Wide Area Networks (WAN) are one of the most durable, longest-standing technologies in the IT world, but the era of WAN as a foundation for corporate computing is nearing an end because of a combination of factors that are quickly making this venerable architecture obsolete. I say this with more than a hint of nostalgia because WANs have been a cornerstone of corporate IT strategies for decades. It pre-dates the internet. It even pre-dates many people reading this article. However, everything in IT eventually reaches the end of the road, and that is even true for technology with the longevity of WAN.
There is no single culprit for what is causing WANs to reach obsolescence, but there are some key factors that are hastening its demise:
• The decentralized geography of company’s operations today
• The rise of cloud-based applications
• Difficulty of connecting branch offices and remote locations via traditional WANs
• Latency and downtime issues that don’t mesh well with today’s applications
• The accelerating speed of change in companies’ IT needs
• The high cost of building and maintaining WANs relative to newer technologies
• And perhaps most importantly, higher expectations from customers who want more from their WAN infrastructure than the technology can provide
“To extend the lifespan of their WANs, most companies have been taking a duct tape approach”
To extend the lifespan of their WANs, most companies have been taking a duct tape approach to shoring up these shortcomings for years, including ad hoc steps like combining MPLS with dedicated Internet access with IPSEC, and/or DSL, Wave Services, Ethernet, etc. As an example, a mid-market company might keep a dedicated SLA with a primary core network connection but then adopt a smaller set of dedicated bandwidth solutions complemented with 4G, broadband, etc. As a result, some branch offices and remote sites might rely completely on broadband connectivity with no dedicated core transport.
It is no wonder, then, that IDC estimates that as many as 60percentof the U.S.-based companies plan to migrate from a traditional WAN to an SD-WAN solution in the next 18 months as businesses continue to move mission-critical data into the cloud and as employees rely more and more on a combination of public and private connections to do their jobs. Mid-market companies are feeling the pressure of WAN’s growing obsolescence in a particularly acute way because they do not have the deeper pockets and resources of large enterprises to support the endless series of duct tape fixes needed to keep a WAN propped up. Mid-market companies need a forward-looking strategy for a post-WAN world, which is what is driving such strong interest in NFV (Network Functions Virtualization), SDN (Software-Defined Networking) and SD-WAN (Software-Defined Wide Area Networks).
One of the most common misconceptions I hear is when folks talk about it as if SD-WAN is one single thing that is the same for Company A as it would be for Company B. The reality is that SD-WAN installations are never one-size-fits-all..
For mid-market companies in particular, the best solution is likely to be an SD-WAN implementation that is a hybrid architecture that dynamically routes traffic across distributed branch and remote locations via software intelligence—irrespective of the underlying connection type, access point, or carrier. The most common approach my team uses with this size customer is an SD-WAN overlay network solution that uses the principles of SDN technology to decouple network intelligence and configuration from physical connections and hardware (via an abstracted layer). This creates a centrally managed virtual WAN that connects distributed branches and remote locations irrespective of connection type, access point or carrier.
This is very different from the types of SD-WAN solutions implemented by large enterprises (think Fortune 500 companies), but that is exactly the point. Mid-market customers have very different technical needs, legacy infrastructure and budgets than their much larger cousins, and a hybrid implementation tends to be much more in tune with these needs. It helps companies quickly converge voice, data, and video applications on the same IP backbone in a way that is cost-effective and easy to manage. It virtually plugs into existing legacy hardware so that the transition is quick, and it provides the fastest path to future-readiness that keeps up with evolution of the cloud and evolution of applications.
Here are a few more pieces of advice for mid-market companies looking to move from WAN to SD-WAN. I will start with a tip that relates to something I hear repeatedly from companies moving ahead with this project: moving from the ad hoc management of an aging WAN infrastructure to the strategic planning process for SD-WAN can give a CIO whiplash. It is critically important to be strategic and to design SD-WAN not just for the moment but with the future in mind, and that is dramatically different from the day-to-day challenges that WANs pose to them. That is a big change for CIOs, but it can be a very welcome one because so many CIOs spend so much of their time putting out fires and responding to immediate technical challenges. Every CIO I know wants more opportunities to step back and be more strategic and SD-WAN transitions are a great opportunity to do exactly that. CIOs should be prepared for that abrupt shift from putting out fires to being strategic, and embrace it as a chance to spend time on the big picture.
Specifically, I suggest using multiple access technologies such as Fixed Wireless, 4G and fiber so that diversity is built into the SD-WAN from the beginning to support growth, redundancy, disaster recovery and more. Since SD-WANs are not static, I also recommend that companies ask many questions of their providers to determine how committed they are to investments not just today but in the future. One of the most compelling things about SD-WAN is its adaptability, which can flex, adjust, and grow in ways that WANs ultimately could no longer do. Having a provider who will partner with you for the long term and who will combine a set of services and applications that evolve over time is critical. For example, some providers will suggest a pure OTT solution as the solution today, but that may not hold up as a long-term solution. In the same way that CIOs need to be strategic and think about the future of their SD-WAN, they need a provider who is willing to be just as strategic and forward-looking.
The best strategy is always first to have a strategy, and companies that are willing to look with clear eyes at the limitations of their existing WAN know that it is not a technology that will last forever. They need a strategy for a post-WAN world, and for mid-market companies the most sensible solution may be a hybrid SD-WAN that gives them the performance, flexibility, network efficiency and other technical specifications they need at a cost they can afford.