Yoda said “Do, or do not. There is no ‘try.'” He didn’t run a site on the internet, which is and always will be a “best effort network” with no guarantees of availability and performance. It is up to the customer to optimize these aspects the best they can for their budget. I, Cringely posts a great article on the struggle for network neutrality – effectively the control of the internet, which is a big battleground in the fight for freedom of information. This battle has been heating up with politicians, service providers and network vendors pushing for stricter control over service delivery allowing a form of favoritism called prioritization (a technique developed for private networks) to be implemented on the public internet.
I asked Bob Kahn, the father of TCP/IP, and he made the point that the Internet is a Best Effort network and if you change that, well, you no longer have the Internet.
Why is nobody –- NOBODY -– mentioning Moore’s Law in this discussion? The Internet is today 1,000 times the size it was in 1996, yet these discussions tend to view network growth as static.
The real question is whether the Telecommunications Act of 1996 even needs re-writing and that comes down to something Congress probably doesn’t want to consider –- that they aren’t all-powerful after all. The Internet has grown beyond Congress’ ability to control it. So they can bow to special interests, change the law, and ultimately prove their own impotence. Or they can do nothing and fight again another day.
I think his point about the dymanic nature of the internet is excellent. Do we really need to address current performance problems when cheap 10GE is here and 100GE around the corner? How do you control such a complex, dynamic system for any appreciable length of time before it adapts to your controls? Obviously you would need to create rules (policies) for what actions are permitted, which ones are favored over others, a process which is mainly and rightly left up to the network security professionals working for end users (and increasingly home users). The current internet basically leaves it up to the end users to create and manage their own policies, with some exceptions (typically laws related to harming others) which are built into the acceptable use policies of ISPs. Basically it is an “enduser beware” system offering enormous flexibility and unprecendented access to services and information.
Increasingly, policy makers and ISPs are realizing how powerful they could be if they managed to enforce additional policies related to, for example, voice traffic or file sharing. Unfortunately the historically network neutral network vendors (such as Cisco, Juniper etc) are realizing they can help create a new market for far more complicated network devices which can play a part in a tightly regulated and commericalized internet. It used to be if you needed to enforce policies you turned to security vendors, but now all the big network companies have bought security companies and of course they are trying to move up the value stack since network equipment is effectively commoditized. Routers and switches are dirt cheap and there is no margin to be made in this segment.
So to come back to the question at hand: what do we get when ISPs start controlling our use of applications (they’re already started by the way – blocking certain ports and rate limiting P2P traffic) and start to favour their own applications (or their partners’ applications) over the ones you really want to run? Well the problem comes down to choice. If I have a choice about what network to be on and the extent to which I will be regulated that might be ok. But partial control is not what the stakeholders are envisioning.
This is a political play by lobby groups and a profit motivated move by network vendors and telecom. It is contrary to the philosophy in which the internet was invented and grew. It is a play for power and control. My message to them: stop taking the internet for granted: the public, taxpayer-funded standards-guided internet that allowed you to make a fortune in the first place is being jeopardized and could fragment into a collection of private networks. If too much control and regulation appears on our internet, a new public internet will emerge from the ashes of the original and it won’t use any of your name brand equipment.