Managing containerized applications across a cluster of machines sounds simple in theory and gets complicated fast in practice, which is exactly the gap Kubernetes was built to close. This DenverJUG meetup took a practical look at how Kubernetes approaches that problem.
At its core, Kubernetes is a system for scheduling and managing containers across a group of nodes, designed to smooth over the disconnect between how modern clustered infrastructure actually works and the assumptions a lot of applications and services still make about their environment. The goal of most clustering technology is to give developers a uniform platform to deploy against, so nobody has to think too hard about exactly which machine their code ends up running on. Kubernetes takes that idea and pushes it further, framing deployment around services rather than individual machines.
We walked through the basic building blocks, including pods, services, and how scheduling decisions get made, and talked about where Kubernetes fits relative to other clustering and orchestration approaches the group had used in the past.
As containerized infrastructure keeps becoming more common, understanding the fundamentals behind tools like Kubernetes is increasingly just part of the job for backend and infrastructure-minded developers.
Thanks to everyone who joined us, and to the sponsors who help keep DenverJUG meetups free for the Denver software community.