proving once again that "peering ratios" only matter if the other guy's
customers can live without your "assymetric" content, here are two articles
i saw today via slashdot. what's interesting to me is whether bellsouth
will be sued some time later by some other content provider for de-peering
them without also having applied the same rules to google. note, this isn't
a bellsouth-specific rant, they just happen to be mentioned in today's story.
I refer to a previous post: Best effort is best effort, right? Ergo
setting special QOS for special people=worse QOS for notspecial
people. And who knew these content providers were getting free
bandwidth? Me, I thought they had to pay for their leased lines 
I'd say more, but I'll trigger swearfilters..
I'm sure the hardware vendors don't mind the prospect of wide-scale cycle-intensive QoS being deployed on large networks.
-david
This is the USA. They will be sued no matter what they do. They
will be sued even if they do nothing. The use of the ambigious
pronoun is deliberate. They are both big enough (both goliths,
no davids) that attorneys will target them both for real and
imagined vices.
Peering battles are a good thing, if annoying. Sometimes you can't
watch your favorite television station on your cable system, sometimes
you can't reach your favorite web site through your ISP. If you hide
the pain, you also hide the pressure to fix it.
For those unfamilar with history, insert Benjamin Franklin quote,
I would suggest reading the FCC Walker Report (1938) and the AT&T
Kingsbury Commitment (1913).
More like a preference/QoS war - peering has little to do with it. BS et al
will provide their customers a route to Google, Yahoo, etc - anything else
is economic suicide.
The big question is, can they convince the content players that they need
preferential transport. Anyone with a clue would say that things are working
just fine, and that the bits won't move any faster in an uncongested and
uncontested modern Internet backbone network.
The RBOCs need to get over this - they are floundering around to try and
find a way to recoup network costs. This is one front. IMS is another. I
feel their pain, but this battle has been lost. It has taken ten years, but
content has turned out to be king, at least as far as profit margins go. The
RBOCs are paying for their lack of vision.
Perhaps the RBOCs can do better with IPTV and take on the MSOs? Who knows,
but this effort to wring profit out of done deals is a sign of desperation
from companies that have lost the ability to innovate.
- Daniel Golding
Carriers trying to charge content-providers for access to their
network/customers is just part of a greater picture. The telco industry
is fighting to re-establish their dominant position. Traditionally
they've been able to pocket (extort) a large portion of the revenue for
3rd-party PSTN services (content services) themselves. Over the last
decade they've gained control of the ISP-industry and noe they want to
achieve the same level of control of the internet. The most conservative
are even suggesting to remove internet-governance from the public
domain. The European telecoms industry is openly urging the UN to take
control of ICANN's role. In the process they are trying to place the
functions of IANA and IETF in their belowed ITU. Their ultimate goal is
to eliminate IP as a product, to be able to sell access to sub-protocols
as individual services.
//per
Whatever. No-one's actually trying to do "some packets are more equal
than others" here in Europe, except for the mobile people with IMS and
such. BT just transferred its access network into a new division with
a specific remit to provide open access to all ISPs and alt-
tels who want it.
It's in the US that the RBOCs and cablesters are actually doing this.
My guess would be that basically everybody doing triple play will prioritize the IPTV and VoIP packets in their network including the access. Considering that streaming UDP IPTV requires very very low packet loss, much better than Best Effort, this is needed to provide a good quality service.
If you do LLQ you want to make sure you can control what goes into that class, that can be done several ways, including disallowing anything you don't know about (transit/ix) to go there.
This is preferential treatment for some packets and it makes perfect technological sense.
My guess would be that basically everybody doing triple play will
prioritize the IPTV and VoIP packets in their network including the
access. Considering that streaming UDP IPTV requires very very low
packet
loss, much better than Best Effort, this is needed to provide a good
quality service.
This is preferential treatment for some packets and it makes perfect
technological sense.
But it's no magic bullet. Streaming live media also requires low
jitter, especially if you are selling it as TV because viewers
will join and leave channels often, as they change channels on
their remote controls. This means you can't have big local buffers
to hide jitter, therefore you have to build a network with enough
capacity so that packets are all cut-through switched.
