Spring GDS 25th Anniversary
A logistics company that ships to 190 countries built something to ship to itself.
Kubernetes
What is Kubernetes?
Kubernetes is a system for running containers across a cluster of machines. You tell it the state you want, say five copies of this service, always healthy, reachable at this address, and it works to keep reality matching that description. If a container crashes, it restarts one. If a machine dies, it reschedules the work elsewhere. If traffic climbs, it can add copies.
It grew out of how Google ran software internally and became the standard way to orchestrate containers at scale. Kubernetes handles deployment, scaling, networking between services, rolling updates, and self-healing, all driven by declarative configuration you commit alongside your code. The catch is that it is a large, intricate system with a real learning curve and real operational cost. A small app does not need it, and running it badly is worse than not running it at all. A streaming platform that has to add capacity the moment a show goes live, then scale back down overnight, uses Kubernetes to add and remove containers automatically against live demand.
It pairs naturally with infrastructure as code and CI/CD, since the whole cluster is described in files and changes flow through the same pipelines as application code.
Kubernetes at Dallonses
Our first question is always whether a client actually needs Kubernetes. Plenty of products run beautifully on a managed platform or a couple of containers with far less to operate. We reach for it when the requirements genuinely call for it: many services, real scaling demands, multiple environments that have to behave identically. When they do not, we say so, because unused complexity is just cost with extra steps.
When Kubernetes is the right tool, we keep it boring on purpose. The cluster lives in version control, rollouts flow through CI/CD pipelines, and nothing changes by hand on a live system. That ties into our platform standardization work, so every environment is reproducible, and into cost optimization, because a cluster left on autopilot quietly burns money on capacity nobody is using.
Not sure whether Kubernetes fits your scale or just adds overhead? Let's figure out what you actually need.
Related services
Related works
Vizitio
Turning a brand into a working business.
Access Ticket
Half a million people. One app. Zero chaos.
Ready to work together?
Book a meeting
