In their annual hype cycle on on network technologies, Gartner lists the emerging technologies and an estimate of the timeframe in which they will reach the plateau of productivity. The latest hype cycle on enterprise networking labels Software-Defined Networks (SDN) as an obsolete technology. So on the surface, it would appear that SDN is now semi-officially dead.
While most natural scientists accept the Darwinian theories, the technology industry has traditionally been trying to defy evolution. Every 10-20 years, new revolutionary technology is born with the expectation that it will completely transform how a certain thing is being done. Assuming that Gartner's analysis of SDN is correct, it would appear that it has moved on to the nowhere land of promising technologies that never happened.
Yet when one thinks about it, the development of technologies tends to advance in waves of incremental improvements. A good example is electric cars. Most of the technologies needed to build them have been around for decades. However, it required a major catalyst like climate change to start their journey to the mainstream. And while the transformation of the automotive industry is getting into full swing now, there still are many familiar elements from the previous technology generations that make up a good 90% of the electric cars offered in the showrooms.
Networks are no different in this regard. While SDN may well be dead in the water, it does not necessarily mean that the school of thought it represents should be forgotten or obsoleted. Quite the opposite. SDN is all about programmability, agility and resulting service elasticity, which I believe will remain at the core of future networks.
If not SDN, then what?
When engineers fall in love with a specific technology, they often forget that most people do not care how something has been implemented. Instead, what they care about is convenience and utility. Like the slow uptake of IPv6 shows, even the most eloquent of technologies struggle to be adopted when they are not powered by a megatrend that drives concrete benefits to end-users.
Looking at the industry trends, the 2010s have been all about cloud computing and digital transformation. While networks remain an important piece of this puzzle, SDN, in its purest form, is too big a leap to take at once. Enterprises and service providers around the world have invested hundreds of billions into the network infrastructure they have today. The assumption that they would write off all these investments may err to the side of optimism.
That said, what we are seeing increasingly today are the features pioneered by Software-Defined Networks being implemented using alternate methods. Programmability leading to service agility, better performance characteristics, and lower latency are all key factors when new networks are being designed. In the cloud-native era, any new business infrastructure that lacks these abilities will not have a very long future ahead of it.
Instead of SDN, these architectures are being implemented using automation frameworks and orchestrators such as Ansible and Open Network Architecture Platform (ONAP). These technologies allow organizations to introduce automation to even existing network devices. Network automation reduces the largest cost of networks of today; downtime caused by manual mistakes. Automation is perhaps one of the highest yielding investments an entity running its network can make.
There are, of course, new use cases, especially in the enterprise space where emerging network technologies such as Software-Defined Wide Area Networks (SD-WAN) and virtual extensions of the enterprise network such as AWS VPC and Azure VNET make perfect sense. Both differ significantly from how enterprise networks have traditionally been run. While these technologies are still rather young, the chances are that the future of networks is in automation that allows both these new technologies and the traditional networks to co-exist side by side through the 2020s.
Conclusion
While SDN in the academic sense may not gain the traction it was initially hoped to, it has paved the way for network automation that is used on a much broader scale. Rather than forcing a tightly defined new network paradigm such as SDN upon service providers and enterprises, the industry is headed towards agnostic network automation frameworks that are driven by the business objectives rather than an academic definition. The technologies facilitating the transformation in this area are network automation frameworks such as Open Network Automation Platform (ONAP) and orchestrators such as Ansible.
By Juha Holkkola, Co-Founder and Chief Technologist at FusionLayer Inc. – Juha Holkkola is the Co-Founder and Chief Technologist at FusionLayer Inc. An inventor with several patents in the US and Europe, he is an advocate of technology concepts with tangible operational impact. Juha is an active proponent of emerging technology trends such as cloud computing, hybrid IT and network functions virtualization, and a regular speaker at various industry events.