AWS DevOps Engineer Professional

Correct — It chains source, build, test, and deploy stages into a pipeline.

Correct — It compiles, tests, and produces artifacts without managing build servers.

Correct — It is the recipe CodeBuild executes.

Correct — With in-place or blue/green strategies depending on the target.

Correct — The old (blue) stays until you cut over, enabling fast rollback.

Correct — Alarm-triggered auto-rollback protects production.

Correct — It maps files and runs validation scripts at each hook.

Correct — Stages pass artifacts via S3.

Correct — Insert a controlled checkpoint into automated flow.

Correct — The pipeline assumes a role with only the permissions it needs.

Correct — A controlled, private package source.

Correct — Catch vulnerabilities before deploy.

Correct — Repeatable, version-controlled infrastructure.

Correct — See what will be added/modified/removed to avoid surprises.

Correct — Org-wide, multi-Region provisioning.

Correct — Someone changed a resource outside CloudFormation.

Correct — Use loops, conditions, and abstractions in TypeScript/Python/etc.

Correct — Concise syntax for serverless resources.

Correct — Modularity and reuse across templates.

Correct — Prevent an update from replacing your production database.

Correct — Infra changes go through the same review as application code.

Correct — Fail the build on non-compliant infrastructure.

Correct — Automatic rollback keeps the stack consistent.

Correct — Higher-level constructs encode best practices.

Correct — A suite for managing and automating fleets.

Correct — Auditable, keyless access via the SSM agent and IAM.

Correct — Consistent, compliant patching without manual work.

Correct — Config/secrets management without hardcoding.

Correct — Repeatable, auditable automation of ops procedures.

Correct — Config history + rules = drift and compliance detection.

Correct — e.g. re-enable encryption or a public-access block automatically.

Correct — Reduce configuration drift across the fleet.

Correct — Automation is repeatable and leaves a record.

Correct — Fetch at runtime via IAM, rotate centrally.

Correct — Tag-based targeting for patching, cost, and IAM conditions.

Correct — Apply a compliance framework’s checks together.

Correct — The core monitoring service for AWS resources and apps.

Correct — Alarms drive automated response.

Correct — A query language over your ingested logs.

Correct — Find bottlenecks and errors across service calls.

Correct — Default EC2 metrics lack memory/disk; the agent fills the gap.

Correct — e.g. on a GuardDuty finding, invoke a remediation Lambda.

Correct — CloudWatch metrics/logs plus X-Ray traces.

Correct — Detect problems users feel, not just CPU.

Correct — Catch an outage before users report it.

Correct — Cross-account observability and a tamper-resistant trail.

Correct — Alert on what indicates real problems.

Correct — Self-healing systems act without waiting for a person.

Correct — Limit blast radius; watch metrics before full rollout.

Correct — Gradual replacement maintains availability.

Correct — Bad deploy trips an alarm → CodeDeploy rolls back.

Correct — Systems recover automatically from common failures.

Correct — One place to define and enforce backup policy.

Correct — Automate the failover to meet RTO/RPO.

Correct — Automated, consistent response beats improvisation under pressure.

Correct — e.g. quarantine a resource on a GuardDuty finding.

Correct — Turn the incident into durable improvements.

Correct — Limit exposure and recover fast if something breaks.

Correct — Find weaknesses proactively, not during an outage.

Correct — Automation + observability + resilience across the lifecycle.