It's possible to hide packet loss from IPTV by throwing away
some other application's packets but you can't hide jitter
on your network. And if you have built such a good network
that you don't have jitter, there is not going to be any
packet loss either so QoS does nothing at all.
Preferential treatment can degrade service, but it cannot
improve service. If you prefer an IPTV service then you are
degrading all other services. If a 3rd party measures the
true quality of your service without using IPTV, then they
will see a network with much worse performance than on a
network which does not do preferential treatment.
No magic bullets.
And if you are spending the extra money to implement
preferential treatment, can you be sure that there is
a market willing to pay extra for this?
--Michael Dillon
[...]
But it's no magic bullet. Streaming live media also requires low
jitter, especially if you are selling it as TV because viewers will
join and leave channels often, as they change channels on their
remote controls. This means you can't have big local buffers to hide
jitter, therefore you have to build a network with enough capacity
so that packets are all cut-through switched.
I observe about 3-4 seconds of latency on the UK DVB-T and DAB
broadcasts anyway compared to analogue. Cost-cutting on CPU grunt in
decoder boxes can mean it takes up to ten seconds to change channel.
In contrast, streaming video and audio from iTMS starts to play a lot
quicker. It sounds like the problems with jitter and latency over
private IP networks is overstated if it still works fine over the
Internet.
(FWIW, this is on 1Mb/s ADSL that is 170ms from www.apple.com.)
Mike, can I make:
Preferential treatment can degrade service, but it cannot
improve service.
my motto?
> Whatever. No-one's actually trying to do "some packets are more equal
> than others" here in Europe, except for the mobile people with IMS and
> such. BT just transferred its access network into a new division with a
> specific remit to provide open access to all ISPs and alt- tels who want
> it.
I'm sorry if I made the impression that it is already happening. Now
it's a game on the political arena, and it's important to support the
RIR-communities' efforts to provide balanced information to
decision-makers.
My guess would be that basically everybody doing triple play will
prioritize the IPTV and VoIP packets in their network including the
access. Considering that streaming UDP IPTV requires very very low packet
loss, much better than Best Effort, this is needed to provide a good
quality service.If you do LLQ you want to make sure you can control what goes into that
class, that can be done several ways, including disallowing anything
you don't know about (transit/ix) to go there.This is preferential treatment for some packets and it makes perfect
technological sense.
Preferential treatment of value-added services in the providers own
network is just fine. It's down-prioritizing competing services that may
become a problem. Like blocking all VoIP traffic not using the
providers' own "gateway-service".
//per
[...]
But it's no magic bullet. Streaming live media also requires low
jitter, especially if you are selling it as TV because viewers will
join and leave channels often, as they change channels on their
remote controls. This means you can't have big local buffers to hide
jitter, therefore you have to build a network with enough capacity
so that packets are all cut-through switched.I observe about 3-4 seconds of latency on the UK DVB-T and DAB
broadcasts anyway compared to analogue. Cost-cutting on CPU grunt in
decoder boxes can mean it takes up to ten seconds to change channel.
<AOL>
Here in the US, Comcast's "digital cable" service takes seconds to show a picture after you change channels. I don't know if that's buffering or CPU or what, but consumers are clearly OK with it.
</AOL>
So you _can_ have a large client-side buffer and ignore jitter. That means packet loss is important, not jitter. (A 2 second buffer would be orders of magnitude more than your typical jitter.) Which means queue size is only relevant if you drop things off the back end of the queue.
Which means you can build an intentionally congested network and "sell" the front-end of the queue to services which will pay you more. The rest will just risk being dropped off the end of the queue.
Will consumers care? Hell, they're already used to the Internet not really working right, rebooting their computers every day, and sites being taken down 'cause the next box over is infected and DDoS'ing someone (or their domain has been removed for spamming :). In fact, most consumers probably can't use the speed they have since their computer is using all the available bandwidth & CPU spewing crap onto the 'Net from the 1389 viruses installed.
