Docker & Kubernetes: Ship with Confidence
Package a real API into a lean, secure image and run it on Kubernetes without 3 a.m. surprises.
- 4.8 (976 ratings)
- 8,900 learners
- Intermediate
- 6h 20m total
- Updated July 2026
What you’ll learn
- Write fast, cache-friendly Dockerfiles with multi-stage builds
- Ship small, non-root images pinned by digest and scanned for CVEs
- Run an app and its database locally with Docker Compose
- Deploy with Deployments, Services, ConfigMaps and Secrets
- Add probes, resource requests and autoscaling that hold up under load
- Roll out, roll back and debug a broken release with kubectl and Helm
Who it’s for
Containers are how most software ships in 2026, and Kubernetes is where much of it runs. This course takes one small service, notes-api (Node.js and Postgres), from a laptop to a cluster. You write the Dockerfile, shrink and harden the image, wire up a local stack with Docker Compose, then deploy, scale, update and roll back on Kubernetes.
Your instructor is Amara Diallo, a cloud architect at Skylane who carries the pager for production clusters. Every recommendation here is tied to an incident it prevents: the cache that took builds from forty seconds to eight minutes, the secret baked into a public image, the rollout that sent traffic to Pods that weren't ready. You learn the defaults that keep services boring.
- Who it’s for
- Developers who can build a small web API and now need to package, run and operate it with containers.
- Before you start
- Comfort with the command line, Git and one backend language; Node.js is used in examples.
Hands-on practice
- Guided exercises: spot the bug, fill the command, order the steps.
Course content
5 sections · 20 lessons · 6h 20m total
- What a Container Really Is14 min
- Your First Dockerfile20 min
- Layers and the Build Cache18 min
- Running Containers: Ports, Env, Logs and Signals18 min
- Section checkpoint5 questions
- Multi-Stage Builds and .dockerignore20 min
- Small Bases and Non-Root Users20 min
- Tags, Digests and Registries18 min
- Scanning Images and Keeping Secrets Out20 min
- Section checkpoint5 questions
- Compose Fundamentals: Services and Networks18 min
- Volumes: Data That Survives16 min
- Startup Order with Healthchecks18 min
- Env Files, Overrides and Watch Mode20 min
- Section checkpoint5 questions
- The Control Loop and Your First Cluster16 min
- Pods, ReplicaSets and Deployments20 min
- Services: Stable Addresses for Moving Pods20 min
- ConfigMaps, Secrets and Namespaces20 min
- Section checkpoint5 questions
- Probes and Graceful Shutdown20 min
- Requests, Limits and Autoscaling22 min
- Rolling Updates, Rollbacks and Helm20 min
- Traffic In, Trouble Out: Gateway API and Debugging22 min
- Section checkpoint5 questions
Your instructor
Amara Diallo
Cloud Architect · Skylane
- 4.6 instructor rating
- 19,700 learners
- 2 courses
Amara racked servers by hand in a university data center before containers existed, then spent eight years moving companies onto the cloud. At Skylane she designs the AWS and Kubernetes platform that runs more than 300 services, and she still takes her turn on call.
What learners say
4.8
976 ratings
Sipho Ndlovu
8 months ago
A multi-stage build took our API image from 1.2 GB to 160 MB. The spot-the-bug Dockerfile exercises are why I now read every COPY line twice.
Lina Jaber
5 months ago
Readiness versus liveness probes, explained with a pod that keeps restarting at 3 a.m. That one example fixed a real outage pattern on my team.
Noah Fischer
2 months ago
The Docker half is excellent, the Kubernetes half is good. Without a real cluster the kubectl practice is fill-in-the-blank, so I ran kind locally alongside it.