symmetric vs. asymmetric [was: Verizon Policy Statement on Net Neutrality]

Michael Thomas mike at mtcc.com
Sat Feb 28 17:23:06 UTC 2015


On 02/28/2015 08:59 AM, Mike Hammett wrote:
> 20 years ago was into AOL's prime, so yes they did.
>
> Great, let's re-evaluate the system when demand necessitates it. For many systems, it's literally as simple as changing how many channels are allocated to what directions.
>
> By that logic, we would have been running 486s with 32 gigs of RAM because some people today use that much. *shakes head* Obviously the majority of the dissent here works with OPM.
>
>

The point is that the incumbents (= telephants) at the time looked at 
even the
minuscule AOL user base with disdain saying that their market share was 
irrelevant.
Even into the early 2000's these same guys thought that voice was the 
only thing
that really mattered because the new fangled internet users were 
outliers from their
pots bread and butter. We now know those outliers were important. Being 
dismissive
of them is dangerous.

I think at this point, it's really not too much to ask for PHY's that 
can deliver decent
upstream rates on demand to deal with the bursty nature of upstream 
traffic from
eyeball networks. Nor is it too much to ask for l2/l3 shaping to deal 
with the internet
equivalent of a synchronized toilet flush.

 From the consumer standpoint, I *really* don't think it's too much to 
ask that when
I have the occasional 10 gig image to upload that it takes me << than a 
full day. This
has nothing really to do with symmetry, per se. It's the need to adapt 
to what the
traffic is *actually* doing at peak times, regardless of the average 
up/down byte count.

Mike



More information about the NANOG mailing list