Deployment pipelines: how do you handle server and database differences per stage?

Hi all,

We promote semantic models from dev to test to prod with deployment pipelines, and I am trying to get the environment-specific settings (server, database, file paths) into a clean pattern.

Right now I am leaning towards Power Query parameters for every environment-specific value, with parameter rules in the pipeline so each stage points at its own source. Is that what you would do, or are data source rules less hassle in practice?

Two things I keep hearing about and would like to confirm: do you have to re-enter credentials after each deployment, and what happens to incremental refresh partitions on the first refresh in a new stage?

Also curious whether you automate promotion with the REST API or Azure DevOps instead of clicking through the UI.

Thanks!