Architecting on AWS Certification Welding Procedures & Techniques 5 — Questions and Answers
Question 1: An application experiences database bottlenecks due to read-heavy traffic. Which approach offloads reads from the primary RDS instance?
- Increase the primary instance's storage IOPS
- Add RDS Read Replicas and direct read traffic to them (Correct answer)
- Enable Multi-AZ to distribute reads across AZs
- Use RDS Proxy to cache query results
Correct answer: Add RDS Read Replicas and direct read traffic to them
Read Replicas serve read queries from a copy of the database, reducing load on the primary writer instance.
Question 2: A web application needs to serve static assets globally with the lowest possible latency. Which combination is optimal?
- S3 static website hosting with no CDN
- EC2 instances deployed in every Region
- S3 as origin with Amazon CloudFront distribution (Correct answer)
- EFS mounted across all EC2 instances globally
Correct answer: S3 as origin with Amazon CloudFront distribution
CloudFront caches S3 objects at 400+ edge locations worldwide, delivering assets to users from the nearest point of presence.
Question 3: An application requires session state to be shared across multiple EC2 instances behind a load balancer without sticky sessions. What is the recommended approach?
- Store session data on each instance's local disk
- Use Amazon ElastiCache (Redis) to store session state centrally (Correct answer)
- Enable sticky sessions on the ALB
- Save session data in an S3 bucket per user
Correct answer: Use Amazon ElastiCache (Redis) to store session state centrally
ElastiCache Redis provides a fast, centralized session store accessible by all instances, enabling stateless scaling.
Question 4: Which AWS Auto Scaling policy type scales based on a predictable, time-based traffic pattern such as business hours?
- Target Tracking Scaling
- Step Scaling
- Scheduled Scaling (Correct answer)
- Predictive Scaling
Correct answer: Scheduled Scaling
Scheduled Scaling lets you pre-define scale-out and scale-in actions at specific times to match known traffic patterns.
Question 5: A microservices application has one slow service that causes cascading failures across other services. Which pattern prevents this?
- Retry with exponential backoff on every service
- Circuit Breaker pattern implemented in each service client (Correct answer)
- Increase the timeout on all upstream services
- Deploy each service on a Dedicated Host
Correct answer: Circuit Breaker pattern implemented in each service client
A circuit breaker stops sending requests to a failing service, giving it time to recover and preventing cascading failures.
Question 6: Which Amazon CloudWatch feature allows you to detect unusual spending or API call volume automatically without setting manual thresholds?
- CloudWatch Dashboards
- CloudWatch Anomaly Detection (Correct answer)
- CloudWatch Contributor Insights
- CloudWatch ServiceLens
Correct answer: CloudWatch Anomaly Detection
CloudWatch Anomaly Detection uses ML to establish a baseline and alert when metrics deviate unexpectedly.
Question 7: An architect wants to ensure an Aurora cluster can handle a sudden 10x traffic spike. Which Aurora feature enables instant read capacity scaling?
- Aurora Multi-Master
- Aurora Auto Scaling for Read Replicas (Correct answer)
- Aurora Serverless v1 pause/resume
- Aurora Parallel Query
Correct answer: Aurora Auto Scaling for Read Replicas
Aurora Auto Scaling automatically adds or removes Read Replicas based on CPU or connection metrics to handle traffic spikes.
Architecting on AWS Certification Welding Procedures & Techniques 5
An application experiences database bottlenecks due to read-heavy traffic.
Which approach offloads reads from the primary RDS instance?