OpenShift — Portfolio Projects
Build these to prove you can actually run OpenShift, not just recite SCC and Route trivia. Each project maps to a real EX280-style task area, so finishing all three doubles as certification practice.
Project 1: Multi-Environment App with Routes, SCC, and Quotas
Level: Beginner | Time: 1-2 days
Deploy a small web app (any language — a simple Node/Python "hello" API is fine) into three Projects: myapp-dev, myapp-staging, myapp-prod. Each environment gets its own Route, its own SCC assignment, and its own ResourceQuota.
Steps
oc new-project, each with a ResourceQuota and LimitRange sized differently (dev: tight limits, prod: generous)
oc new-app from a container image
oc expose svc/ and set up edge TLS termination on the Route
anyuid to the app's ServiceAccount — document the exact error message and fix
oc rollout status check to confirm each deploy actually completed before calling it done
oc adm policy add-scc-to-user command you needed and why
Skills Demonstrated
GitHub Repo Name
ocp-multi-env-scc-routes
---
Project 2: S2I Build + Tekton Pipeline to OpenShift
Level: Intermediate | Time: 2-3 days
Take a source repo (not a pre-built image) and get it running on OpenShift two ways: first with Source-to-Image (oc new-app ), then with a proper Tekton Pipeline that rebuilds and redeploys on every push.
Steps
oc new-app https://github.com//.git --name=myapp
oc get bc,is)
Pipeline with Tasks for clone → build (buildah) → deploy (oc set image or oc rollout restart)
EventListener + Trigger so a webhook push kicks off the PipelineRun automatically
oc set triggers to confirm a new image tag actually rolls out the Deployment
Skills Demonstrated
GitHub Repo Name
ocp-s2i-tekton-pipeline
---
Project 3: Day-2 Operations — Upgrade, Backup, and Incident Drill
Level: Advanced | Time: 4-5 days
Simulate the operational work an OCP admin actually does after the cluster is already running: a controlled upgrade, an etcd backup/restore, and a manufactured incident you diagnose and fix.
Steps
oc get co, oc get nodes, oc get mcp are all healthy before touching anything
oc adm upgrade to see available updates, then upgrade one minor version, watching clusterversion, co, and mcp roll through in order
cluster-backup.sh, then practice (on a disposable cluster only) restoring from it — document each command and what it actually restores
DEGRADED, then diagnose using oc describe mcp, node logs, and oc adm must-gather, and roll it back
Skills Demonstrated
GitHub Repo Name
ocp-day2-ops-runbook
---
Tips for Great Projects
Make it real. "Deployed a 3-tier app on OpenShift with SCC-hardened pods and a Tekton pipeline" is a sentence an interviewer will ask follow-up questions about — that's the point.
Keep the error messages. The SCC error in Project 1, the MCP degraded state in Project 3 — paste the actual oc describe output into your README. Employers care that you've seen real failures, not just happy paths.
Show the diff between OCP and vanilla Kubernetes in your writeup wherever it's relevant (Routes vs Ingress, SCC vs PSA, DeploymentConfig vs Deployment) — that comparison is exactly what OCP interviews probe for.
Portfolio Checklist
oc new-app