Add option to limit flow run concurrency for separate served flows

[jrbourbeau]

First check

• I added a descriptive title to this issue.
• I used the GitHub search to find a similar request and didn't find it.
• I searched the Prefect documentation for this feature.

Prefect Version

2.x

Describe the current behavior

Currently I can limit the number of concurrent flow runs when using the .serve() API with the limit= parameter:

from prefect import flow

@flow
def my_flow():
    ...
    
my_flow.serve(name="flow", limit=1)

This works great for a single flow, but when serving multiple flows I can't (to my knowledge) limit concurrent flow runs in a similarly convenient way. Setting limit= on the top-level prefect.serve(...) method limits concurrent flow runs for all the flows that are being served:

from prefect import flow, serve

@flow
def flow1():
    ...

@flow
def flow2():
    ...
    
d1 = flow1.to_deployment(name="flow1")
d2 = flow2.to_deployment(name="flow2")
serve(d1, d2, limit=1)

Describe the proposed behavior

It'd be great if I could do something like this:

from prefect import flow, serve

@flow
def flow1():
    ...

@flow
def flow2():
    ...
    
d1 = flow1.to_deployment(name="flow1", limit=1)
d2 = flow2.to_deployment(name="flow2", limit=10)
serve(d1, d2)

to independently limit the flow run concurrency for each flow being served.

Example Use

No response

Additional context

No response

[serinamarie]

Hi @jrbourbeau, thanks for creating your first issue in Prefect :) This sounds like a nice add and we'll add it to our backlog! This will require some design for the runner to separate concurrency limits between different flows.

[aaazzam]

This has been achieved in recent work, see #14934.