This message was deleted.

I typically just treat pulumi like any other JS/TS project. wrap exceptions in

try/catch/finally

"dealing" with errors in

catch

accordingly. If I want to log and exit, I log what I want and process.exit. If I want to retry something, I retry. etc ... I very much prefer this to relying on uncaught exceptions to stop my program and log the "throw". Just my opinion.

tbh, I always throw. Pulumi is already handling logging and process.exit'ing, manually killing the process is unnecessary imho

that's fair!

Thanks for the insights into your workflow!