server less architecture implementation on AWS

Question

I am going through serverless architecture documents and some videos on AWS.

I understand we can implement serveless with API gateway , Lambda , DynamoDB and have our front end on S3 bucket.

I have brief idea how it works on legacy EC2 instance.
Example with .net application with reactjs.

In instance we install all the dependencies like .net sdk react ..ect Then we place our source code file within the instance and run the respective execution file. (with .net it would be .dll file)

Now the question is: How can this be implemented on serverless?

where should i have the .dll file?

where should all the dependencies should be installed?

and how the architecture would look?

Answer 1

This really depends on your app, S3 will only serve static HTML so this is entirely upto your application whether it can run its frontend from static HTML files.

For dynamic part of this, you would use API Gateway (calling a Lambda function) to perform any backend calls. These would allow the standards flows your application is accustomed to but is a lot of rewrite if your applications are tangled up heavily between the frontend and backend.

Lambda currently only supports .Net Core so if you're using a legacy version I am not sure you'll be able to run this as a Lambda (you can use Lambda Layers to build support for languages that are supported, but as far as I am aware .Net could only be installed on Windows).

If you are really keen to run your application on serverless without the major rewrite required I would advise looking at migrating your application to run as a container and then run it using Fargate.

Then once this is up and running start migrating/rewriting the backend to work on Lambda, you would then eventually be able to move the frontend to S3 as it becomes detangled from the backend.