Answer: Use AWS Application Auto Scaling with target tracking policies to scale when ECS metric breaches trigger an Amazon CloudWatch alarm.

## Explanation **Correct Answer: D** - Use AWS Application Auto Scaling with target tracking policies to scale when ECS metric breaches trigger an Amazon CloudWatch alarm. ### Why Option D is Correct: 1. **AWS Application Auto Scaling** is specifically designed for scaling AWS services like Amazon ECS, Amazon EC2 Spot Fleets, Amazon DynamoDB, and Amazon Aurora. It provides automatic scaling for these services based on CloudWatch metrics. 2. **Target Tracking Scaling Policies** are the most efficient way to scale ECS services with Fargate launch type. These policies: - Automatically adjust the number of tasks to maintain a specific target metric value (like CPU utilization at 70%) - Scale out when the metric is above the target - Scale in when the metric is below the target - Automatically create CloudWatch alarms and manage them 3. **Fargate Launch Type Compatibility**: Since the company is using Fargate launch type, EC2 Auto Scaling (Options A and C) is not appropriate because Fargate abstracts away the underlying EC2 instances. With Fargate, you scale tasks, not EC2 instances. 4. **Cost Optimization**: Target tracking policies automatically scale in when utilization decreases, which directly addresses the requirement to "reduce costs when utilization decreases." ### Why Other Options are Incorrect: **Option A**: Incorrect because: - EC2 Auto Scaling is for scaling EC2 instances, not ECS tasks - Fargate doesn't use EC2 instances that you manage - Scheduled scaling based on previous patterns may not respond to actual utilization changes **Option B**: Incorrect because: - While Lambda could be used, it's unnecessarily complex - AWS Application Auto Scaling is the native, managed service designed specifically for this purpose - Requires custom code and maintenance compared to a managed service **Option C**: Incorrect because: - EC2 Auto Scaling is for EC2 instances, not ECS tasks - Simple scaling policies are less sophisticated than target tracking - Doesn't work with Fargate launch type ### Key AWS Services Mentioned: - **Amazon ECS with Fargate**: Serverless compute engine for containers - **AWS Application Auto Scaling**: Service for scaling various AWS resources - **Amazon CloudWatch**: Monitoring service for collecting metrics and setting alarms - **Target Tracking Scaling Policies**: Policy type that maintains a specific metric value This solution provides automatic, responsive scaling based on actual utilization, ensuring optimal performance during high traffic while minimizing costs during low utilization periods.