SD-WAN - A Quick 101
Updated: Mar 4
What follows is a general summary of the principles at work. If you already understand how SD-WAN works and the typical use cases and want to know about the technology and provider choices out there, take a look at our provider overview here.
How does SD-WAN work?
SD-WAN evaluates network traffic patterns and chooses the most efficient route across the network in real time. It can combine multiple lower-cost networks that have variable performance characteristics, and even combine them with more expensive networks like MPLS to achieve even better performing connectivity than an individual network alone.
For example, if broadband is performing better than MPLS in that particular moment, it will choose the broadband connection. It also has the ability to prioritise traffic based on user-defined rules, ensuring the most important or network-sensitive traffic is handled accordingly.
SD-WAN also enhances network efficiency by leveraging some of the most important characteristics of the cloud, namely the many-to-many connections possible with the internet, instead of transmitting data from point-to-point in predefined networks utilising the old hub-and-spoke or star models, thereby adding to latency and cost.
How does SD-WAN support cloud-based applications?
To further enhance cloud-based application performance, some SD-WAN providers also have direct connections into the most popular cloud-based applications and global datacentre providers, enabling customers to leverage their networks’ SD-WAN edge deployments.
Other SD-WAN providers have their own backbone network, with dedicated high performing network connections regionally or around the globe, providing faster and potentially more reliable connectivity to countries where internet latency can often be a problem.
The ability of SD-WAN to dynamically make a decision on where to route traffic makes it an ideal solution for higher reliability network requirements. Multiple service providers can be used in an all-active configuration in order to minimise the impact of an outage.
Pre-SD-WAN technology was not very friendly to managing multiple simultaneous networks, at times requiring backup networks to remain dormant awaiting an outage, or taking too long to transition and reroute traffic, which would drop calls and other network sensitive applications. SD-WAN allows for the simultaneous use of the backup network, achieving more combined bandwidth availability and higher performance than any single network due to dynamic routing.
SD-WAN in Transition
The ability to simultaneously handle multiple networks makes SD-WAN a useful network transition tool. One can implement SD-WAN in advance of the expiration of an MPLS contract for example, still leveraging the MPLS connectivity for the combined SD-WAN network, and then decide to add different networks, or supplement the MPLS network without needing to decommission and transition off the MPLS WAN immediately.
In some cases, it’s been used to test the performance and suitability of DIAs versus an MPLS tail in remote sites. SD-WAN providers often speak of this as being a proof point, and of course talk freely about customer installations where the MPLS has not been renewed because a DIA (or multiple diverse DIAs) worked as well as if not better than their traditional corporate WAN.
What about Security?
This is an interesting one and warrants its own blog. For now, we’ll generalise. Depending on the solution, SD-WAN can also incorporate many security features, replacing standalone security appliances, firewalls, VPNs and more advanced security functionality.
Lastly, it can also be a replacement for the corporate edge router, and some even have WiFi access capabilities making it an ideal solution for a remote office or retail chain.
Some Points to Consider
SD-WAN is showing rapid uptake, meaning that many of your competitors may be taking advantage of its benefits.
If you’re doing UCaaS or CCaaS, Office365, salesforce.com, AWS, Azure, Google, or cybersecurity, the use case for SD-WAN may be more pronouced.
Costs versus value: Bear in mind that some estimates for cost reduction associated with SD-WAN may be overstated. It can often come down to comparative costs of traditional connectivity. Uptake in the US is therefore considerable because of cost of national networks, and potential latency challenges with the distances involved. Outside of the US if you are a transnational / international business, it should be part of your consideration.
SD-WAN will not immediately displace MPLS across the broad market. But the continued use of MPLS should not delay trialing or even implementing SD-WAN along with it.
Many carriers are adopting SD-WAN, even though they can often make more money by selling MPLS. They do so out of recognition of the direction of the market plus their own need to evolve.
Firewalls and routers will potentially be displaced by multi-function SD-WAN boxes, meaning that your current infrastructure mix is likely to change over the course of time.
With the drive to move applications to cloud and the increasing mobility of workers, taking an application- / user-centric approach to network technology, design and management makes sense in many cases.
SD-WAN supports international operations and the new ways of access and remote working, so it’s always worth considering if your business has one or more of these characteristics. However, the fundamental reason to explore SD-WAN is based on the network performance and the latency-sensitive applications that are enabled by it.
We’re not saying SD-WAN is the answer to everything. And MPLS is not going away. It’s evolving. Tagoro work with dozens of providers, operating the whole gamut of services and technology. We look religiously at use cases, recommending a best-fit approach.
Drop us a line or call us if you have network challenges or projects where you’d benefit from independent oversight and full access to the SD-WAN provider market.