After updating to 2.8 we found that each sub flow in a parent flow no longer scales to separate Kubernetes pods. Any idea on why this is happening?

subflow as in you are calling run_deployment?

Subflow as being called in another flow in a deployment. Triggering the deployment through Quick run doesn't scale the subflow

have you seen https://github.com/PrefectHQ/prefect-recipes/blob/main/flows-advanced/parent-orchestrator/pokemon_weight.py#L59

Yes, but the behavior here without using run_deployment and just invoking flow within another flow changed in 2.8, in 2.7 each subflow within a flow occupies a k8s pod by itself.

Invoking a flow in another flow has never spawned new infrastructure

So basically every flow will only run on a single kubernetes pod it was created on?

Unless you do

run_deployment

, yeah.

We would like to make some developmental contributions here to allow for that, is there someone we should talk to in the Prefect team to make sure how we can best work that out?

How best to work out having subflows automatically run in new pods?

Yes

I'm the best one to start talking to for something like that.

We meant to do some development in this area and would be happy to potential make a PR. Just want to make sure it's not conflicting with some internal roadmap