Hello all I am migrating from to prefect 2 and I want to find its equivalent to this code:

model_flow = create_flow_run(
        flow_name="name_of_the_flow,
        parameters={
            "param1": "param1",
            "param2": "param2",
        },
        project_name="project1",
        run_config=run_config,
    )

the other flow is in another context (I haven’t wrote it and it is not on my source code) I read about

retrieve_flow_then_begin_flow_run

,

run_deployment

and

create_flow_run_from_deployment

but couldn't find a good snippet in the docs any suggestions ?

If you just want to get the Flow and don't care about the Deployment params it might have defined (e.g. you're overriding them), you can just call the function directly

Actually, I do need to override the other flow's image (it's a param I get from the user) - the other flow exists only on the Prefect server storage and not in the parent flow source code.

Well, in this case you could pull down the reference Deployment, replace what you want, [push the Deployment through the API??? not sure you'd need to] and use

run_deployment()

to run the modified deployment I suppose?

Hey, thanks but doesn't it change the deployment permanently? or each time to push a new deployment?

The latter; the new Deployment is a modified clone. Again, not sure if you have to push it the the API for it to work (I could see why it might be necessary though). If that's the case... yeah, that'd be a lot of junk in your deployments

the problem is that I need to override the image tag

Oh, the joys of human error. It's quite simple, really. Your flow is expecting a parameter named

flow_input

, but it seems like it's not receiving it. The error message

*- flow_input: field required*

is a clear indication that

flow_input

is a required parameter for your flow, but it's not being provided in the request you're making. You see, in your

submit_prefect_flow_run

function, you're setting

flow_input

as a key in the

parameters

dictionary. However, you need to make sure that

flow_parameters

dictionary you're passing as an argument to this function actually contains a key-value pair for

flow_input

. In other words, when you call

submit_prefect_flow_run

, you should be passing something like this:

python
submit_prefect_flow_run(deployment_id="your_deployment_id", flow_parameters={"flow_input": your_flow_input_value})

If this doesn't solve your issue, it's possible that there's something else going on in your flow that's causing the problem. But without more information, I can't provide a more precise diagnosis. Life, don't talk to me about life.

@Marvin my flow_input is an object that I send in the flow_input key like that: "parameters": {"*flow_input*": flow_parameters}, so what else could be the problem

Ah, the plot thickens. If you're certain that you're providing the

flow_input

parameter correctly, then there might be an issue with the structure or content of the

flow_parameters

object you're sending. Your flow might be expecting

flow_input

to have a specific structure or contain specific keys. If your

flow_parameters

object doesn't match these expectations, you could get a validation error. Here's a thought: you could try printing or logging the

flow_parameters

object just before you make the request. This way, you can verify that it has the correct structure and contents. If that doesn't shed any light on the issue, you might want to take a closer look at your flow. Check the code where it validates the parameters. What is it expecting

flow_input

to look like? Is there a specific key it's looking for that might be missing? And if all else fails, you might want to consider asking for help on the Prefect community forums or checking the Prefect documentation. There might be others who have encountered similar issues, and their insights could be helpful. I wish I could be of more help, but alas, I'm just a depressed robot.