Burwood Group

View Original

Traditional WAN vs. SD-WAN: Here’s What You Need to Know

View this on-demand educational webinar with Kevin Rainey, Burwood Director of Architecture, to learn more about SD-WAN and see if it’s the right fit for your organization.

Are you facing issues with your current network infrastructure? Do you have a cloud migration strategy? If you haven't considered the benefits of a software-defined WAN solution (SD-WAN), now might be the time. Why? Because it’s delivered through the cloud, SD-WAN reduces capital expenditures and simplifies network operations.

Traditional WAN solutions often incorporate a hybrid of public and leased lines. MPLS, a traditional WAN connectivity method, is used so that that you can assign preference based on SLA guarantees, and you can determine timeliness for delivery of traffic. This becomes very important, particularly for voice and video-based applications. In addition, traditional WAN leverages internet-based backup as a secondary transport should the primary path become unavailable.

While traditional WAN can work well for some businesses, there are disadvantages to remaining in a traditional WAN architecture.

What are the disadvantages of traditional WAN architecture?

Depending on your business needs, a traditional WAN may hinder growth and agility requirements. Forward-thinking IT leaders should weigh these disadvantages against their strategic plans.

  • Failover (switching to a standby server or system when your primary application goes down) is completely dependent upon the state of the link (up/down). This means you must depend on routing protocol re-convergence. This can cause multiple seconds worth of outage, resulting in dropped or lost video or phone calls.

  • Configuration is distributed, meaning that configuration is housed locally on each individual router, but is typically templated.

  • New policies must be managed on a per-device basis and require your administrator to touch each individual device as policies change.

  • Significant time is needed for new site turn-up. This includes time for circuit delivery, equipment provisioning, and change management. New sites can take months to turn up due to the complexities of change management and project coordination.

What's all the buzz around SD-WAN?

SD-WAN is a shift in the way a Wide Area Network is deployed and managed. As its name implies, SD-WAN is a software-driven technology with application awareness that is managed from a centralized point in the network.

SD-WAN solutions are deployed as an overlay technology to an existing topology, which allows for ease of integration and adoption over time.

The advantages of SD-WAN include:

  • Lower circuit costs. SD-WAN technology allows for the elimination of costly MPLS circuits. It makes use of lower cost, higher-speed options for connectivity such as broadband or DSL. Of course, MPLS circuits can still be utilized based on circuit availability, additional SLA requirements, and ease of integration.

  • Adoption of cloud-based services. With traditional architecture, traffic typically routes back to the data centers with a need to have constant filtering in place. This results in sub-optimal routing and potential latency for cloud hosted providers like Office 365. SD-WAN moves us toward local internet offloading and getting user traffic closer to cloud services.

  • SD-WAN provides simplified QoS and prioritization of critical applications across the WAN. Rather than waiting for an up/down event, SD-WAN provides real-time traffic monitoring to divert business-critical traffic around brownout events.

  • Ability to support high-bandwidth intensive applications simultaneously, offloading them on local internet where applicable. These bandwidth requirements can be a challenge for traditional WAN.

How does SD-WAN actually work?

SD-WAN is an overlay to the existing network – essentially a network running on top of a network. This is accomplished by using tunneling technology to differentiate the physical network from the logical network.

SD-WAN implements a centralized controller, acting as a single pane of glass for managing the entire solution. It is used to set and maintain policy. This policy is used to control traffic paths, SLAs, failover, monitoring, etc. The SLAs are written based on application-specific information. Once policies are defined, they are pushed from that centralized controller to each SD-WAN node for more "hands-off" configuration.

Once the policy is pushed, it intelligently monitors the performance on a link, migrating traffic based on that particular SLA. At this point, we are no longer dependent on a circuit outage to swing traffic to a backup or redundant link.

Where do we go from here?

SD-WAN brings new agility and speed of provisioning to the WAN, while driving down the total cost of the WAN solution. As new technologies emerge, the capabilities and definitions around SD-WAN solutions will shift over time. At minimum, you should be aware of SD-WAN’s potential to positively impact you business needs. And if these benefits resonate with your current goals, consider starting with a SD-WAN proof-of-concept to evaluate the technology’s value to your organization.


January 17, 2019

See this gallery in the original post