What we can learn from a massive ISP outage—and how SD-WAN can help


In late December, happy holidays became a headache for enterprises nationwide when a massive ISP outage disrupted Internet and telecommunications service—including 911 calls. The two-day outage affected non-customers, too, because every major ISP provides a backbone for other service providers. It wasn’t the first time a telecom carrier has experienced an outage, but this was arguably the most significant one.

 Whether or not your organization was among those affected by this dramatic outage, the incident should serve as a wake-up call—particularly if your organization uses cloud services. The fact is, a carrier outage can happen at any time for any number of reasons—a natural or man-made disaster, human error or, in this case, a bad networking card that ultimately brought down 15 data centers.

 Lessons learned? SD-WAN should become the solution of choice

 While it’s unlikely that we’ll see another nationwide service outage any time soon, some hard lessons became evident in this recent episode. The following are our take-aways.

  • Don’t assume your carrier has it covered. Carrier outages are common, although they’re usually limited to local or regional impact. You might assume that your ISP will quickly address the problem—but what if it can’t? A better approach is to take control proactively with a network strategy that supports the level of high availability and connectivity you need.

  • The multi-provider carrier strategy is your only real protection. If your organization relies upon a single telecommunications carrier, a significant outage uncovers the risks. You may have deployed MPLS routing through your ISP and created different physical connectivity paths across your region to protect against a fiber disconnect in one part of the network. However, those redundancies won’t help if the carrier’s entire network fails.  What does work is a multi-carrier strategy—which we consider to be a best practice to ensure physical redundancy and high availability. The catch? Creating a multi-carrier network, along with negotiating rates and managing contracts with multiple ISPs, is time-consuming, and you risk paying more than you need to.  

  • SD-WAN is the simplest strategy for creating a high-availability network. MPLS was designed for a world in which a company’s branch offices needed efficient data routes only to and from the corporate headquarters or the data center, not to and from a cloud service provider. SD-WAN provides many advantages, including centralized, streamlined management of policies and multiple ISPs, and it gives your geographically dispersed offices direct access to the cloud. If one carrier network fails, SD-WAN automatically re-routes traffic without missing a beat.

Like any major solution implementation, however, deploying SD-WAN is not without pitfalls—especially for organizations with large footprints or complex data needs. It can be more cost-effective to tap SD-WAN-as-a-service to take the headaches out of deployment and to move the investment from the cap-ex to the op-ex side of the ledger.

A chief advantage of SD-WAN-as-a-service is that your provider will already have relationships with multiple telecommunications providers around the world. A qualified SD-WAN service provider will be a circuit aggregator able to negotiate the most favorable rates on your behalf—and they’ll monitor service like a hawk to ensure that you’re getting the service you’re paying for. 

Ultimately, your best protection against connectivity disasters is having the right network architecture and design for the needs of your enterprise. SD-WAN isn’t your only option, but it’s rapidly becoming the solution of choice for organizations that need high-availability cloud access.

January 31, 2019