Building A Containerized Lambda Function

AWS Lambda recently added container support and we’re going to look into what this offer actually includes and how we could build a containerized Lambda function ourselves.

Start your IDEs and open your AWS consoles because we’re going in!

AWS Lambda Previous To Container Image Support

Until recently, AWS Lambda only allowed you to write some code and upload a zip-archive containing all files necessary to run your function. Some frameworks eased up this process, but that’s how it went.

Lambda also imposed some pretty strict limits, especially the 50MB default deployment size limit. You could circumvent this limit, technically, by pulling more deployment dependencies from S3 or by issuing an AWS Service Limits support request. But this wasn’t the most intuitive or clear thing to do.

This could make it pretty difficult sometimes to create Lambda functions out of a larger deployment. Especially when using data science dependencies like NumPy, pandas, or larger JS libraries, you could easily hit that limit.

What Container Image Support Includes

Container Image Support includes, well…containers. This means that you can now build an image with your favorite tool (Docker…