Enterprises have a love-hate relationship with the wide area network. They love the fact that they can now reach a global workforce and customer base with their important data and applications. But they hate the fact that congestion plagues long-haul Internet links, and the only way around the congestion is through expensive, slow-to-deploy services, such as MPLS (multi-protocol label switching).
The problem with MPLS and similar technologies is that they weren’t designed with today’s business challenges in mind. Today, a company may need to launch an overseas R&D office overnight, or it may acquire a startup and want to immediately network with offices in distant regions and countries.
Older technologies just don’t have the flexibility to do this on the fly.
At the same time, as more enterprise applications move out from behind the firewall and into the cloud, global enterprises need speedy connections to these often shifting service providers. Static networking architectures just can’t accommodate the rapid change that is becoming the norm in our knowledge-based, global economy.
From SDN to SD-WANs
Before we discuss why we need SD-WANs, we should talk about SDN (software-defined networking). Most analysts are bullish on SDN – and for good reason. According to IDC, the SDN market will reach $3.7 billion by 2016. SDN products are intended to create the flexible, agile networking capabilities that virtualization and the cloud have promised but have yet to fully deliver.
SDN decouples the control and forwarding functions from the underlying hardware. The result is that networking is simplified, vendor lock-in is (in theory) eliminated, and manual tasks can be automated by a network administrator to meet changing business conditions in near real-time. Moreover, new applications can emerge, ones not tied to hardware, such as optimization and security apps. These can be added and updated at will.
That’s the promise. The reality is rather different, though. As the SDN sector gains momentum, many SDN vendors are pitching forklift-upgrades to make everything work smoothly.
And you thought SDN was supposed to remove vendor lock-in?
But that’s not the biggest problem with SDN. For all of its promise in the data center, SDN quickly runs into a serious bottleneck: the WAN.
Efficiency in the datacenter is desirable, but it only gets you so far, and without optimizing the WAN, SDN will be a small-bore, incremental advancement.
That’s not a knock on SDN. SDN programmability within the data center has the potential to be a very big deal, but what global, cloud-dependent businesses need is the ability to extend that programmability all the way across the WAN.
And therein lies the problem.
The typical service provider today offers legacy services that are a poor fit for cloud-enabled, knowledge-based, global businesses. Let’s turn our attention back to MPLS, the traditional tool enterprises have used to overcome the distance problems of the Internet. While it’s an outdated technology, MPLS providers serve a $25 billion worldwide market, which, believe it or not, is still growing.
Why? Part of the reason is inertia. Businesses have signed contracts in the past, have relationships in place, and so it’s just easier to renew. You’re paying a premium, and you’re blocking yourself from future agility and flexibility, but if everything stays the same and you have the budget for it, you’ll be OK.
However, if your organization is embracing change, getting involved in M&A activities, expanding globally, moving more resources to the cloud, or even inking the kinds of partnerships that rely on data sharing, MPLS won’t serve you well. It’s too expensive, requires too much equipment, and if anything changes, such as adding a new location overseas, it’s far too slow. It could take months to get that new office online.
What the global enterprise really needs from service providers
What today’s global, agile enterprise really needs is a global network every bit as flexible as the businesses fueling the global economy. Enterprises need intelligence – and even security – baked into the network, rather than added on in an ad hoc fashion.
Enterprises require the ability to scale capacity up and down in an instant, so they can adjust capacity to meet fluctuating demand, rather than overpaying for capacity to be ready for those very occasional spikes.
The global enterprise needs something that can connect to and accelerate everything, from legacy on-premises apps to cloud-based services to applications in private clouds and everything else in between.
The global enterprise longs for the elimination of vendor lock-in once and for all, not simply the replacement of one set of locked-in vendors with another.
What the enterprise wants from vendors is what we all want: vendor reliance that is earned, not mandated, that is made possible only through competitive prices, a product or service that over-delivers on its promises, and outstanding customer service.
This is the promise of the SD-WAN: intelligence built into the network; the ability to deploy in minutes; the elimination of vendor lock-in; and the ability for anyone, anywhere to quickly access all enterprise resources, no matter where they are located.
And whatever Internet connections you have, you can use anything from MPLS and LTE to IP VPN connections to business-class broadband, and these can all come from different providers. These disparate links can be consolidated into one virtual WAN, allowing you to set policies and load-balance over various links.
Knowledge workers demand speed, reliability, and predictability. The old methods can’t deliver that. We need something new – and that something is the SD-WAN.