Back in 2015, when I changed my role from a project technical network engineer to a network solution architect, there was an element of trepidation that I was walking away from my core bedrock skills and into a potential world of uncertain employability. Fortunately, I need not of worried, there was a constant need for my skills regarding solutions and reading customer needs.
The SD-WAN network revolution
WAN solutions were relatively straightforward, with most built on MPLS, which was relatively easy to define. If you wanted to get complicated, there was Internet backup with an IPSec tunnel. These MPLS solutions and their variants served most customers well. But with SD-WAN on the horizon, things were, we were told, going to become much simplified with this new synergy of the existing WAN technological able to create neat WAN solutions from within a simple GUI and easily acquired Internet access.
One vendor said that software defined networking would put me out of a job as datacentre SDN would rapidly swallow up network WAN architects’ work. I have to admit I didn’t have much of an argument as it looked like customers and their non-IT network teams would be able to build their own SD-WANs with easy-to-use GUIs. Suddenly, it was possible to leverage a software-defined approach to any kind of transport services – including MPLS, LTE, and broadband internet services to connect users to applications securely.
Supporting the future network
Did any of this come to pass? Yes, it has given some enterprises more control and allowed them to build their own networks. But, truthfully, architects like myself working for managed service providers (MSPs) such as Telstra have never been so busy with WAN solutioning.
Why? Because we have to dig deeper. We now have to go right into the Internet underlay of the network, its peering, and cloud connectivity. For each enterprise we need to compare and contrast, the various overlay solutions by understanding the benefits, limitations, and implementation together with the unusual oddities between them to make it all work smoothly.
The conundrum of cloud connectivity
Then there is the elephant in the room, cloud connectivity for software as a service (SaaS) and platform-as-a-service (PaaS). This we find is probably an enterprise’s number one driver for moving across to SD-WAN.
The complexity of cloud connectivity is often not acknowledged. There are many cloud destinations to solve, and each SD-WAN vendor has more than one feature to help get you there. And there is a no one-size fits all option.
Customers also have their own ideas. You can guarantee that you think you’ve seen all the variations of customer requirements - something new pops up the next day.
So, SD-WAN keeps us busy and we continue to provide the best-in-market SD-WAN vendors teamed with Telstra’s underlay and valued added services. As enterprises look to support hybrid workforces and cloud-native network infrastructures, SD-WAN is a valuable technology in terms of agility and optimized connectivity. Last year, for example, the SD-WAN market grew a staggering 35% and achieved a record revenue of more than $2 billion, according to analysts. The growth continues.
SASE has landed
Just when we thought we had a clear vision in network solution architecture, secure access service edge (SASE) lands.
SASE, like SD-WAN, is a concept rather than a ready-to-go off the shelf product. SASE is designed to simplify WAN and security by offering both cloud services directly to the connection instead of using the enterprise data centre.
SASE and SD-WAN play in the same team with the common goal of ensuring users are securely connected wherever they are.
This simplification doesn’t mean you have to ring the death knell for network solutions architects. SASE is the next frontier in cloud security and another challenge for our architectural skills.