Resize Image Using AWS Lambda, S3, and SNS

This project automatically resizes images uploaded to an S3 bucket using AWS Lambda, and sends notifications via SNS when the resizing is complete.

Architecture Overview

Project Description

This Terraform project sets up a serverless image resizing pipeline that:

• Monitors an S3 bucket for new image uploads
• Triggers an AWS Lambda function automatically when images are uploaded
• Resizes images using the Pillow library
• Stores resized images in a separate S3 bucket
• Notifies users via email when the resizing is complete

Key Components

1. S3 Buckets

Two S3 buckets are created:

• Source Bucket (image-need-to-resize): Where users upload original images

  

• Destination Bucket (image-resized): Where resized images are stored

  

2. AWS Lambda Function

The Lambda function performs the image resizing:

• Runtime: Python 3.9
• Memory: 256 MB
• Timeout: 60 seconds
• Layer: Klayers-p39-pillow for image processing
• Handler: CreateThumbnail.handler

3. SNS Topic

Email notifications are sent when images are successfully resized:

• Protocol: Email
• Subscriber: Configured in terraform.tfvars

4. IAM Roles and Policies

Lambda execution role with policies to:

• Read from source S3 bucket
• Write to destination S3 bucket
• Publish to SNS topic

Prerequisites

• AWS Account with appropriate credentials
• Terraform installed (v1.0+)
• Python 3.9+ (for local testing)
• AWS CLI configured

Files Structure

.
├── providers.tf              # AWS provider configuration
├── variables.tf              # Variable definitions
├── terraform.tfvars         # Terraform variable values
├── locals.tf                # Local values
├── s3-bucket.tf             # S3 bucket resources
├── iam-role.tf              # IAM role definitions
├── iam-policy.tf            # IAM policy definitions
├── lambda-function.tf       # Lambda function resources
├── sns-topic.tf             # SNS topic and subscription
├── backend.tf               # Terraform backend configuration
├── CreateThumbnail.zip      # Lambda function code (zipped)
├── README.md                # This file
└── assets/                  # Architecture diagrams and screenshots

Configuration

1. Update terraform.tfvars

aws_access_key = "YOUR_AWS_ACCESS_KEY"
aws_secret_key = "YOUR_AWS_SECRET_KEY"
aws_region     = "us-east-1"
mail-id        = "your-email@example.com"

2. Variable Defaults

• s3-bucket-lambda-code: Source bucket for original images (auto-suffixed with account ID)
• s3-bucket-dest: Destination bucket for resized images (auto-suffixed with account ID)
• sns-name: SNS topic name
• mail-id: Email address for notifications

Deployment

Initialize Terraform

terraform init

Plan the deployment

terraform plan

Apply the configuration

terraform apply

When prompted, type yes to confirm.

Usage

1. Upload an image to the source S3 bucket (non-resized-image-[account-id])

2. Lambda automatically triggers and resizes the image

3. Receive an email notification with a link to the resized image in the destination bucket

4. Access the resized image from the destination bucket (resized-images-[account-id])

Lambda Function Logic

The CreateThumbnail.handler function:

1. Receives S3 event when image is uploaded
2. Downloads the original image from source bucket
3. Resizes the image to a thumbnail
4. Uploads the thumbnail to destination bucket
5. Publishes a message to SNS topic with the result

Cleanup

To remove all resources created by this project:

terraform destroy

Type yes when prompted to confirm deletion.

Security Considerations

• Store AWS credentials securely (use AWS IAM roles when possible)
• Restrict S3 bucket access to Lambda function only
• Use encryption for sensitive data in environment variables
• Monitor Lambda execution logs in CloudWatch

Troubleshooting

Lambda function fails

Check CloudWatch Logs:

aws logs tail /aws/lambda/Image-Resizing-Lambda --follow

Email not received

1. Confirm SNS subscription by clicking the link in the confirmation email
2. Check spam/junk folder
3. Verify the email address in terraform.tfvars

S3 bucket name already exists

The bucket names are auto-suffixed with your AWS account ID. If you still get conflicts, modify the locals.tf file to add a random suffix.

Cost Estimation

• Lambda: ~$0.20 per 1 million requests
• S3: ~$0.023 per GB stored
• SNS: ~$0.50 per 1 million notifications
• Data Transfer: Varies by region

License

This project is provided as-is for educational and demonstration purposes.

Support

For issues or questions, please refer to the AWS documentation:

• AWS Lambda Documentation
• Amazon S3 Documentation
• Amazon SNS Documentation
• Terraform AWS Provider