Dear all, I guess there are many ways to Rome, but

how does prefect suggest to set up CICD workflows to deploy flows in dev, staging and production environments

? I set up a CICD pipeline, in which I once want to check in a staging environment whether the flows run properly on the staging infrastructure, before deploying the flows to production. Does it therefore make sense to have • one work queue for each environment • one deployment for each environment Currently, I create an

{environment} deployment

and a

{environment} work queue

on

{environment} specific infrastructure

so that I can make a full integrated test and avoid that a staging flow is run on the production infrastructure in the same

general work queue

. One could of course argue that one may differentiate between "infrastructure environment" and "runtime environment" etc. but that may make things more complicated. Maybe I am also overcomplicating things and prefect has a simple, elegant solution for this? 🤔

I prefer 1 workspace per environment, of course there are cost implications however I prefer everything to be isolated and no danger of crossing the streams, also reduces confusion

Oh yeah, that makes sense