AWS DevOps Engineer Professional Course Notes

Automate the full release path: orchestrate with CodePipeline, execute builds/tests with CodeBuild, manage packages/images/artifacts correctly, and deploy safely across EC2, ECS, and Lambda with health validation and rollback.

• Using CodeBuild as the pipeline orchestrator or EventBridge as a build engine.

• Identify the pipeline stage and failure boundary first: source, build/test, artifact, approval, deploy, verify.

Provision and govern infrastructure as code across lifecycles, accounts, and fleets using CloudFormation/SAM/CDK, StackSets, Organizations/Control Tower, Service Catalog, and Systems Manager automation.

• Calling a template and stack the same thing.

• For CloudFormation lifecycle questions ask: preview, update protection, deletion protection, retention, drift, or failure evidence?

Design for failure and variable demand: distribute across failure domains, scale the constrained layer using meaningful metrics, and automate/test recovery against RTO and RPO.

• Sending reads to a traditional RDS Multi-AZ standby.

• Translate the exact business requirement into failure scope, demand pattern, RTO, and RPO before choosing services.

Collect useful telemetry, convert it into reliable signals, analyze incidents with the right observability tool, and automate responses safely across accounts and Regions.

• Calling 3-of-5 alarm logic 'hysteresis' instead of configuring M/N.

• Identify the telemetry type first: metric, log, trace, client experience, synthetic result, API audit, or config history.

Route events, retain delivery failures, automate safe remediation, and troubleshoot using service-native evidence before making speculative changes.

• Using archive as DLQ.

• Separate event routing, delivery-failure handling, and remediation execution—they are distinct concerns.

Apply least-privilege identity at scale, automate layered protection and data controls, and know exactly which AWS service provides each type of security evidence.

• Expecting SCPs to grant access.

• Classify the problem first: identity authorization, network/web enforcement, encryption/secret lifecycle, API audit, config compliance, network flow, vulnerability, threat, external access, or aggregation.