ImageResizer AWS Lambda

A high-performance, serverless image processing service built with AWS Lambda, TypeScript, Sharp, and AWS SDK v3.
The function dynamically resizes, converts, rotates, and fits images stored in Amazon S3, and returns a redirect URL to the processed result.

The project is fully optimized with serverless-esbuild, uses Amazon SNS for error notifications, and integrates AWS X-Ray for distributed tracing of S3/SNS calls.

---

Project Overview

ImageResizer_AWS_Lambda is a scalable serverless image transformation engine.
It retrieves images from S3, processes them using Sharp, and stores output images back into S3 under a /resized/ directory.
All functions are exposed through API Gateway using Lambda proxy integration.

This architecture requires no servers, scales automatically, and is cost-efficient — suitable for:

• Dynamic CDN image resizing
• Thumbnail generation
• Automated image pipelines
• On-demand image transformation APIs

---

Tech Stack

Languages & Frameworks

• Node.js (18+)
• TypeScript
• Sharp — high-performance image manipulation
• Serverless Framework

AWS Services

• AWS Lambda — compute engine
• Amazon S3 — original + processed image storage
• Amazon SNS — error alert delivery
• AWS X-Ray — tracing for S3 / SNS calls
• Amazon API Gateway — request routing

Build & Packaging

• serverless-esbuild for fast bundling
• AWS SDK v3 modular client for S3 & SNS

Logging

• Custom logger
• CloudWatch Logs integration

---

AWS diagram

✨ Main Features

1. Resize Images

Resize to a given width while maintaining aspect ratio.

GET /resizeImages/{files}/{width}

Output filename format:

resized/resizeImages_<width>_<filename>

---

2. Convert to Specific Width & Height

Resizes to an exact W × H using fit: fill.

GET /convertSizeImages/{files}/{width}/{height}

resized/convertSizeImages_<W>_<H>_<filename>

---

3. Rotate Images

Rotate by any angle (0–359).

GET /rotateImages/{files}/{angle}

resized/rotateImages_<angle>_<filename>

---

4. Fit Images Within Dimensions

Resizes the image to fit within W × H using fit: contain.

GET /fitSizeImages/{files}/{width}/{height}

resized/fitSizeImages_<W>_<H>_<filename>

5. Cached Output Optimization

For each transformation:

• Checks existence via HeadObjectCommand
• Skips processing if already generated
• Improves speed & reduces Lambda cost

---

6. Error Handling & SNS Notification

If S3 retrieval fails:

• Lambda logs the event
• SNS sends a notification to the registered phone number

Useful for production monitoring.

---

7. AWS X-Ray Tracing Enabled

Instrumented using:

AWSXRay.captureAWSv3Client(new S3Client(...))

Captures subsegments for:

• S3 GetObject
• S3 PutObject
• S3 HeadObject
• SNS Publish

---

Prerequisites

1. Required System Tools

• Node.js 18+
• npm or yarn
• Serverless Framework

  npm install -g serverless

---

2. AWS Credentials

Configure via AWS CLI or env variables:

export AWS_ACCESS_KEY_ID=xxxx
export AWS_SECRET_ACCESS_KEY=xxxx
export AWS_REGION=ap-northeast-2

Or located at:

~/.aws/credentials

---

3. Required Environment Variables

Variable	Description
bucket	S3 bucket name for image storage
redirect	Base redirect URL after processing
phone	Phone number for SNS alerting
message	Message body used by SNS
access	Optional custom AWS key
secret	Optional custom AWS secret

---

4. AWS Resources Needed

• S3 bucket with images/ folder
• SNS publish permissions
• Lambda execution role for S3 + SNS
• API Gateway (auto-generated by Serverless)

---

Build automation with CircleCI and Docker

1. How to configure docker configuration for build

• See the Dockerfile in this repository or the bash script below

docker run -it  amazonlinux:2023 /bin/bash

yum upgrade -y
yum update -y
yum install -y gcc-c++ make  
  
curl -sL https://rpm.nodesource.com/setup_18.x | bash -  
dnf install nodejs -y
yum install -y nodejs  
npm install -g serverless  
npm install -g npm@11.6.4
npm install web3 --unsafe-perm=true --allow-root  

yum install git -y

serverless config credentials --provider aws --key <ACCESS_KEY> --secret <SECRET_KEY>

• Read .circleci/config.yml from CircleCI connected to this Git Repository and build & deploy

2. build and deploy automation diagram

3. local test

• Since serverless-offline is installed
• After setting the environment variable of process.env locally
• If you enter the following command, you can launch Lambda in Local and test it.

sls offline start

How to check at AWS call traces

• First, Go to Lambda console monitoring, configuration tap and let X-Ray avaliable

• If you have enabled Lambda X-Ray, you can see the API call trace and API response time as below.