SDWAN remains a topic of high interest among Gartner clients. While many networking technologies are over-hyped as the next big thing, SDWAN is delivering on the promise. In just three short years, adoption has taken off and we now estimate 6,000+ paying SDWAN customers with more than 4,000 production implementations. We recommend you look at SDWAN when refreshing WAN edge equipment, renegotiating a carrier contract, building out new branches, or aggressively moving apps to the cloud (among other reasons).
SDWAN has several very real and tangible benefits (cost, performance, availability to name a few) with clear ROI when compared to alternatives. In fact, this is what is driving rapid adoption. However, one question that often comes up with clients is “what are the downsides” or “it sounds to good too be true”. With that as the backdrop, here are some of the downsides (or “challenges” as the marketers call them) affiliated with early SDWAN deployments…
- Substantial functionality gaps exist with many well-known SDWAN vendors/products. For example, features that are missing from many SDWAN products include T1/E1 interfaces, integrated 4G/LTE capability, Wan Optimization, and even support for IPv6. This will change and vendors are adding these capabilities as we speak, but at this point in time, most vendors don’t have them all.
- SDWAN can facilitate a move towards lower-cost Internet transport, but this can lead to challenges in managing heterogeneous connectivity. MPLS did something that everyone knows about, but that most people don’t really appreciate – manage that last mile. I’ve talked to many SDWAN early adopters that started bringing in lots of Inet connections only to find that the heterogeneous nature of vendor and/or service offerings prompted them to bring in a third party aggregator to manage it, or go with a higher cost provider with a larger footprint.
- Despite the name, there’s a good bit of hardware involved as most SDWAN vendors lead with hardware appliances in accounts.
- While SDWAN is much easier to configure and operate than traditional routers (probably 10X when making complex application routing policy), many SDWAN implementations have actually added technical debt to the branch. This is because many of the early adopters are not actually replacing an existing router (at least not yet) for a variety of reasons including aversion to risk, lack of functionality, etc… Instead they are placing it behind and/or in conjunction with the existing router. I anticipate this will (and should) change soon.
For reference, our most recent SDWAN research is here:
Technology Insight for Software-Defined WAN (SD-WAN)
Summary: SD-WAN is an early mainstream technology that offers several benefits compared to traditional, router-based WANs. I&O leaders responsible for planning, sourcing and managing WANs can reduce costs and improve agility and uptime by using SD-WAN products.
Market Guide for WAN Edge Infrastructure
Summary: WAN edge infrastructure is changing rapidly as I&O leaders responsible for networking face dynamic business requirements, including changing application architectures. This guide covers a fragmented vendor and technology landscape, with a focus on the increasingly relevant topics of SD-WAN and vCPE.
Regards, Andrew
The Gartner Blog Network provides an opportunity for Gartner analysts to test ideas and move research forward. Because the content posted by Gartner analysts on this site does not undergo our standard editorial review, all comments or opinions expressed hereunder are those of the individual contributors and do not represent the views of Gartner, Inc. or its management.
6 Comments
Great article as always Andrew. It’s interesting to see how these next levels of complexity emerge as enterprises move past the marketing promises and see the reality of SD-WAN deployments.
The IPv6 comment is interesting – this is a topic that many people easily dismiss but it has some important implications in SD-WAN. We’re already seeing some ISP services in APAC that use a private (CG-NAT) IPv4 address and a routed IPv6 prefix. For inbound connections to the site (like a dynamic SD-WAN spoke-spoke tunnel) the IPv6 path may be the only option. Even if the enterprise has no immediate plans for internal IPv6 adoption (and most don’t) they can be accidentally dragged into this world based on the ISP services they buy.
The ISP vendor management overhead is another great point that is often overlooked, and it’s essential to get the cost base right at this layer for most SD-WAN business cases to work!
6000+ paying clients and 4000 production installs. Those 2000+ clients without installs gives me a slight headache but interesting none the less.
We have WAN SDN in top 3 private telcos/1 Government Ministry & 2 BFSI Clients doing Physical/logical inventory/performance/fault/configuration/traffic engineering & IP MPLS Planning & Resilency building over L2/L3 layers.there is scaled down Enterprise version too
Mahesh Bhatt
Can anyone share the earnings from SD WAN DIY Vs Managed Services approach – is there really a cost benefit ?
What are the parameters to consider ?
Hi Shubha, we’ve published on both the DIY vs MNS and cost ramifications of SDWAN:
DIY vs. MNS: Enterprises Must Reassess Their Network Sourcing Model to Prepare for SD-WAN
https://www.gartner.com/document/3796585
Published 8 September 2017 – ID G00333193 – 14 min read
Summary: Compared to legacy technology, software-defined WAN products are often easier to implement and manage. The buying decision is more complex; I&O leaders must reassess their current sourcing models to analyze the short- and long-term impacts of do-it-yourself versus managed SD-WAN services.
Fact or Fiction: Does SD-WAN Really Save You Money?
https://www.gartner.com/document/3861480
Published 27 February 2018 – ID G00345417 – 16 min read
Summary: SD-WAN improves WAN performance and reduces bandwidth pricing, but may not always result in lower WAN expenses. When planning SD-WAN projects, I&O leaders can use this research to compare before-and-after expenses to confirm that the value of achieving business goals exceeds any increased expenses.
Can anyone share the learnings from SD WAN DIY Vs Managed Services approach – is there really a cost benefit ?
What are the parameters to consider ?