There is a widespread assumption that cloud computing will one day reach maturity. But that belief overlooks the possibility that the cloud may not need to mature for it still to have a radical impact, says Lori MacVittie.
Some pundits and providers have dismissed concerns about the readiness of cloud computing for mass adoption, arguing that the technology is part of an evolutionary process. They say it will take time before providers have crossed all the Ts and dotted all the Is, and are in a position to roll out new services. After all, providers are responsible for important data and applications. Surely, we wouldn't want them to rush?
We should probably ask, then: what if the cloud never fully matures? But before we can address that question, we have to examine where the technology is right now and hence where it might be for the foreseeable future.
The definition of cloud computing that is least contentious is that provided by the National Institute of Standards and Technology (Nist), Information Technology Laboratory.
Nist distils cloud computing down to "a pay-per-use model for enabling available, convenient, on-demand network access to a shared pool of configurable computing resources — for example, networks, servers, storage, applications, services — that can be rapidly provisioned and released with minimal management effort or service provider interaction".
The important part of this definition is that computing resources includes infrastructure resources such as networks and storage. That's important because today, apart from separate storage and application offerings, very few cloud computing providers have gone beyond offering basic compute resources as a service.
A few have ventured into load balancing as a service, but almost none has taken the next steps to cloud computing maturity that require infrastructure as a service.
What's needed to move through the remaining stages of maturation is standardisation. First, we require the network standardisation that yields a services-based infrastructure and then the ultimate goal of cloud framework standardisation that yields true inter-cloud interoperability and portability.
But given the complexity involved in network standardisation — not just across cloud implementations but even just across a single infrastructure — it's hard to imagine that the cloud will mature past its current provision of compute standardisation and services.
What we have today is a flexible, accessible, nearly standardised method of rapidly provisioning and exploiting compute resources. Those compute resources can be used for a variety of purposes.
The idea that an organisation will live completely in the cloud is a fantasy for the moment — and probably always will be — not just because of the immaturity of cloud computing but because of practical concerns about control, regulatory compliance, security, performance and availability that surround not just the deployment model but also...
...the required reliance on providers that a complete cloud deployment would entail.
Today organisations are taking advantage of cloud computing across the entire application lifecycle based on specific resource requirements. They are capitalising on the relatively inexpensive and rapidly provisioned compute in the cloud to extend their datacentres and their ability to scale.
These facilities allow them to meet business and customer demands, and to decrease the operational costs of managing and maintaining hundreds of duplicated servers and systems.
They're using cloud computing as a utility, in a manner that makes sense given the technological and architectural state of computing and networking.
And that is leading to some challenges in networking, in application delivery and in management. Those challenges are leading to innovation in those problem areas, many of which have not radically changed since their inception.
New computing model
Vendors are rethinking systems and updating them to adapt to this new model of computing and of resources on-demand. These innovations may in fact result in systems that make the need to evolve further not only unnecessary, but impractical. Cloud computing as it exists today is forcing change in every corner of the network to adapt to the challenges of the new model. Those solutions are often viewed as stop-gaps that, once cloud matures, will no longer be needed.
But it may be the case that reality works the other way around: cloud computing maturation will no longer be needed precisely because of the innovation it engenders. If the cloud never matures, perhaps it's because it doesn't need to. It will simply change the very fabric of computing as we know it.
In fact it may be the case that the cloud revolution isn't about the cloud at all, but about the way it has pulled the rug from under our complacent view of networking and application delivery and is forcing that view to change, too.
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.