Auto Scaling practice questions

Refactor the Node.js frontend to run on AWS Lambda with Amazon API Gateway

Why wrong: Lambda is suitable for Node.js, but the question asks for TWO strategies; the Java backend and database are better candidates for modernization.

Migrate the database to Amazon DynamoDB

Why wrong: DynamoDB is NoSQL and would require application changes.

Deploy the Java backend on Amazon EC2 with Auto Scaling

Why wrong: EC2 requires managing servers; not as modernized as Fargate.

Configure lifecycle hooks in the Auto Scaling group to drain connections

Why wrong: Lifecycle hooks are not needed for scaling based on CPU.

Use a single large EC2 instance (e.g., c5n.18xlarge) to handle all traffic

Why wrong: Single large instance is not highly available.

Use a Network Load Balancer to distribute traffic to web servers

Why wrong: NLB is for TCP/UDP; for HTTP traffic, ALB is better.

Amazon DynamoDB

Why wrong: DynamoDB is a NoSQL database and does not support MySQL.

Amazon Redshift

Why wrong: Redshift is a data warehouse, not suited for online transaction processing (OLTP) workloads.

Amazon Aurora

Why wrong: Aurora is MySQL-compatible but is a specific engine under RDS; the broader service answer is RDS.

Use On-Demand Instances only for the batch job to ensure availability.

Why wrong: On-Demand is costlier than Spot for this workload.

Purchase Reserved Instances for the batch job capacity.

Why wrong: Reserved Instances require a 1- or 3-year commitment, not suitable for short monthly jobs.

Configure a simple scaling policy based on CPU utilization.

Why wrong: Simple scaling is reactive and adds capacity after metrics exceed thresholds.

Configure a scheduled scaling policy to add instances during known peak hours.

Why wrong: Scheduled scaling requires manual setup and does not adapt to unexpected spikes.

Configure a target tracking scaling policy based on average CPU utilization.

Why wrong: Target tracking is also reactive, maintaining a target metric.

Increase the ALB idle timeout to 120 seconds.

Why wrong: Increasing idle timeout does not address high CPU utilization.

Enable cross-zone load balancing on the ALB.

Why wrong: Cross-zone load balancing distributes traffic evenly but does not add capacity.

Reduce the target tracking scaling threshold to 50% average CPU.

Why wrong: Lowering threshold may cause frequent scaling but does not guarantee faster response to spikes.

Enable Multi-AZ for the RDS instance.

Why wrong: Multi-AZ provides failover, not read scaling.

Implement DynamoDB Accelerator (DAX) in front of the database.

Why wrong: DAX is for DynamoDB, not RDS for MySQL.

Scale up the RDS instance to a larger instance type.

Why wrong: Scaling up helps but does not fully address read-heavy workload.

Create a custom AMI that pre-populates session data from Amazon S3.

Why wrong: Session data is dynamic and cannot be pre-populated in an AMI.

Increase the Auto Scaling group's cooldown period to 600 seconds.

Why wrong: Longer cooldown delays scaling, which can cause performance issues.

Enable sticky sessions (session affinity) on the ALB.

Why wrong: Sticky sessions route traffic to the same instance but do not protect against instance termination.

Enable Multi-AZ for the cluster to improve read availability.

Why wrong: Multi-AZ does not improve read performance.

Increase the instance size of the writer instance to improve write throughput.

Why wrong: Writer size helps writes but not replica lag directly.

Refactor the application into microservices and use Amazon DynamoDB for data storage.

Why wrong: High refactoring effort; not directly reducing licensing costs.

Rehost the application on Amazon EC2 and use Amazon RDS for Oracle with license-included.

Why wrong: License-included is more expensive.

Rehost the application on Amazon EC2 and use Amazon RDS for Oracle with BYOL.

Why wrong: Still uses Oracle licensing; does not reduce costs.

Configure the ALB with an HTTPS listener using an ACM certificate. Configure the target group with HTTP health checks on port 80.

Why wrong: This does not encrypt traffic between ALB and instances.

