YOU ARE AT:Network InfrastructureThe rise of multicloud networking and Connectivity-as-a-Service (Reader Forum)

The rise of multicloud networking and Connectivity-as-a-Service (Reader Forum)

It’s hard to imagine an application today that isn’t reliant on connectivity to different clouds, multiple microservices and highly distributed users. To paraphrase the poet John Donne for our modern IT world — “No application is an island, entire of itself.” A simple migration of a stock-taking application from a data center to a cloud provider is likely to generate many new connectivity requirements from the single sign-on (SSO) system, to backend databases still in the private cloud, to connectivity of remote users and devices, such as barcode scanners and smart cameras. This layering of applications across multiple infrastructures — for example, where AI stock-item recognition on a remote camera is now part of the application too — creates a nightmare for the network and security teams responsible for managing infrastructure they do not own or control. This application connectivity paradigm, writ-large across our industry means most enterprises, regardless of their stated intent, are now operating in multiple private and public clouds and are caught on the back-foot without the tools or visibility to run their (now virtual) networks in a secure and reliable manner.

A 2021 study conducted by IDC on the state of multicloud networking for enterprises found 75% of enterprises are moving to edge, hybrid and multicloud networks. This current networking climate fosters dispersed application usage, a scattered user base, and sometimes, non-compliant data streams. Thus emerges two major challenges: enterprises losing control of their networks and falling behind in digital transformations for network infrastructure.

This shift isn’t lost on service providers who see opportunities in private 5G, in private cloud peering and in SD WAN services. According to TMForum research, 44% of CSP decision-makers believe B2B service revenue growth in the telecom industry will command a CAGR of greater than 10% over the next five years.

But these solutions miss a bigger challenge to network operators: How to deliver end-to-end network management and service assurance to customers without ownership, from the operator or customer of the entire infrastructure end-to-end?

Operators have much to offer in solving this challenge — the physical infrastructure at the edge, in 5G and in transport. However, these assets need to be coordinated in a management plane, in application visibility and through SDN to link seamlessly between clouds, users and sites. Doing so delivers an end-to-end application experience that is controlled, secured and managed on a second-by-second basis based on how each part of the infrastructure is performing. This management and control plane is at the center of an emerging technology area focused on Network-as-a-Service (NaaS) and Connectivity-as-a-Service (CaaS), and behind the explosive growth in SDN, projected to hit USD 37.71 Billion at a 21.24% CAGR by 2030.

Consider a car manufacturing factory floor. There is an abundance of robots, cameras, computers, phones, machines and devices all connected to the network. To ensure peak productivity, security and operability, the network must be functioning at its best. If a machine fails or the network buffers, the factory could face severe delays, leading to lower output and other implications. Thus emerges the need for Connectivity-as-a-Service (to ensure complete observability and security. Powered by AI, CaaS can predict outages and automatically service the device or network, resulting in smooth, consistent and continuous service. It also simplifies the back-end management supporting all the various applications powering the devices.

Does this mean that network operators need to become experts in manufacturing plants? No. Because while the operator remains responsible for connectivity of a holistic network, the customer and/or the application vendor is responsible for defining the experience and the virtual connectivity requirements of their applications, albeit in a way that now calls network resources dynamically.

So as enterprise network demands have become more diverse and complex, they require a new era of network service — both physical (think private 5G) and virtual (think overlay). By telcos shifting toward a Connectivity-as-a-service model, they can meet network and connectivity needs while simplifying operational challenges and creating a network that supports the needs of hyper-connected applications.

ABOUT AUTHOR