bit quantity”.  Except in the case of a  
> constant load over the entire measured period, these are very differ 
> ent things.  The so-called “bandwidth caps” are for the most part  
> not that at all. They limit the total amount a user can transfer ove 
> r some period (generally it seems a month).  I understand that there 
>  is likely a correlation between people who transfer a lot of data o 
> ver a month and those that try to run a high instantaneous rate, but 
>  it is precisely the difference that makes capacity caps unlikely to 
>  be the right tool.  There may be two problems that ISPs have.  The  
> most common is what capacity they provision in their networks and ca 
> pacity caps only address this incidentally via the correlation above 
> .  For handling fairness on an immediate basis they need not caps bu 
> t rather better scheduling algorithms for the traffic that penalize  
> proportionally to the “excess” traffic that users are generating  
> right then.  The other problem may be one of costs related to dispro 
> portionate total traffic exchange (which could justify a capacity ca 
> p) but this one doesn’t get spoken of much.  Placing a capacity cap  
> on a user who goes out of her way to heavily use the service only in 
>  off peak hours does nothing to help problem one.  Conversely, a use 
> r who stays well under the capacity cap but puts all that load in th 
> e peak hour or two every day is impacting the system considerably.   
> Independent of any of our beliefs about what ISPs should or shouldn’ 
> t do, or how neutrally they may do it, could we at least be precise  
> about what we are talking about on this issue?
>
> ____________________________________________________________
> Kevin C. Kahn
>
> Intel Senior Fellow, Director Communications Technology
>
> Intel Labs
>
>
>
> JF2-96
>
> 2111 NE 25th Ave
>
> Hillsboro, OR 97124-5961



-------------------------------------------
Archives: https://www.listbox.com/member/archive/247/=now
RSS Feed: https://www.listbox.com/member/archive/rss/247/
Powered by Listbox: http://www.listbox.com