I have about five years of experience spanning DevOps, cloud, and backend development. I've worked with Kubernetes, EKS, AKS, Helm, CI/CD, Docker, Terraform, Argo CD, AWS, Azure, and some Java and Spring Boot. I'm comfortable deploying applications, writing manifests, using kubectl, building pipelines, and handling cloud networking and IAM, but my understanding is less consistent in areas such as Kubernetes internals, Linux networking, storage and CSI, the control plane, scheduling, CRDs and operators, upgrades, observability, and deep production troubleshooting. I'm thinking of starting with one comprehensive Kubernetes course to create a solid mental model, then spending several months on hands-on labs, troubleshooting exercises, Linux and networking fundamentals, and production-style projects. For people who became genuinely strong in Kubernetes or platform engineering, does that sequence make sense? What would you change, and which activities made the biggest difference in developing real depth? I'm mainly looking for advice on the learning approach rather than a huge list of tools.
2 Answers
The right path also depends on your goal. If you already have a senior DevOps role, don’t feel that you need to master every implementation detail before your experience is valuable. Use the fundamentals to help teams ship more reliable systems, improve platform design, and make better operational decisions. For the gaps you specifically mentioned, choose a few realistic projects and investigate them deeply: trace a network request end to end, diagnose broken scheduling or DNS, test a storage failure, perform an upgrade, and examine control-plane and workload metrics. That gives you depth tied to real engineering outcomes instead of collecting isolated facts.
That sequence makes sense, but make the course the starting point rather than the main event. The biggest jump usually comes from understanding what Kubernetes is doing underneath the abstractions. Spend time with Linux namespaces, cgroups, process isolation, mounts, and filesystems, and recreate small pieces by hand from the command line. Then study how CNI and CSI components work, including how pod interfaces connect to host interfaces and how volumes are mounted inside containers. Packet-level walkthroughs of a request moving through Kubernetes are especially useful. Build small failure scenarios and troubleshoot them instead of only following successful deployment labs. That combination of a broad mental map and deliberate experiments should turn your existing experience into much deeper understanding.
This is exactly the kind of actionable direction I was looking for. I’ll focus more on reproducing the underlying behavior and troubleshooting it instead of just completing another tool-focused course.
I’m already working as a senior DevOps engineer at a large company, so this is mostly about strengthening my foundation. I may be underestimating how much I already know, but I still want to level up deliberately.