Some have said that telcos will have a reduced role in the era of cloud computing, but Paul Gainham writes that service providers are the only players in the value chain that can ultimately shape the overall end-user experience. Co-sponsored feature: Juniper Networks 

Cloud computing will bring a need to coordinate network
resources from a common point. No single service layer
architecture will dominate how services are delivered

Paul Gainham: service providers with an established
business relationship with enterprises have a real
advantage in cloud computing

Cynic or optimist, hype or reality? Whenever a major new development is positioned and discussed within the IT industry, it has a tendency to polarise strong views either side of the debate with very little middle ground.

In that context, cloud computing is the latest major development to be debated within both the application and networking industries.

There are also just about as many definitions of cloud computing as there are opinions, so for context in this article, a definition that captures the holistic essence of this development is “a pool of highly scalable, abstracted infrastructure, capable of hosting end-customer applications, that is billed by consumption” — a definition from James Staten of Forrester Research.

From the service provider perspective a key point of such definitions is that cloud computing is more than the application, more than the data centre, more than the network: it should be viewed as a comprehensive experience-oriented offer to consumers and enterprises of any size alike, based on market need and opportunity.

That places the service provider in a potentially extremely advantageous position and could lead to cloud computing becoming a major, positive business inflection point for those providers that choose to view this as an opportunity rather than as a threat.

Cloud computing will be of growing interest to those enterprises and small and medium businesses which want to focus their corporate investment on core business activities and are seeking to reduce significantly both the capital investment and operational expense involved in running an IT operation for its employees.

The key view is that in many cases outsourcing elements of application development, deployment, hardware and support to the cloud will deliver significant business benefits, characterised by a new level of flexibility and agility previously unavailable from static in-house IT deployments.

Significantly, enterprises and small and medium businesses looking to outsource key business-critical applications will in most cases already have outsourced or will be looking to outsource the very network infrastructure those applications are running over.

This neatly summarises why, overall, those service providers with an established, well-developed business relationship with and diverse service offerings for enterprises have a real advantage to build upon with cloud computing.

When considering the positive opportunity that could develop for service providers, it’s also worth looking at why some actually regard cloud computing as a threat.

The early developments and deployments of cloud-based services have arguably seen most activity in the consumer space, with key content players offering a combination of platform, infrastructure and software as a service offerings to internet-based application houses.

These houses have used the flexibility and agility to deliver key application experiences that may require a huge amount of hardware and network capacity in short, sharp bursts.

A good example was January’s presidential inauguration ceremony for Barack Obama, which by some estimates generated in excess of 60-70 gigabits a second of internet video traffic for 60 minutes maximum.

Having to build network and application capacity to meet temporary and infrequent peaks like this is clearly uneconomical for individual enterprises. Renting that capability only when needed and treating it as a transactional business is far more cost-effective.

The development of internet-based offerings such as this, using over-the-top delivery to end customers, has led some to suggest that the only role for service providers in this scenario is that of a simple bit-pipe provider, providing raw capacity to content players and users alike and adding little else in terms of any added value.

The reality will be that there will be different models at play that service different parts of the customer base. But, interestingly, the service providers are the only players in the end-to-end value chain that can ultimately shape the overall end-user experience.

That experience is based on knowledge of who you are, where you are, what device you are using, what access to bandwidth you have, what bandwidth is required for the application and what security posture you require or need.

When combined with the ability to treat this development as a transactional business at scale the unique role of the service provider becomes clear.

These experience shaping take on extra significance as cloud computing develops, from a predominantly consumer-based over-the-top offer to a mainstream business critical service offering with demanding service level agreements as a prerequisite.

A number of studies from various industry watchers and analysts have repeated common themes when asking enterprise customers their opinions on the business appeal of cloud computing.

The positive perceptions typically revolve around reduced cost, speed of deployment and “pay for what you use”.

The top three concerns typically focus on security, performance and availability.

While it is true that many may well be thinking of application security, performance and availability when asked the question, it’s clear that when any application is delivered remotely, its security, performance and availability has to be considered end-to-end from the point of ingress to the point of usage.

This is why the ability that service providers have to shape experience potentially puts them right at the top table in terms of unique value-add in the delivery of cloud computing services.

Is today’s NGN up to the challenge?

So what does all this mean to the continued development and transformation of the IP next generation network as a flexible platform for innovative new service deployments?

It seems that there will be three business models at play within the cloud computing space:

• the continued delivery of over-the-top, internet-based services that are simply carried without any real value add by the service provider;
• an open garden model, based on customer demand, where the service provider partners with an off-net cloud content or application owner to enhance the experience of delivery to customers. This model may be most applicable to services such as software as a service or platform as a service where the service provider does not have the business desire or skill to become involved in the intricacies of the application, but can play a key part in delivering application-level SLAs as part of a partnership agreement;
• A walled garden model, where the provider owns the complete application infrastructure and network delivery. This may be more suited to “infrastructure as a service” opportunities where the service provider is responsible for the application delivery hardware, data centre and network and rents that capacity to third-party software houses or application delivery organisations or enterprises themselves.

Focusing on the open and walled garden business models, the potentially dynamic nature of cloud-based services and the concept of flexible demand and delivery may well trigger a key architectural transition in network design and thinking — the move from a predominantly static IP NGN to a much more dynamic and agile environment.

In this scenario resources will be demanded at scale from the network at an application level and the network has to not only deliver on those demands and meet end-to-end application-level SLAs but also deliver the ability to transact for those occurrences at the desired scale.

In many cases the underlying IP network has proven its ability to deliver a wide range of applications at varying levels of quality delivery, from best effort to assured.

This move from a predominantly static to a far more dynamic environment should be considered alongside the variety of applications the network is now being asked to carry — including IPTV, VPNs, IMS and cloud computing. At the same time there is a mix of walled garden, open garden and over-the-top business models.

This will drive a need to coordinate network resources from a common point. No single service layer architecture — IMS, Web 2.0, SOA or proprietary — will dominate how services are delivered.

This reality will drive three key inflection points in network design and philosophy.

• The application-oriented “micro SLA” will become a necessity for end-to-end service delivery. Where SLAs exist there is value, and where there is value there is revenue.
• A converged policy and identity capability will become a key requirement to help support the delivery of those micro SLAs and the key experience attributes detailed previously.
• Virtualisation of the NGN will become a necessity and drive significant flexibility and agility benefits in line with the changing application and service delivery landscape.

Cloud computing could well revolutionise the way applications are developed and delivered to end customers and become a significant opportunity for service providers who regard this whole development as an opportunity and not a threat.

After all, leaders and innovators define the future; followers simply live in it. GTB