So, yeah, I think the end user will put up with the fact some sites are slower on their DSL line and not look to change providers. And they will slowly migrate to the faster sites - i.e. the ones who pay for the front of the queue.
Also, no one has talked about the ideas proposed in Vixie's second link: That the big content providers are willing to pay a 'little' to raise the bar of entry. A few million bux a year to each of the RBOCs in the US would be a rounding error in Google's bottom line, but it would make it nearly impossible for a 'start-up' to make it.
Doesn't that scare anyone?
In contrast, streaming video and audio from iTMS starts to play a lot
quicker. It sounds like the problems with jitter and latency over
private IP networks is overstated if it still works fine over the
Internet.(FWIW, this is on 1Mb/s ADSL that is 170ms from www.apple.com.)
Yeah, but you don't get iTMS stuff from www.apple.com. I'm betting you are a LOT closer to iTMS. 
Perhaps this additional networking complexity (and hence cost, at some level, presumably) will allow peoples' eyes to be opened to the fact that the majority of television being viewed over the Internet today is done asynchronously, through peer-to-peer, file-sharing networks.
It amuses me to think of early-adopting consumers receiving all their expensive, network-optimised television shows in real-time on their TiVOs, only to have them recorded to disk and watched days later. (Recorded onto hard disks with no DRM, no less, ready to be encoded and uploaded to eDonkey
If content distribution companies would accept this as the final outcome, then sticking a torrent client on the set-top-box and feeding it from an RSS feed starts to seem a lot cheaper than encumbering every access network with traffic shaping.
Joe
Agreed - mostly.
Things like sports events will still require real-time feeds, and people will pay for them. But satellite seems like a perfectly reasonable and cost-efficient means of distribution without going through anyone's right-of-way.
I mean, seriously, do you think anyone is going to WAIT to see Victoria's Secret Fashion Show?
That and breaking news seem like reasonable exceptions to point out in contrast to my rampant generalisations.
For news, however, stories seem to break on the web long before they usually reach the television. Anybody who really wants to hear about things as they happen are probably best to avoid the traditional news networks anyway.
As far as sports go, there is no timely coverage of rugby in North America anyway, I can't imagine why anybody would waste their time watching inferior games like football, hockey, baseball or basketball at all, never mind in real time.
Joe (running away quickly now)
Things like sports events will still require real-time feeds, and people will pay for them.
That and breaking news seem like reasonable exceptions to point out in contrast to my rampant generalisations.
I think we are in very close agreement here.
Although you bring up a good point. At least here in the US, there is the "emergency broadcast system", a way to break into the TV feed in "real time" in case of emergency. It was designed because, well, us dumb americans are glued to the boob tube 24/7, so what better way to say "GET THE HELL OUT NOW!"?
Things like "breaking in" to TV feeds are not really useful if everything is pre-recorded and stored locally.
For news, however, stories seem to break on the web long before they usually reach the television. Anybody who really wants to hear about things as they happen are probably best to avoid the traditional news networks anyway.
As far as sports go, there is no timely coverage of rugby in North America anyway, I can't imagine why anybody would waste their time watching inferior games like football, hockey, baseball or basketball at all, never mind in real time.
I didn't say they were BRIGHT or TASTEFUL, just that people would pay for it.
Hell, people use Pay-Per-View for WWE, even after they admitted it was staged. No one has ever gone broke underestimating the US public....
Then again, I like US "football".
Joe (running away quickly now)
As you should. We might not be smart, but we can kick Canada's ass!
If something like the slingbox catches on....
www.slingmedia.com
From the sling community forum:
Hello before yall get to excited about verizon it looks like they are cancelling users who use too much bandwith.
" Unlimited NationalAccess/BroadbandAccess services cannot be used (1) for uploading, downloading or streaming of movies, music or games, (2) with server devices or with host computer applications, including, but not limited to, Web camera posts or broadcasts, automatic data feeds, Voice over IP (VoIP), automated machine-to-machine connections, or peer-to-peer (P2P) file sharing, or (3) as a substitute or backup for private lines or dedicated data connections. "
Might as well stick with 56K dialup at that point....