Learning Outcomes And Review Rubric¶
Use this guide to evaluate whether the curriculum teaches real Azure skills instead of only showing copy-and-paste commands.
Learner Outcomes¶
| Phase | Lessons | Learner can demonstrate |
|---|---|---|
| Workstation readiness | 090 | Optionally install and verify the full Windows contributor toolchain |
| Terraform foundation | 100-160 | Evaluate Terraform locally, verify Azure readiness without paid resources, then run the first deployment and learn state/locking |
| Network foundation | 170-180 | Design a VNet, subnets, address ranges, and NSG rules with explainable intent |
| Windows compute | 190-270 | Deploy Windows VMs, bootstrap IIS, use Bastion, configure load balancing, and scale VMSS workloads |
| Platform services | 280-350 | Add NAT Gateway, DNS, remote state, Key Vault, and private endpoints |
| Operations | 360-390 | Add monitoring, CI validation, Azure SQL private access, and a final reference architecture |
Review Rubric¶
| Area | Strong evidence | Needs work |
|---|---|---|
| Learning clarity | The lesson explains what is being built, why it matters, and how to verify it | Steps work but the reason behind them is unclear |
| Safety | Secrets, public access, cleanup, and costs are called out before risky commands | Learners could create resources without understanding exposure or cost |
| Terraform quality | Code is formatted, validated, named consistently, and split only where it helps readability | Code works once but is hard to review or repeat |
| Azure accuracy | Resources match current Azure patterns and avoid obsolete defaults | Resource choices are unexplained or likely to drift quickly |
| Windows path | PowerShell commands are first-class and tested on the documented workstation path | Windows users must translate from another shell |
| Documentation | Diagrams, runbooks, validation, and cleanup sections are aligned with the code | README content and Terraform behavior disagree |
| Operations | CI, linting, scanning, and dependency updates are documented and repeatable | Quality checks exist but are difficult to run or interpret |
Evidence Checklist¶
Before calling a lesson ready, verify:
- The lesson README names the goal and expected result.
- The Terraform files pass
terraform fmt -check. - Lockfile-readonly, backend-free static validation passes; Azure plans remain explicit local/manual actions.
- The sample variables avoid real names, real credentials, and private tenant data.
- The cleanup path is clear and tested.
- Catalog-generated previous/next navigation is correct.
- Any diagram matches the deployed resource shape.
Phase Review Questions¶
- Can a learner explain what changed from the previous phase?
- Can a learner predict what Azure resources will be created before running
terraform plan? - Can a learner identify which resources cost money while idle?
- Can a learner explain which traffic is public, private, inbound, and outbound?
- Can a learner recover safely from a partial apply or failed destroy?