Bring your towel and join one of the fastest growing data communities. Welcome to our second-generation open source orchestration platform, a completely rethought approach to dataflow automation.

Prefect Community

Annotation 2023-09-04 112221.png

Hi everyone, I'm having a weird issue when I trigger my prefect deployment through the prefect api.
I have the flow registered and deployed already, but If I go the "create flow run" or "create flow run from deployment" passed in necessary request body parameters, and I hit execute, the deployment will get triggered but it won't run the tasks in the flow.

This is the request body I passed to it under the "create_flow_run" :
{
  "state": {
    "type": "RUNNING"
  },
  "name": "weather_flow_run",
  "flow_id": "xxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}

Then, that's the flow run status below: it doesn't run the tasks in my flow and it keeps running forever
<@U03KWSKFN8L> <@U05340YCP0C>

hey Peter, I had the same issue and resolved it with using "SCHEDULED"

```{
  "state": {
    "type": "SCHEDULED",
    "name": "deployment-name",
    "message": "Run started",
  },
  "parameters": {
    "some_parametaer": "bla_bla" 
  },
  "tags": [
  ],
} ```

and the flow id is in the url:
```api/deployments/{FLOW_ID}/create_flow_run```

sorry do you mean deployment_id instead of flow_id, the parameter it asks for is the deployment_id

Annotation 2023-09-04 161138.png

This is what it's resulting in: it's not running the flow immediately

image.png

maybe it has something to do with work pool or work queue?
because when I run with "scheduled" I immediately see them in the UI

sorry, how did you configure your work pool and work queue to use default?, was it in the deployment script

hi <@U05RJ6HAXS4> - are you running <https://docs.prefect.io/latest/concepts/work-pools/?h=worker#worker-overview|a worker>?

i see that your flow run is scheduled but marked late, it could be that whichever work pool you have deployed your flow to does not have a worker that would pick up and execute your flow run

&gt; configure your work pool and work queue
you select a work pool (optionally a work queue as well) at deployment creation time, which is basically selecting the type of infra (Process, Docker, k8s) you want your flow runs to execute on - the default infra is Process, which would be created by a process worker if you ran `prefect worker start -p my_new_process_pool`

Alright, Noted. It works fine now, Thanks very much