What is docker-lambda?

docker-lambda was a project that let you run and test AWS Lambda functions on your own computer before deploying them to Amazon. AWS Lambda is a cloud service where you upload a small piece of code and Amazon runs it for you, charging per execution rather than per server. The catch is that Lambda runs in a specific controlled environment, so code that works on your laptop might behave differently in the cloud. docker-lambda solved this by packaging that exact Lambda environment inside Docker containers that you run locally. The README opens with a deprecation notice. This project is now archived, and AWS itself publishes official replacement images. The README links to those replacements, so if you are looking for this type of tool today, you would use the official AWS images rather than this project. When it was active, you could mount your function code into the Docker container and run it as if Lambda were executing it. The container matched the Lambda environment so closely that file paths, installed libraries, environment variables, and even the user account running the process were identical to what you would get on AWS. You could use this to test your function logic, verify that native compiled dependencies would link correctly on Lambda, and catch environment-specific bugs before a deployment. The project also supported a mode where the container stayed running and exposed a local HTTP API, so you could invoke your function repeatedly without the startup cost of launching a new container each time. There was also a watch mode that detected file changes and reloaded your function automatically, useful when you are iterating quickly on code. Build variants of the images included extra tools for compiling code and packaging functions using common build systems. Docker images were provided for most Lambda runtimes: Node.js, Python, Java, Go, Ruby, and others.