2018 is poised to be year of reckoning for legacy networks in Asia Pacific (APAC). Built long before the cloud was formed, they’ve rapidly become an Achilles heel for businesses embracing digital transformation, sparking a tipping point in the adoption of next-generation networking technologies like SD-WAN.
In a survey conducted by Riverbed Technology late last year of more than 400 APAC IT decision-makers, nearly all (97%) agreed that within two years, SD-WAN technology will be critical in next generation networks to manage cloud and hybrid infrastructure. And while only 6% were currently benefiting from SD-WAN at the time of the survey (slightly higher than the global average of 4%), within two years 57% planned to migrate to SD-WAN, and 92% within four years.
As someone who’s spent his nearly 25-year career in the networking industry, playing a role in the advancement of SD-WAN – a revolutionary new approach to connecting people and data – has been nothing short of thrilling.
But what’s even more exciting is that it’s just the beginning. While today, SD-WAN can automate secure connectivity and guarantee performance across a global enterprise network, it’s not yet intelligent enough to self-protect against an outage. But that’s changing – and fast.
Today, most large network and application infrastructures are complex and sprawl across the globe, making it incredibly challenging for IT teams to pinpoint and troubleshoot issues when they arise with traditional approaches. When outages occur – and they inevitably do at the most inopportune times – business owners become frustrated, and IT teams are left scrambling to identify the root cause and to shore up the infrastructure to prevent future outages from occurring, but they inevitably occur again.
It seems to be a cycle that is destined to repeat itself again and again, with no end in sight.
However, we’re now on the cusp of being able to create a network that mirrors the human body’s capacity to find issues and heal them before the whole body succumbs. Much like the simple injection of a vaccination enables our immune system to squash future bugs, so too will we be able to build a resilient network that is not brittle -- an organic network that can heal itself, a network that can be vaccinated from infection and infrastructure that can protect the most important applications for the most important people at the most important time.
This is the promise of a software-defined approach or SD-WAN. However, current infrastructure needs more to become a truly organic network.
The combination of SD-WAN, cloud, built-in monitoring, artificial intelligence (AI) and machine learning will soon allow us to realize the dream of a resilient organic network. This will be a system that can protect itself from a viral onslaught. If a serious problem is diagnosed, the combination of SD-WAN and automation will allow the operators to create a vaccine so that the network becomes immune to similar threats in the future.
For example, let’s say you are out of hypervisor resources, then why not provision portions of the application in infrastructure as a service (IaaS)? Once the symptoms of a problem are identified, a vaccine can be created to inoculate the system from future problems, just like our bodies do every day.
Built-in monitoring will be a central component of this new, resilient network. Today, we deploy troubleshooting tools, usually after a problem has already occurred. Much to the annoyance of the operations teams, we have to wait for that same issue to arise again in order to even have a chance to pinpoint it and fix it.
This is what I refer to as bolt-on monitoring. But that is not how the human body works.
Contrast that approach with this one: If every computer, mobile device, network equipment and application can provide telemetry to a central command center, early warnings can be dealt with before they develop into a catastrophic failure.
Taking it back to our analogy of the human body, this is similar to the clotting factors that rush to the site of a cut to prevent blood loss. At the very least, all users’ transactions should be monitored instead of relying on synthetic traffic. We should monitor every end-user by using the actual traffic generated by the end users.
This is the definition of built-in digital performance monitoring.
Every outage causes the business to ask, “What are you going to do to the next time a problem like this occurs?”
As managed service provider or an operations professional, wouldn’t it be great to be able to explain that a vaccine has been created, the organic network has been inoculated and future outages will be at most a minor bump if that? We are not too far away from that future.
Hansang Bae is chief technology officer at Riverbed Technology