Anyone who knows me professionally knows I've been enthusiastic about Kubernetes at the Edge for years: running Kubernetes in places you might not immediately expect. Think factories, greenhouses, and utility closets, but also other environments where IT doesn't naturally run in a datacenter or public cloud.
Between 2022 and 2026, I spent a lot of time championing this vision and showing what the technology can mean in practice. That wasn't always easy. Especially in the early days, I regularly heard the same question: why would you bring complex technology like Kubernetes to environments that are so unpredictable?
That hesitation was understandable. Kubernetes was originally known mainly as a platform for orchestrating workloads in datacenters and cloud environments. The ecosystem evolved quickly, though. With the arrival of projects like K3s, and later Talos Linux and Kairos, more and more solutions became available that made it easier to run Kubernetes in smaller, distributed locations. The technology became more accessible, while its management and robustness kept improving.
Still, conveying that shift in words alone wasn't always easy. Presentations and architecture diagrams are useful, but they don't tell the whole story. I wanted to show something that actually worked, preferably a physical system you could bring along to a customer or a conference. Precisely because physical presence is such an important part of Edge computing, it seemed worthwhile to make the infrastructure itself tangible too.
There were a few clear requirements. The system had to be compact and easy to transport, capable of running a full Kubernetes cluster, and manageable remotely. That way I could not only explain the technology, but also demonstrate how you manage a cluster in a location without extensive local IT facilities.
That's how the idea for the Edgecase was born.
A Kubernetes cluster in a suitcase
The name was a nod to Edgecase, the conference TrueFullstaq has organized since 2022 around Kubernetes at the Edge and related technologies. The device had to make that same idea tangible: showing what's possible when you push the boundaries of traditional infrastructure.
The base became an aluminum case with two layers. We routed the cabling away in the bottom compartment, while the top layer held the hardware and the remaining components. The result had to be functional, but also look good. Looks matter too, after all, especially when you're telling a technical story on stage.
Three OnLogic CL250 nodes went into the case. OnLogic specializes in industrial hardware, which fit well with the story I wanted to tell. This wasn't just a random collection of computers, but compact, industrial-grade systems that gave a realistic picture of the hardware you might actually encounter in other environments. Each node had an Intel processor, 8 GB of memory, and a 64 GB SSD. Thanks to passive cooling, no fans were needed either.
A simple desktop switch connected the three nodes. For external connectivity, we added a 4G router. That meant the Edgecase could connect to the internet even in locations without existing network infrastructure.
For the operating system, I chose Talos Linux. Talos is purpose-built for running Kubernetes and has a minimal, largely immutable base. The absence of a traditional OS onto which all kinds of additional software can be installed reduces the number of components that need managing. That fit well with the goal of the demo: a compact system that could operate as autonomously and reliably as possible.
The three nodes were then connected to Sidero Omni, the platform used to manage Talos systems centrally. Through Omni I could configure the machines, bring them together into a Kubernetes cluster, and maintain them remotely. We opted for three control-plane nodes that could also run workloads. That kept the available compute usable, instead of the cluster architecture consisting solely of control-plane capacity.
The combination of Talos, Omni, and the 4G connection made it possible to set up the Edgecase in virtually any location and then manage it centrally. Running updates, accessing the cluster, and checking the status of the infrastructure didn't depend on physical access to the device. That was exactly one of the key properties I wanted to demonstrate.
We also mounted a screen in the lid showing the output of one of the nodes, along with a few LEDs. These additions played no essential role in the cluster's operation. They mostly made the technology a bit more visible and gave the whole thing the look of a device you could proudly place on a stage.
A weekend to build, years to demo
Building the Edgecase was a joint project. I handled most of the technical installation and configuration myself, but for the physical construction I got help from my father-in-law. He's exceptionally skilled with his hands and works with great precision, exactly the qualities you need when mounting hardware securely and neatly inside a case.
We worked on it together for a weekend. While he handled the assembly, I installed and configured the various components. Eventually everything fell into place and there was a working system inside the case. The Edgecase was born.
Plans quickly formed to show the device on stage for the first time at the third edition of the Edgecase conference, in 2024. The reveal would be part of a talk about Kubernetes at the Edge and the evolution of the ecosystem.
In hindsight, that moment turned out to be the start of far more than a single demo. The talk grew into a story I've since been able to tell more than fifteen times, at events and conferences across Europe. The Edgecase also sparked hundreds of conversations with people interested in the technology, building similar solutions themselves, or thinking about applications in their own environment.
The device's power wasn't just in the technology. It gave people a concrete starting point for conversations that might otherwise have stayed abstract. You could look at the hardware, ask questions about the architecture, and see right away that a real Kubernetes cluster was actually running on it.
