Blog - Cloud ComputingCommentary from our cloud experts around the globeCommentary from our cloud experts around the globePUTINGCOMCLOUDBLOG

June 20, 2017
Network capabilities that you need, when you need them
By: Larry Socher

Today, companies have a new option for consuming network capabilities—Network as-a-Service. It’s an important trend, although it will involve a fair amount of retraining and likely redefine IT’s role in the business.

In a traditional networking environment (see my earlier blogs: Your network and the journey to cloud, How do solve the biggest problems of a dated network), a company buys a tightly wound network appliance where the software and hardware are embedded together. Today, a different option is available: leasing network capacity and having it managed through an as-a-Service model that’s priced by the device or by the port.

An as-a-Service model offers multiple advantages:

  1. Scalable services: Pay only for what you need

  2. Speed to value: Tap into world-class capabilities in a matter of days or weeks, not months

  3. Continuous improvement and innovation: Leverage a provider highly incented to provide the latest technologies and solutions

Network as-a-Service 2.0

At the same time, network technologies and capabilities are evolving rapidly, so companies need to make sure they’re tapping into an as-a-Service model that adapts to change. One effort Accenture is helping to lead, along with our service provider partners, is implementing Network as-a-Service 1.0 while also pushing toward version 2.0.

Networks are becoming much more software defined and virtualized. That will require new commercial models reflecting management complexity and policy. Network as-a-Service today is mostly grounded in devices and ports. By contrast, Network as-a-Service 2.0 will embrace network function virtualization. Although it will still need to factor in the hardware and bandwidth costs, pricing will be driven by the number of virtualized functions and by management complexity—usually reflected in the number of policies being provisioned.

The whole integrated story: Better cost efficiencies and service levels

In other words, we’re evolving Network as-a-Service to operate in a much more flexible environment—a combination of more virtualized functions, modularized using the newest technologies like virtual CPE and SDWAN running on standardized compute. Network as-a-Service 2.0 will also leverage the latest operational tools—the analytics and the automation to get to much more predictive assurance.a

This approach tells a more holistic, integrated story. It's about network transformation and operational transformation—leveraging standard compute, network virtualization and software-defined networks—to achieve better cost efficiencies and service levels.

Taking practical steps toward Network as-a-Service

Many times, when companies tackle network transformation projects, they pick one or two parts of the network such as:

  • Building out the data center or private cloud

  • Interconnecting to the public cloud

  • Overhauling the site networks with managed WiFi and SDWAN

Although it’s possible to drive transformation in those ways—focusing on different parts of the network—you ultimately need an end-to-end view of how this is all eventually going to come together. For example, don't look at your SDN solution in isolation. Look at how it will integrate with WiFi and then how it’s going to be integrated with your private cloud.

Similarly, make sure that, whatever product you pick for SDWAN, you can integrate the policies that authenticate and identify devices and users, classify traffic, mark that traffic with the right quality of service and security, and put it on the appropriate SDN circuit.

Bottom line: Don't box yourself into a corner with a solution. Make sure it will ultimately integrate with your end-to-end vision and evolving architecture.

Popular Tags

    More blogs on this topic