Serverless Image Resizer

Automatically resize uploaded images using AWS SAM, Lambda, S3, API Gateway, and CloudFront.

Architecture

User Upload → S3 Input Bucket → Lambda Function → S3 Output Bucket → CloudFront

Features

• Automatic Resizing: Creates 3 sizes (thumbnail: 150x150, medium: 500x500, large: 1200x1200)
• Multiple Formats: Supports JPEG, PNG, and other PIL-compatible formats
• Global CDN: Optional CloudFront distribution for fast image delivery
• REST API: Optional upload endpoint via API Gateway
• SAM Framework: Simplified serverless deployment
• Cross-Platform: Works on any OS without Docker

Prerequisites

1. Install SAM CLI:

   # macOS
   brew install aws-sam-cli
   
   # Or follow: https://docs.aws.amazon.com/serverless-application-model/latest/developerguide/install-sam-cli.html

2. Install AWS CLI:

   # macOS
   brew install awscli

3. Configure AWS Credentials:

   aws configure

Quick Start

1. Deploy Infrastructure:

   chmod +x *.sh
   ./deploy.sh

   The script will prompt for:

   • Project name (default: serverless-image-resizer)
   • AWS region (default: us-east-1)
   • CloudFront CDN (y/n)
   • API Gateway (y/n)

2. Note the outputs after deployment for bucket names and URLs.

3. Test the system:

   # Upload a test image
   aws s3 cp sample-images/your-image.jpg s3://YOUR-INPUT-BUCKET/
   
   # Check results
   ./check-results.sh

Usage

Method 1: Upload via AWS CLI

# Get your input bucket name from deployment output
INPUT_BUCKET="your-input-bucket-name"

# Upload an image
aws s3 cp sample-images/your-image.jpg s3://$INPUT_BUCKET/

# Check resized images in output bucket
OUTPUT_BUCKET="your-output-bucket-name"
aws s3 ls s3://$OUTPUT_BUCKET/resized/ --recursive

Method 2: Upload via API Gateway (if enabled)

# Get API endpoint from deployment output
API_ENDPOINT="your-api-gateway-url"

# Upload image via API
curl -X POST "$API_ENDPOINT?key=test-image.jpg" \
     --data-binary @sample-images/your-image.jpg \
     --header "Content-Type: image/jpeg"

Method 3: Upload via AWS Console

1. Go to S3 in AWS Console
2. Find your input bucket
3. Upload images directly through the web interface

Accessing Resized Images

Direct S3 URLs

# Replace with your output bucket name
OUTPUT_BUCKET="your-output-bucket-name"

# Access resized images
https://s3.amazonaws.com/$OUTPUT_BUCKET/resized/thumbnail/your-image.jpg
https://s3.amazonaws.com/$OUTPUT_BUCKET/resized/medium/your-image.jpg
https://s3.amazonaws.com/$OUTPUT_BUCKET/resized/large/your-image.jpg

CloudFront URLs (if enabled)

# Replace with your CloudFront domain from deployment output
CLOUDFRONT_URL="your-cloudfront-domain"

# Access via CDN (faster, cached globally)
https://$CLOUDFRONT_URL/resized/thumbnail/your-image.jpg
https://$CLOUDFRONT_URL/resized/medium/your-image.jpg
https://$CLOUDFRONT_URL/resized/large/your-image.jpg

Project Structure

├── template.yaml          # SAM template
├── src/
│   └── lambda_function.py  # Lambda function code
├── sample-images/          # Test images directory
├── deploy.sh              # Full deployment script
├── update-lambda.sh       # Quick Lambda code updates
├── check-results.sh       # Check resized images
├── cleanup.sh             # Cleanup script
└── README.md              # This file

Testing

1. Add test images to sample-images/ directory
2. Upload using any method above
3. Check results:

   ./check-results.sh

4. Verify sizes: thumbnail (150x150), medium (500x500), large (1200x1200)

Quick Results Check

After uploading images, use the results checker:

# Check if resized images were created
./check-results.sh

This will:

• List all resized images in output bucket
• Show direct S3 access URLs
• Confirm processing worked correctly

Configuration

Modify src/lambda_function.py to customize:

• Output sizes
• Image quality
• Supported formats
• Processing logic

Quick Lambda Updates

For code-only changes to src/lambda_function.py:

# Fast update (no infrastructure changes)
./update-lambda.sh

Use this instead of full deployment when:

• ✅ Modifying Lambda code only
• ✅ Testing function logic
• ✅ Avoiding SAM build time

For infrastructure changes, use ./deploy.sh

Monitoring

• CloudWatch Logs: Check Lambda function logs
• S3 Events: Monitor bucket notifications
• Lambda Metrics: View function performance

Cleanup

⚠️ Important: Run cleanup to avoid ongoing AWS charges

chmod +x cleanup.sh
./cleanup.sh

This will:

• Empty S3 buckets
• Delete CloudFormation stack
• Remove all AWS resources
• Clean up local files

Cost Estimation

• Lambda: ~$0.20 per 1M requests
• S3: ~$0.023 per GB storage
• CloudFront: ~$0.085 per GB transfer (if enabled)
• API Gateway: ~$3.50 per 1M requests (if enabled)

Troubleshooting

Common Issues

1. "No module named PIL": Layer not installed correctly

   • Re-run ./deploy.sh

2. "Access Denied": Check AWS credentials

   • Run aws configure

3. "Stack already exists": Use cleanup first

   • Run ./cleanup.sh then ./deploy.sh

Logs

# View Lambda logs
aws logs describe-log-groups --log-group-name-prefix /aws/lambda/serverless-image-resizer

# View specific log stream
aws logs get-log-events --log-group-name /aws/lambda/your-function-name --log-stream-name LATEST

Development Workflow

1. Initial deployment: ./deploy.sh
2. Test uploads: Use AWS CLI, API Gateway, or Console
3. Check results: ./check-results.sh
4. Modify Lambda code: Edit src/lambda_function.py
5. Quick update: ./update-lambda.sh
6. Test again: Upload and check results
7. Infrastructure changes: ./deploy.sh
8. Cleanup when done: ./cleanup.sh

Use Cases

• Social Media: Profile picture resizing
• E-commerce: Product image thumbnails
• Portfolios: Performance optimization
• Content Management: Automatic image processing
• Mobile Apps: Multiple resolution support