OpenShift Administrator (EX280) - 6-Week Practice Plan¶
This plan assumes 1-2 hours of theory + 2-3 hours of hands-on lab per day. EX280 is performance-based - you cannot pass on theory alone.
Lab setup (do this first)¶
Choose one path:
- OpenShift Local (CRC) - free single-node OpenShift on your laptop (download)
- Developer Sandbox - free 30-day shared cluster (sandbox)
- ROSA / ARO - paid managed clusters; usable for prep if you minimize uptime
Verify:
oc whoami
oc whoami --show-server
oc get nodes
oc get co
Week 1 - Cluster basics + Projects + RBAC¶
Reading¶
- README.md full read
- fact-sheet.md skim
- notes/01-cluster-install-config.md
- notes/02-projects-quotas-rbac.md
Hands-on¶
- Inspect cluster:
oc get co,oc get nodes,oc get clusterversion - Cordon/drain/uncordon a node
- Create a project, switch into it, create a deployment, scale it
- Create a ResourceQuota and LimitRange
- Add an htpasswd user, edit OAuth, log in as that user
- Grant alice
adminonmyapp, verify withoc auth can-i
Self-check¶
- Can I add an htpasswd user end-to-end without notes?
- Can I write a ResourceQuota and LimitRange YAML?
- Do I know the difference between
cluster-admin,admin,edit,view?
Week 2 - Workloads¶
Reading¶
- fact-sheet.md - workloads section
- OpenShift docs: Building applications
Hands-on¶
-
oc new-appfrom Git repo -
oc new-appfrom container image - Scale, rollback, set image, set env, set resources
- Configure HPA:
oc autoscale - Use
oc rollout(status, undo, restart, history) - Diagnose a pod stuck in CrashLoopBackOff using
oc logs -p
Self-check¶
- Can I deploy an app, expose it, and update its image rolling-restart-style?
- Do I know how to override container resource requests/limits?
Week 3 - Networking and Routes¶
Reading¶
Hands-on¶
- Expose a Service as a ClusterIP, NodePort, LoadBalancer
- Create a Route (HTTP, edge TLS, passthrough TLS, reencrypt TLS)
- Path-based routing on the same hostname
- Canary route weighting (90/10)
- Default deny NetworkPolicy + allow specific source
- Allow from a different namespace
- Test that NetworkPolicy actually blocks (use
oc rshfrom a pod that should be denied)
Self-check¶
- Can I write a NetworkPolicy from memory for a "deny all, allow frontend" pattern?
- Do I know the four TLS termination types for Routes?
Week 4 - Storage¶
Reading¶
Hands-on¶
- List storage classes, identify the default
- Mount a PVC into a deployment using
oc set volume - Create PVC YAML directly with specific access mode and size
- Set a different default StorageClass
- Expand a PVC and verify size grew (after pod restart)
- (If on cluster with NFS) statically provision an NFS PV
- Diagnose a Pending PVC (
oc describe pvc)
Self-check¶
- Can I write a PVC YAML with specific access mode and storage class?
- Do I know which access modes are supported by EBS / Azure Disk / NFS?
- Can I expand a volume and have the pod see the new size?
Week 5 - Operators, Monitoring, Identity Providers¶
Reading¶
- notes/05-operators-monitoring.md
- OpenShift docs: Operators
Hands-on¶
- Install an operator from OperatorHub (e.g., OpenShift Pipelines)
- Verify the operator reaches
SucceededCSV - Manual approval of an InstallPlan
- Enable user-workload monitoring via cluster-monitoring-config
- Configure HTPasswd identity provider (multiple users)
- Configure a cluster admin via
add-cluster-role-to-user
Self-check¶
- Can I install an operator from raw YAML (Subscription + OperatorGroup)?
- Can I enable user workload monitoring?
- Can I add and remove htpasswd users on an existing IDP?
Week 6 - Security, Troubleshooting, Mock Exams¶
Reading¶
- fact-sheet.md - SCC and Security sections
- scenarios.md - work through all 15
Hands-on¶
- Add
anyuidSCC to a service account - Diagnose a pod failing on SCC ("unable to validate against any security context constraint")
- Run
oc adm must-gather - Run
oc debug node/<name>andchroot /host - Recover a pod stuck in
Pendingdue to PVC issues - Recover from broken NetworkPolicy
Mock exams¶
- Sander van Vugt's EX280 practice exam (book or video course)
- scenarios.md - aim to do all 15 in 2 hours
- Score 80%+ on a clean lab session before scheduling
Schedule the exam¶
- Book Red Hat Individual Exam (online) or testing center
- Confirm 3-hour slot
- Plan exam-day logistics
Daily routine (suggested)¶
| Time | Activity |
|---|---|
| 30 min | Read notes / fact-sheet section |
| 90 min | Hands-on lab on the day's topic |
| 15 min | Verify with reboot/rollout, write a one-line cheat sheet |
Stop signals (when you're ready)¶
- You can do all 15 scenarios in scenarios.md without notes in under 2 hours
- You can configure HTPasswd identity provider end-to-end from memory
- You can write a Deployment + Service + Route + NetworkPolicy YAML stack from memory
- You can install an operator from raw YAML (no console)
- You can debug a stuck pod / PVC / route to root cause without help
- You've watched at least 80% of an EX280-focused video course (Sander van Vugt or RH364)