Configure the ALB with an HTTPS listener using an ACM certificate. Configure the target group with HTTPS on port 443 using a self-signed certificate. Configure health checks to use HTTPS on port 443.

Why wrong: Self-signed certificates are not trusted by default and health checks may fail.

Configure the ALB with an HTTPS listener using an ACM certificate. Configure the target group with HTTPS health checks on port 443 using a separate ACM certificate.

Why wrong: Using a separate ACM certificate for health checks is unnecessary and could cause complexity.

Use AWS CloudFormation StackSets to deploy the Auto Scaling groups consistently.

Why wrong: StackSets are for initial deployment, not ongoing management.

Use AWS Config rules to enforce desired scaling policies.

Why wrong: Config rules are for compliance, not management.

Use Amazon CloudWatch cross-account dashboards.

Why wrong: Dashboards are read-only and cannot modify resources.

Use Amazon EC2 Auto Scaling to add more Fargate tasks.

Why wrong: EC2 Auto Scaling manages EC2 instances, not Fargate tasks; ECS Service Auto Scaling is needed.

Use AWS Auto Scaling Plans with predictive scaling.

Why wrong: AWS Auto Scaling Plans is for multi-resource scaling, but for ECS services, Application Auto Scaling with target tracking is more appropriate.

Manually adjust the desired count of tasks in the ECS service based on traffic analysis.

Why wrong: Manual scaling is not automated and does not meet the requirement for automatic scaling.

Amazon Kinesis Data Firehose to deliver events to destinations.

Why wrong: Firehose is for streaming data delivery, not event-driven routing.

AWS AppSync to provide real-time subscriptions for events.

Why wrong: AppSync is for GraphQL, not for event processing architecture.

Amazon DynamoDB

Why wrong: DynamoDB is a NoSQL database, but for session state, in-memory stores like Redis are more appropriate for low latency.

Amazon S3

Why wrong: S3 is object storage and not designed for low-latency session state access.

Amazon RDS for MySQL

Why wrong: RDS is a relational database, which adds latency and overhead compared to in-memory caching.

Amazon CloudFront

Why wrong: CloudFront is a CDN.

Amazon Route 53

Why wrong: Route 53 is DNS.

Elastic Load Balancing

Why wrong: ELB distributes traffic, does not scale instances.

Use DynamoDB Streams to reduce write capacity.

Why wrong: Streams are used for change data capture, not for reducing capacity.

Use DynamoDB global tables for multi-region replication.

Why wrong: Global tables increase cost due to replication.

Use DynamoDB Accelerator (DAX) to reduce read capacity.

Why wrong: DAX is a caching layer that reduces read latency but does not directly reduce cost of the table; it adds cost.

Increase the instance size to handle more load per instance.

Why wrong: Larger instances may still be overwhelmed; horizontal scaling is more effective.

Manually increase the desired capacity when traffic spikes.

Why wrong: Manual intervention is not automated and may not react quickly enough.

Use scheduled scaling to add instances at expected peak times.

Why wrong: Scheduled scaling is for predictable patterns, not sudden spikes.

Use AWS CloudHSM to generate a key and store it in a hardware security module.

Why wrong: CloudHSM is for FIPS 140-2 Level 3 compliance, but key rotation and management are more complex compared to KMS.

Use an AWS managed key for S3 and enable automatic rotation.

Why wrong: AWS managed keys cannot be rotated manually and do not allow customer control over rotation schedule.

Use a customer-managed CMK and generate a new key every 90 days, re-encrypting all data.

Why wrong: Re-encrypting all data every 90 days is unnecessary and operationally complex; KMS retains old backing keys.

Purchase Reserved Instances (RI) for a 1-year term.

Why wrong: RIs require upfront commitment and are not ideal for variable workloads.

Use On-Demand Instances to ensure availability.

Why wrong: On-Demand is more expensive.

Use Dedicated Hosts for compliance.

Why wrong: Dedicated Hosts are costly and unnecessary.