Network-as-a-service (NaaS) is poised to make dramatic inroads into enterprise infrastructure, and why not? Once the network is defined as a virtual architecture, there is no reason to continue managing it as a piece of fixed infrastructure.
But just like there are differences in service-based compute, storage and applications, there are differences in NaaS that can make or break any given digital initiative.
According to HTF Market Intelligence, the global NaaS market is poised to top 50 percent growth per year through 2022, when it is expected to top $22.5 billion in revenue. This will comprise everything from subscription fees to technology development to expanding global connectivity, although HTF notes that growth of the overall market is actually tempered due to lingering concerns over privacy and security. Much of the activity will be centered in the Asian Pacific region, which is experiencing massive growth in telecommunications infrastructure and services across the board.
Still, it’s important not to misinterpret NaaS’ overall impact on the networking world by assuming it will become the de facto solution going forward. As Telus’ David Steele points out, one of the most pervasive myths about NaaS is that it is basically the same as SD-WAN, and by deploying one you automatically get the other. SD-WAN is essentially the virtualization of wide area networks so they can be centrally managed. But that management layer is not necessarily service-based, nor does it automatically allow network resources to be consumed as services. Also, NaaS does not guarantee top-notch quality of service, as this is more a function of the targeted policies that guide network consumption rather than the more flexible management model.
What is becoming clear is that there will be no shortage of NaaS solutions as the market evolves. An Israeli firm called Meta Networks is one of the latest entrants with a platform that is intended to replace multi-site VPNs with a single IPSec approach that provides always-on security and availability on a cloud-native backbone. The Meta NaaS system works with either managed or personal devices that can be tailored for either employee or partner/contractors regarding access, authentication and other policies. In this way, the enterprise is able to tap multiple cloud services without opening themselves up to attack vectors tied to each one.
So far, NaaS has been viewed as a way to manage connectivity in distributed, cloud-based architectures, but it has potential for on-premises environments as well. Big Switch recently released the Enterprise Virtual Private Cloud (EVPC) component to its Cloud-First Networking portfolio. It will give organizations the same service options regardless of whether the network resources are located in the data center or on the cloud. The idea is to give all infrastructure the same level of elasticity without having to set up multi-layered architectures that drive up costs and complexity. The system will be integrated into the company’s Big Cloud Fabric and Big Monitoring Fabric where it can take advantage of advanced analytics, data visualization and automation. (Disclosure: I provide content services to Big Switch.)
Like all as-a-service offerings, the real benefit of NaaS is that it replaces a fixed cost with a more flexible, consumption-based model that lets you pay only for what you use. This will become increasingly important as enterprise infrastructure expands into the cloud and all the way to the IoT edge, connecting all manner of endpoints on wired, wireless, virtual and physical infrastructure.
The need for fixed connectivity will not recede, of course, and in fact, it might actually expand. But a network defined as services will provide optimal support for everything else that is currently experiencing a service transition, which in the end will likely enhance all forms of traditional infrastructure.
Arthur Cole is a freelance journalist with more than 25 years’ experience covering enterprise IT, telecommunications and other high-tech industries.