AWS DevOps DevOps Monitoring, Testing & Deployment Strategies 4 — Questions and Answers
Question 1: Which AWS service should you use to implement synthetic monitoring that continuously tests your API endpoints from multiple geographic regions?
- Amazon CloudWatch Synthetics (Correct answer)
- AWS X-Ray sampling rules
- Amazon Route 53 health checks
- AWS CodeBuild scheduled builds
Correct answer: Amazon CloudWatch Synthetics
CloudWatch Synthetics runs Canary scripts on a schedule to simulate user interactions and monitor endpoint availability and latency.
Question 2: When using AWS CodePipeline with a manual approval action, what happens to the pipeline if no one approves or rejects within the configured timeout?
- The action times out and the pipeline fails (Correct answer)
- The pipeline continues automatically after timeout
- The previous stage is automatically rolled back
- A Lambda function is invoked to auto-approve
Correct answer: The action times out and the pipeline fails
Manual approval actions have a configurable timeout (default 7 days) after which the action expires and the pipeline execution fails.
Question 3: A team needs to run different test suites in parallel to speed up a CodePipeline stage. Which CodeBuild feature supports running multiple build environments concurrently for a single project?
- CodeBuild batch builds with parallel build graph (Correct answer)
- CodeBuild build queuing with multiple agents
- CodePipeline parallel actions within a stage
- CodeBuild fleet with auto-scaling reserved capacity
Correct answer: CodeBuild batch builds with parallel build graph
CodeBuild batch builds allow you to define a build graph where multiple build environments run in parallel, sharing artifacts.
Question 4: Which metric in Amazon CloudWatch is most useful for detecting a memory leak in a running application over time?
- Custom memory utilization metric published via CloudWatch Agent (Correct answer)
- Default EC2 MemoryUtilization metric
- CloudWatch built-in RAM metric
- VPC Flow Logs byte count
Correct answer: Custom memory utilization metric published via CloudWatch Agent
EC2 does not publish memory metrics by default; you must install the CloudWatch Agent and configure it to collect and publish memory utilization.
Question 5: In AWS Fault Injection Simulator (FIS), what is an 'experiment template' used for?
- Defining the fault actions, targets, stop conditions, and IAM permissions for a chaos experiment (Correct answer)
- Creating CloudFormation templates with injected resource failures
- Scheduling automated chaos monkey runs on EC2 instances
- Generating synthetic load to trigger auto-scaling
Correct answer: Defining the fault actions, targets, stop conditions, and IAM permissions for a chaos experiment
An FIS experiment template specifies what faults to inject, which AWS resources to target, stop conditions, and the IAM role to use.
Question 6: Which deployment configuration in Elastic Beanstalk replaces all instances simultaneously but first launches a fresh set with the new version before terminating old ones?
- Immutable deployment (Correct answer)
- Rolling deployment
- Blue/green using swap URL
- All at once deployment
Correct answer: Immutable deployment
Immutable deployments launch a new Auto Scaling group with the new version, then decommission the old group only after health checks pass.
Question 7: What does a CloudWatch Composite Alarm allow you to do that individual alarms cannot?
- Combine multiple alarm states using Boolean logic to reduce alert noise (Correct answer)
- Trigger a Lambda function when multiple metrics breach simultaneously
- Create alarms that span multiple AWS accounts in an Organization
- Set different threshold values depending on the time of day
Correct answer: Combine multiple alarm states using Boolean logic to reduce alert noise
Composite Alarms evaluate AND/OR/NOT conditions across other alarms, so you can alert only when a meaningful combination of conditions is true.
AWS DevOps DevOps Monitoring, Testing & Deployment Strategies 4
Which AWS service should you use to implement synthetic monitoring that continuously tests your API endpoints from multiple geographic regions?