Kubernetes - Does It Live Up to the Hype?

​

Twelve years ago, I landed my first DevOps role.

I was a junior engineer, responsible for infrastructure and security at a startup (yeah, not the best call on their behalf), that would later become a unicorn.

One evening, our CEO landed the first huge client, and we had 24 hours to prepare.

I panicked.

I spun up extra servers, scaled out load balancers, and froze every scale-down rule I could find to make sure we overkill this and that I'm not the single point of failure.

Somehow, we survived.

But that night made one thing clear: we needed a better way to scale infrastructure.

Fast forward to today, Kubernetes is solving exactly this kind of challenge.

But at what cost? Do you, or your team really need it?

Let’s talk about when Kubernetes makes sense, when it’s overkill, and what alternatives exist.

---

The Problem: Scaling Infrastructure Without Wasting Resources

Before 2013, companies running SaaS applications typically scaled by upgrading to bigger servers or adding more as needed.

But this manual approach was inefficient—it was hard to predict demand, leading to either under-provisioning (downtime) or over-provisioning (wasted costs).

Load balancers helped distribute traffic, but maintaining and tuning them was a constant challenge.

Then came containerization, which promised a more flexible way to run applications.

But managing containers at scale required something more—a system that could orchestrate and optimize resource usage automatically. That’s where Kubernetes comes in.

​

What most teams do (And why it doesn’t always work)

Many companies rush to Kubernetes as the default solution for container orchestration.

It provisions servers, balances workloads, and automatically scales resources.

It also integrates with load balancers and supports redundancy for high availability.
​

However, Kubernetes is complex.

Setting up a production-ready cluster, with monitoring and automation can take weeks or months.

Hiring engineers who understand it is expensive.

And if you’re running a small to mid-sized application, the overhead might not be worth it.
​

A Different Approach: Choosing the Right Tool for the Job

Kubernetes is a powerhouse, but it’s not the only solution.

If you’re running a simple application, a managed container service like AWS ECS, Google Cloud Run, or even a simpler orchestrator like Nomad might be a better fit.

You know what, maybe you don't even need an orchestrator?

For local development, tools like Minikube offer a lightweight Kubernetes experience without the complexity.

For remote, yet lightweight deployments, like on edge-locations, or a low-in-resources set of hosts, K3s might be an excellent alternative to its bigger sibling.

The key is knowing whether Kubernetes aligns with your needs—or if you’re adopting it mainly because it’s trendy.

Now, about those needs, many teams end up running too fast to pre-mature optimization, which K8s can be a classic example of; A 2-container system, with Postgres and Redis, might not need an entire control plane, a fleet of nodes and controllers to support them.

Maybe deployment platforms such as Vercel, Fly, Heroku can be great to hit the ground running.
If there's no extreme time constraints, I'd actually opt for a cloud platform and maybe start with their proprietary deployment platform, such as AWS ECS that I've mentioned earlier. Giving you a lot of the power of Kubernetes, without the overhead complexity that comes with it.

---

Key Takeaways and Actionable Steps

1. Assess your needs – Do you truly require Kubernetes, or is a simpler solution sufficient?
2. Assess your skills – Do you have the in-house skills to deploy, run, and maintain it?
3. Understand the trade-offs – Kubernetes is powerful but comes with a steep learning curve. If you don't foresee an extremely special use-case, a 3-containers deployment with a DB might not be a good fit.
4. Start small – Experiment with Minikube or K3s before committing to full-scale Kubernetes deployments.
5. Look at alternatives – Tools like Nomad, Docker Swarm, or managed cloud services can simplify your workflow.

Kubernetes is here to stay, and learning it can be beneficial—even if you don’t use it immediately.

The real question is: are you using it for the right reasons?

All that said, if you did chose to go down that road, Kubernetes is an absolute beast when it comes to infrastructure. The control plane and the huge ecosystem will ensure you can scale vertically and horizontally, both pods and nodes, run functions, stateful workloads, daemons and pretty much anything you can thing of!

​
Thank you for reading.
Feel free to reply directly with any question or feedback.
Have a great weekend!