I’m fascinated with the entry of IT vendors into the “Edge” market. The Edge includes everything from the data center (as opposed to the public cloud) to windmills and fitness bands on consumers. I’ll let my peers continue to battle out the definition of Edge. However, realize that the IT management discipline doesn’t align with the financial drivers of all of the Edge deployment I’ve managed. To help illustrate the challenge, let’s talk about printers.
Code Red on a Copier
In 2001, I worked at a value-added reseller, and the Code Red computer worm roiled our clients. It was one of the first viral worms of the internet era. Unpatched Windows NT systems were especially susceptible to the virus. And back in 2001, there were plenty of unpatched Windows NT. A few years later, I ran into the worm in an unexpected environment.
In the mid-2000’s, I was an IT manager for a manufacturing firm. We maintained a fleet of 1999 era Xerox Multifunction Printers (MFP).
If you think you know where I’m going with this story, you are right. I noticed suspicious network traffic from one of the devices. The device was running an embedded Windows NT operating system. And as you might have guessed, the MFP had Code Red. Because who patched printers for Windows worms back in 2005?
Copiers aren’t refreshed every 3 to 7 years, like desktops and servers. Before today’s as a service model, new copiers were purchase when they stopped working or became too expensive to fix. After all, printers are purpose-built devices that perform their purpose until they break.
Now take this printer example and apply it to other edge-like solutions. When I worked in fast-food in the early ’90s, I was shocked to see an IBM 8inch Floppy disk for the first time. The fast-food restaurant’s point of sale (POS) solution used 8inch Floppy disks to update the POS devices’ software. As a teenager, I was amazed at the museum-quality piece of computer history.
However, I’d discover these weren’t isolated incidents. Most edge solutions are purpose-built to run until they no longer work. Today’s platforms are no different. I recently watched an introduction to KubeEdge, an opensource project. The goal of the project? To enable Kubernetes to orchestrate workloads on the Edge. These solutions are still basically a snapshot in time. For the majority of Edge customers, the deployment will become a living time capsule.
The challenge is driven by the depreciation schedule of Edge technology purchasers and enterprise IT sales and support cycles. What does a Farmer call an internet-connected tractor? They call it a tractor. The typical life cycle of a tractor is 15-years. Fast-forward ten years from now after a farmer purchases a modern tractor. Will the Kubernetes control plan still be supportable 10-years from now? Or will we see today’s Code Red equivalent running on that tractor, and the solution is to replace the 10-year old iPad Air with a brand new iPad Air based on 10-year old technology?
I’m glad you asked. It’s just another question for business leaders to consider. Buying the right technology at the wrong time can ruin your business. The timing of these purchases is crucial. If a farmer purchases tractors with technology displaced within a year, she simply doesn’t roll out a new set of tractors running the latest code.
Manufacturers and technology companies must come together with modular solutions that enable software upgrades for these assets’ lives. I haven’t seen these agreements in the wild. I’d love to hear if you’ve had a different experience @ me on Twitter.