portfolio

☁️ Infrastructure Cost Optimization

AWS Python Terraform Lambda

Automated cost-saving architecture on AWS that reduced monthly infrastructure expenses by 20%

📋 Problem Statement

Rising AWS costs in non-production environments due to:

💡 Solution: “Budget Police”

An intelligent, serverless system that automatically manages AWS resource states based on predefined schedules and budget thresholds.

Key Features

🏗️ Architecture

```mermaid graph LR A[EventBridge Rule] –>|Scheduled Trigger| B[Lambda Function] C[CloudWatch Alarms] –>|Budget Alert| B B –>|boto3| D[EC2 Instances] B –>|Logs| E[CloudWatch Logs] F[Terraform] –>|Provisions| A F –>|Provisions| B F –>|Provisions| C

style B fill:#FF9900
style F fill:#7B42BC
style D fill:#FF9900 \`\`\`

How It Works

  1. EventBridge triggers Lambda function on schedule (e.g., 6 PM daily)
  2. Lambda function queries EC2 instances with specific tags
  3. Boto3 SDK stops non-production instances to save costs
  4. CloudWatch monitors budget and sends alerts
  5. Morning schedule automatically restarts instances for business hours

📁 Project Structure

``` infrastructure-cost-optimization/ ├── src/ │ └── budget_police/ │ └── lambda_function.py # Main Lambda handler └── terraform/ └── modules/ └── scheduler/ └── main.tf # Infrastructure as Code ```

🚀 Setup & Deployment

Prerequisites

Installation Steps

  1. Clone the repository ```bash git clone https://github.com/Abhinandansinha01/portfolio.git cd infrastructure-cost-optimization ```

  2. Configure Terraform variables ```bash cd terraform/modules/scheduler

    Edit variables in main.tf or create terraform.tfvars

    ```

  3. Deploy infrastructure ```bash terraform init terraform plan terraform apply ```

  4. Tag your EC2 instances ```bash

    Tag instances you want to manage

    aws ec2 create-tags –resources i-1234567890abcdef0
    –tags Key=AutoStop,Value=true Key=Environment,Value=dev ```

Configuration

Lambda Environment Variables

Variable Description Example
SCHEDULE_TAG Tag key to identify managed instances AutoStop
ENVIRONMENT_TAG Environment filter dev,staging
STOP_TIME Time to stop instances (UTC) 18:00
START_TIME Time to start instances (UTC) 08:00

EventBridge Schedule

Edit the cron expression in main.tf: ```hcl schedule_expression = “cron(0 18 ? MON-FRI)” # Stop at 6 PM weekdays ```

📊 Cost Savings Results

Metric Before After Improvement
Monthly EC2 Costs $500 $400 20% reduction
Idle Instance Hours 720 hrs/month 240 hrs/month 67% reduction
Manual Interventions 15/month 0 100% automation

Calculation Example

For a t3.medium instance ($0.0416/hour):

🔧 Lambda Function Logic

The lambda_function.py implements:

```python def lambda_handler(event, context): # 1. Get current time and action (start/stop) # 2. Query EC2 instances with AutoStop tag # 3. Filter by environment (dev, staging) # 4. Execute stop/start action # 5. Log results to CloudWatch # 6. Send SNS notification (optional) ```

Key Features in Code

🛡️ IAM Permissions Required

The Lambda function needs:

```json { “Version”: “2012-10-17”, “Statement”: [ { “Effect”: “Allow”, “Action”: [ “ec2:DescribeInstances”, “ec2:StopInstances”, “ec2:StartInstances”, “ec2:DescribeTags” ], “Resource”: “” }, { “Effect”: “Allow”, “Action”: [ “logs:CreateLogGroup”, “logs:CreateLogStream”, “logs:PutLogEvents” ], “Resource”: “arn:aws:logs:::” } ] } ```

📈 Monitoring & Alerts

CloudWatch Metrics

SNS Notifications

Configure SNS topic for:

🔄 Future Enhancements

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

📄 License

This project is open source and available under the MIT License.

👤 Author

Abhinandan Sinha

Built with ❤️ for cost-conscious cloud engineers
This site is open source. Improve this page.