April 27, 2015. The world is changing. Containers are mainstream now—Docker’s momentum is unrelenting. Microservices are everywhere, and everyone wants to be a DevOps engineer. But amidst all this excitement, I find myself asking: where do we fit in?

Last week, I debugged a system that was running into container sprawl issues. It wasn’t glamorous, but it highlighted the very real challenges of managing infrastructure at scale. We had a bunch of microservices deployed across multiple Docker containers and Kubernetes pods, each with its own set of dependencies and configurations.

The problem was simple enough: one of our services was intermittently failing to start properly. The logs didn’t give us much to go on; it just died silently during startup. I spent hours tracing the issue, pulling hair (literally), until finally, the culprit emerged: a version mismatch in a library that was dynamically linked at runtime.

This kind of problem is what sysadmin work is all about. It’s not fancy or cutting-edge—it’s grunt work. But it’s also where real knowledge and expertise lie. The tools may be changing—Docker, Kubernetes, etcd—but the core principles of understanding your infrastructure remain the same.

I can’t help but think back to a time when I was more involved in day-to-day ops. When my colleagues and I were responsible for keeping our servers running, monitoring them with Nagios and Munin, manually deploying updates, and doing our best to anticipate failure before it happened. Those days are long gone, replaced by the promise of automation and self-healing systems.

But here’s the thing: those “old” skills haven’t disappeared entirely. They’ve just transformed into something more abstract. Understanding how your infrastructure works is still crucial—especially when dealing with complex distributed systems like containers and microservices. The difference now is that you’re leveraging tools to manage these complexities, rather than wrestling with them manually.

Take Kubernetes, for example. It’s a fantastic piece of software that automates so much of the work I used to do by hand. But there’s still an art to setting it up right, understanding how your services interact with each other, and ensuring everything runs smoothly. And when things go wrong—well, those moments can be just as frustrating as they were in the pre-container era.

This week, we also had a discussion about whether or not to move away from traditional VMs entirely. Some argued that we should dive deep into Kubernetes and Mesos/Marathon, while others preferred keeping our options open with a mix of both. It was a lively debate, but ultimately, I found myself siding with the idea of staying flexible.

The tech landscape is constantly evolving. But at its core, what matters most are the principles of good ops: reliability, performance, and resilience. These don’t go away just because you’re using containers instead of VMs. And that’s why, despite all the buzz around new technologies, I still root for sysadmin.

Because in the end, it’s not about the tools—it’s about knowing your infrastructure inside and out. It’s about being able to troubleshoot issues quickly and effectively. And while those skills may be less visible in an age of automation, they are no less important.

So here’s to the old days—long as they are—and to the new ones that await us. May our sysadmin knowledge never truly die.