Hi Team, I have been going through the serverless ...
# general
m
Hi Team, I have been going through the serverless stack guide. and I had been using AWS CDK for a while now. what I have not used before is Serverless Framework. Now I see that CDK, as well as SST, is equipped enough to build even the lambda API infrastructure. I was going through the Best Practices section and here is my question: • Should I explore Serverless Framework for (only) lambda API infrastructure or • Should I initialize SST app(s), as separate services for my (Lambda API microservice infrastructure ) ◦ Meaning I use SST/CDK for importing cross-references. (like dynamodb cognito) ◦ Will the "2 repo" design still hold good? (I liked it.) • Is there a way to create nested stacks in SST?
f
Hey @Mr.9715, we haven’t done a good job updating the guide to use SST. Now with the Lambda support and Live Lambda Dev, you don’t need Serverless Framework. And 1 SST app for both infrastructure and Lambda code is the recommended approach.
You can still have services, each will just be a stack in the same SST app.
SST doesn’t support nested stack. There is an open discussion on the topic - https://github.com/serverless-stack/serverless-stack/discussions/181
m
Thanks, @Frank.
a
Server less framework is great and simple to use compare to SAM, but we are in a time of CDK, so server less framework is pretty much deprecated... somehow. Sad, but this is the truth
t
if you are sitting on AWS only - it's a no brainer - just use CDK and SST - forget about Serverless framework 😄