If I have a flow run in scheduled state with one parameter, then i try to run that flow again with a different param, why does it cancel the first flow run?

I’ll try to replicate. Is the process doing a Run with one parameter, and then clicking the Run button and queueing another one with a different value?

im playing with having a lambda trigger a flow run, and my code looks something like

flow_run = StartFlowRun(flow_name="ingest-refresher-single", project_name="meow")
    flow_run.run(parameters=dict(id=id))

where id is some parameter the lambda get. I didn’t have my agent running, and I saw that invoking this lambda multiple times with different lambdas caused the prior ones to cancel

Ah I see.

StartFlowRun

uses an idempotency key. It uses the current task-id. From the docs:

Duplicate flow runs with the same idempotency key will only create a single flow run. This is useful for ensuring that only one run is created if this task is retried. If not provided, defaults to the active task_run_id.

So if you want multiple calls, you need to supply unique idempotency keys to each of those (like a timestamp with second resolution) as a string.

ahhh got it

Btw the new task

create_flow_run

in 0.15.0 provides a better interface than

StartFlowRun

oh? i think im on 0.14

Mainly consistent return values. If you do

StartFlowRun(wait=True)

, it does not return the

id

but if you do

StartFlowRun(wait=False)

it does. So the new paradigm is

create_flow_run

and

wait_for_flow_run

so that the return is consistent

ah

See this

ah cool

And this for the two ways to set upstream tasks in general