AWS FFmpeg Guide 🚀

This repository contains scripts and examples to accompany the article Running FFmpeg on AWS Spot Instances. It demonstrates how to set up AWS, install FFmpeg, and process videos at scale while keeping costs low. Read more on the blog How to Run FFmpeg on AWS Spot Instances for Scalable, Low-Cost Video Processing.

📖 Overview

In this repo, you’ll find:

• Scripts for installing FFmpeg, uploading/downloading files to S3, and running FFmpeg jobs.
• Infrastructure templates (IAM policy + EC2 user-data script).
• Examples for Spot Instances and AWS Lambda usage.
• Step-by-step guides for transcoding workflows.

This makes it easier to reproduce the guide in practice and adapt it for your own workflows.

---

🗂 Repository Structure

aws-ffmpeg-guide/
│
├── README.md                  # Main documentation (this file)
├── scripts/                   # Helper scripts for setup and jobs
│   ├── install_ffmpeg.sh
│   ├── upload_to_s3.sh
│   ├── download_from_s3.sh
│   └── run_ffmpeg.sh
│
├── examples/
│   ├── transcode_example.md   # Walkthrough: uploading, processing, saving
│   ├── transcode_example.sh   # Single script for processing
│   └── lambda_thumbnail/      # AWS Lambda example for lightweight FFmpeg tasks
│       └── index.js
│
├── infrastructure/
│   ├── ec2_spot_user_data.sh  # Auto-install FFmpeg, AWS CLI on EC2 launch
│   └── iam_policy.json        # Example IAM role for S3 access
│
├── .gitignore
└── LICENSE

---

⚡ Quickstart

1. Launch an EC2 Spot Instance

From the AWS Console → EC2 → Launch Instance → choose Ubuntu 22. Enable Spot Instance under Advanced details → Purchasing option.

2. Connect via SSH

ssh -i "your-key.pem" <username>@<your-instance-public-dns>

3. Clone this repo to the instance

git clone https://github.com/your-repo-name/your-repo.git
cd your-repo

4. Install FFmpeg

⚠️ If you used the user data file from /infrastructure, then you can skip this step.

chmod +x scripts/install_ffmpeg.sh
./scripts/install_ffmpeg.sh

4. Upload Video to S3

./scripts/upload_to_s3.sh sample.mp4 your-bucket-name

5. Download & Process

./scripts/download_from_s3.sh your-bucket-name sample.mp4
./scripts/run_ffmpeg.sh sample.mp4 output.webm
./scripts/upload_to_s3.sh output.webm your-bucket-name

---

☁️ Spot Instances vs Lambda

• AWS Lambda → best for small jobs (e.g. thumbnails, short clips).
• EC2 Spot Instances → best for heavy, long-running jobs.

Rule of thumb:

• Quick & lightweight → Lambda
• Large & restartable → Spot

See examples/lambda_thumbnail/ for a Lambda thumbnail function.

---

🛡 Handling Spot Interruptions

• Break videos into smaller chunks (-ss and -t with FFmpeg).
• Store partial results in S3 immediately.
• Orchestrate jobs with SQS or Step Functions for scalability.

---

🔮 Next Steps

• Try different EC2 instance families (C5 for CPU, G4dn for GPU).
• Automate pipelines with SQS/Step Functions.
• Explore CE.SDK if you need an in-browser editor to complement backend processing.

---

📜 License

MIT License. Free to use and adapt.

---

🙌 Credits

Authored as a companion repo for the guide on Running FFmpeg on AWS Spot Instances.