From customer conversations to conferences across Europe
In the years that followed, the Edgecase traveled with me to customer meetings and conferences in, among others, the United Kingdom, Germany, Belgium, and France. At TrueFullstaq events, the device regularly sat in our booth. It drew attention and provided a natural way to start conversations with visitors about Kubernetes, Edge computing, and managing distributed infrastructure.
The device found a place beyond our own events too. Our partner Sidero, for instance, asked us to bring the Edgecase to KubeCon so it could be demonstrated at their booth. That worked out great. The Edgecase wasn't just a demonstration of a Kubernetes cluster, but also a tangible example of what Talos and Omni can do together.
A personal highlight was the extensive conversation I got to have with Kelsey Hightower about the device and why I'd built it. Moments like that show exactly how a relatively small, self-built project can grow into a reason for conversations with people you might otherwise never have met.
The Edgecase's final flight
Last month, during my visit to Sofia in Bulgaria, the Edgecase made its final trip. For the first time, the device had to fly. That brought a new challenge: how do you make sure a compact but fragile collection of hardware survives a flight?
When building it, we'd already accounted for the fact that the device would need to travel. The hardware was mounted with Velcro, so the components stayed firmly in place while still being easy to detach. For the flight, that let me take the three servers out of the case and carry them as hand luggage. The case itself traveled nearly empty, with only the cabling left in the bottom compartment.
The trip there went off without a hitch. Once I arrived in my hotel room, I put the hardware back together, connected everything, and powered the system on. As I'd come to expect by now, the cluster booted up without issues. The combination of Kubernetes and Talos did exactly what was expected of it.
It was a nice example of the resilience I wanted the Edgecase to demonstrate. The hardware had been transported, rebuilt, and connected at a different location, without needing any complicated recovery procedure.
At KCD Sofia, the Edgecase then took the stage once more. There too, it drew interest and sparked new conversations with visitors. It was a fitting finale, though at the time I didn't yet know it would truly be the last one.
On the way back, things went wrong. The infrastructure itself had survived the trip again, but the aluminum case hadn't. The latches were completely broken and turned out to be beyond repair. That made it no longer responsible to keep transporting the device the same way.
The Edgecase retired.
A successor, but not a copy
Naturally, that raises the question of what happens next. Buying a new case and remounting all the components would be the simplest solution. But this is also a good moment to rethink the device's purpose.
Technology keeps evolving, after all. What was an interesting demo a few years ago can look very different today. Maybe there should be an Edgecase 2.0, with a different architecture or a whole new theme. A compact demo environment for local LLMs and private AI, for example, could fit well with the developments we're currently seeing in the market.
For now, those are just ideas, not concrete plans. What is clear, though, is that any successor needs to be more than a replica of the original. What matters most isn't the case itself, but what you can show with it and the conversations it makes possible.
The value of building something yourself
Looking back, the Edgecase gave me much more than a handy demo device. Designing and building it was a genuinely enjoyable project in its own right. In our work, we spend a lot of time on digital infrastructure, virtual machines, cloud platforms, and software that largely stays invisible. Those are important parts of our craft, but there's also something satisfying about working with physical hardware. You hold the parts in your hands, build something that actually exists, and can show the result to someone who might not know anything about Kubernetes at all.
I find that last part just as important. Explaining a working Kubernetes cluster using architecture diagrams can be difficult, especially for people who don't work with the technology daily. A physical case full of computers, cables, a screen, and lights speaks to the imagination far more directly. You can point at what you built and show what it does. It makes an abstract story tangible.
The project also reminded me how much you learn by experimenting yourself. Books, training courses, and certifications are valuable and can provide a solid foundation. But some insights only emerge once you build a system yourself, run into problems, and discover why certain choices do or don't work. You often only truly understand a technology once you get your hands dirty with it.
That's also one of the reasons I set up my own homelab at home, with a compact 10-inch rack and various systems for trying out new technologies. It gives me room to experiment without every project needing a business purpose. Sometimes you build something just to find out if it's possible. Sometimes you discover an application along the way that you hadn't planned for. And sometimes it's simply fun.
In my view, that freedom to try, build, and learn deserves more attention. Not every experiment needs to grow into a product or a commercial success to be worthwhile.
Goodbye, Edgecase
The physical case may have reached the end of its life, but the idea behind it is still very much alive. The Edgecase helped me explain Kubernetes at the Edge in a different way, took me to remarkable places, and led to conversations I might otherwise never have had.
Above all, though, the project reminded me of something that easily fades into the background in an increasingly digital world: the value of making something yourself, simply out of curiosity and because it's fun to discover what's possible.
Thank you, Edgecase. You more than proved your worth. And who knows, maybe one day a successor will take the stage, with a new case and a whole different story.
