Net neutrality: Will tiered access hit cloud?

Lori MacVittie explores what the exact impact of tiered-access models on a network level could be for cloud computing
Written by Lori MacVittie, Contributor

Organisations work at optimising web apps to make them as fast as possible in the cloud. The danger is tiered access at a network level could simply negate all those efforts, says Lori MacVittie.

Net neutrality is a touchy subject. It seems to bring out passionate views on both sides and makes it hard for those unfamiliar with the issue to grasp the implications.

So for the sake of brevity, I'm not going to argue for or against net neutrality. Instead, I want to look at the potential technical ramifications of putting tiered-access structures and other traffic-limiting services into place.

One thing is certain: a tiered network access structure would affect everyone from consumers to providers, including those desperately trying to stay out of the debate and simply do business on the net.

Impact on the cloud

The question then is what exactly the impact of a tiered-access structure might mean for cloud computing, the application owners who take advantage of cloud, and the consumers of those applications.

Could a tiered-access structure or traffic discrimination simply negate at a network level all the acceleration and optimisation techniques employed by application owners and providers to improve delivery performance?

Everyone today has a need for speed. We know that. We know that if customers are forced to wait even a few seconds for a response from a website, they'll simply go elsewhere.

So organisations have optimised and employed acceleration techniques, ranging from compression and caching to content-delivery networks, to ensure web applications are as fast as they can possibly be. Microseconds matter in the ether, and IT people today are well-versed in eradicating those problems that cause delays in application delivery.

Performance issues

The IT staff in the trenches are also well aware that once data leaves their network there's very little they can do to affect performance. Misbehaving routers and other intermediaries between the network perimeter and the client can easily offset all the optimisations and enhancements performed to ensure the best performance possible.

So it stands to reason that any technological impediments placed in the way will harm performance. It may become necessary for organisations to invest in web-application acceleration and optimisation simply to maintain a minimum performance level as set by their particular business.

This kind of tiered-access structure or traffic discrimination could also be the final straw that...

...drives WAN optimisation into the consumer domain. A soft client, after all, could easily be installed and exploited by organisations to provide some relief from intermediate impedances imposed by such implementations.

The vision of infrastructure as a service is a lofty one, and we're only on the first steps of a stairway to cloud that certainly won't be realised for quite some time.

Combining WAN optimisation and acceleration with web application optimisation and acceleration would probably provide an overall benefit to performance regardless of the tiering structure — providing more benefit for those in higher access tiers than those in lower, but benefits nonetheless.

The problem is that these options aren't readily available in cloud-computing environments today. The vision of infrastructure as a service is a lofty one, and we're only on the first steps of a stairway to cloud that certainly won't be realised for quite some time.

Aside from provisioning more resources to address volume and some performance issues caused by heavy load, there's nothing really available from cloud providers today to address the kind of performance impact that could be seen from a tiered-access structure.

Market demands for change

It's often the case that vendors, service providers and manufacturers will not and do not develop or expand offerings unless the market demands they do so.

But it might be that the impact on the bottom line for providers and organisations would be such that cloud-computing providers would be forced to turn into what can realistically be termed infrastructure-as-a-service, with a full line of service-based acceleration and optimisation offerings that allow customers to address performance issues raised by a tiered-access structures.

Or not. After all, if tiered access or traffic discrimination were part of the equation, organisations and providers alike always have the option of pointing out it's not their problem.

Technically they'd be correct. But merely pointing the finger at the guilty party would probably not satisfy customers who would quickly find alternatives that are not impeded by their providers.

Lori MacVittie is responsible for application services education and evangelism at application delivery firm F5 Networks. Her role includes producing technical materials and participating in community-based forums and industry standards organisations. MacVittie has extensive programming experience as an application architect, as well as in network and systems development and administration.

Editorial standards