I'm looking for guidelines regarding how to work with

serverless

and

pulumi

side by side Serverless does a good job at providing code boilerplate and packaging functions but Pulumi far outclasses it when it comes to infrastructure provisioning. What's the best way to make those two work together ? Should I just add a "Managed By" tag to my resources depending on wether they were deployed by Serverless or Pulumi ? How do I automatically import in my Pulumi stack resources that were created by Serverless ? Is it better to not use Serverless's deployment features altogether ?

Using the two together seems like it’d just make for extra work. I’ve used a little Serverless, but not a ton, so I’m curious where Serverless provides a better experience than Pulumi for you (assuming it’s not just about having to scrap work you’ve already done, which is totally understandable).

Personally I’ve found it easier to use whatever provisioning system I’m using for functions etc too, even if they’re developed with serverless or similar.

@miniature-musician-31262 I used Serverless to package my code & dependencies, then I use Pulumi's

sourceCodeHash

to detect changes and update Lambda accordingly One might argue that I could just

zip -r9

and package files myself, but I find that Serverless packaging utilities have nice glob pattern include/exclude mechanisms