I'm working to put some best practices together for our team for developing flows, specifically taking advantage of the new record-trigger feature found in the Winter '21 release. One of the best practices I am looking at is having a single record-triggered Flow per object, so that we can control the order of execution.
However, I'm already running into a challenge with this since it seems you can create a flow for Before and After triggers, which would mean we'd need 2 flows (one before, one after) to cover those use cases. We could have a 3rd Flow for delete.
Another challenge is the it appears one cannot call other flows from a a record-triggered flow, so this means that all logic would have to be embedded into one (or 2 or 3) monolith flows, instead of calling re-useable, auto-launched flows.
So in summary, my questions are:
- What are the best practices to create record-triggered flows to control order of execution?
- Can record-triggered flows call auto-launched flows, and if not, why?