Skip to content

AWS & Azure Infrastructure — Phase 3 Field Guide ​

Source-reviewed 11 October 2026. These are learning and inspection exercises built from public repository files. Commands marked read-only query systems but may require permissions. Configuration examples are illustrative; they are not instructions to create paid cloud resources without a separate cost/security review.

This field guide builds on AWS Chapter 5, Terraform Chapter 7, and the Kubernetes/platform labs. Each lesson separates what the repository currently implements, what is documented as prior cloud validation, and what remains a proposed exercise.

Learning route ​

LessonOutcomeRepository evidence
01 — AWS VPC to EKSTrace private workloads, ingress, endpoints, and availability trade-offsTerraform EKS
02 — Terraform state and plansExplain S3 locking, guardrail tests, saved plans and CITerraform EKS
03 — AWS identity boundariesDistinguish GitHub OIDC, EKS Pod Identity and CognitoPharmacy API EKS
04 — Azure Private LinkDiagnose network/DNS path separately from identityAZ-104 / PropertyOps
05 — Bicep and RBACReview modules, What-If and scoped accessAZ-104 Azure Lab
06 — PropertyOps deploymentIdentify deployed vs prepared networking and CI trustPropertyOps
07 — Cross-cloud troubleshootingPractise safe hypothesis-led incident investigationAll four repositories
08 — Cost and cleanupPrevent unattended EKS, NAT, Bastion and other chargesTerraform and Azure projects
09 — Interview & reviewExplain a design with evidence, alternatives, failures and gapsAll four repositories

Working principles ​

  1. Start with scope: record account/subscription, Region, resource group, cluster, and identity before running commands.
  2. Read before changing: inspect the actual source and state before any apply, deployment, credential modification or cleanup.
  3. Do not confuse boundaries: AWS IAM, Kubernetes RBAC, application authorization and network controls address different problems.
  4. Never claim a live system solely from code: a manifest proves a declared configuration, not a running deployment.
  5. Keep durable evidence: record the commit SHA, command, time, expected result, observed result and redacted output.

Lab safety: Never paste production tokens, Terraform state, bearer credentials or connection strings into public issues, screenshots or documentation. Prefer dedicated sandbox subscriptions/accounts and explicitly approved maintenance windows.

Built from verified projects, decisions, commands and lessons.