Hi all Running into a bit of an odd issue around pickling of

Question

Hi all Running into a bit of an odd issue around pickling of a result with a recent change to an existing flow The return type from the task is a List lt dataclass gt Within the object the properties are both strings I ve tried re deploying the flow and currently in progress of upgrading from 2 2 0 to 2 3 2 Anyone experience issues around cloud pickle on returning an object from a task Stacktrace linked in comments

Cole Murray · Answer

```Can t pickle local object create fn lt locals gt init Traceback most recent call last File opt venv lib python3 8 site packages workflow etl flows flow py line 265 in hello world device paths = get object paths lt redacted gt File opt venv lib python3 8 site packages prefect tasks py line 294 in call return enter task run engine File opt venv lib python3 8 site packages prefect engine py line 727 in enter task run engine return run async from worker thread begin run File opt venv lib python3 8 site packages prefect utilities asyncutils py line 136 in run async from worker thread return anyio from thread run call File opt venv lib python3 8 site packages anyio from thread py line 49 in run return asynclib run async from thread func args File opt venv lib python3 8 site packages anyio backends asyncio py line 970 in run async from thread return f result File usr local lib python3 8 concurrent futures base py line 444 in result return self get result File usr local lib python3 8 concurrent futures base py line 389 in get result raise self exception File opt venv lib python3 8 site packages prefect engine py line 859 in create task run then submit return await future result File opt venv lib python3 8 site packages prefect futures py line 227 in result return final state result raise on failure=raise on failure File opt venv lib python3 8 site packages prefect orion schemas states py line 145 in result raise data File opt venv lib python3 8 site packages prefect task runners py line 214 in submit result = await run fn run kwargs File opt venv lib python3 8 site packages prefect engine py line 1000 in begin task run return await orchestrate task run File opt venv lib python3 8 site packages prefect engine py line 1120 in orchestrate task run terminal state = await return value to state File opt venv lib python3 8 site packages prefect states py line 129 in return value to state return Completed data=DataDocument encode serializer result File opt venv lib python3 8 site packages prefect orion schemas data py line 42 in encode blob = lookup serializer encoding dumps data kwargs File opt venv lib python3 8 site packages prefect serializers py line 59 in dumps data bytes = cloudpickle dumps data File opt venv lib python3 8 site packages cloudpickle cloudpickle fast py line 73 in dumps cp dump obj File opt venv lib python3 8 site packages cloudpickle cloudpickle fast py line 632 in dump return Pickler dump self obj AttributeError Can t pickle local object create fn lt locals gt init ```

Cole Murray · Answer

``` task def get object paths lt redacted gt gt List DeviceDataFilePath return get object paths def get object paths lt redacted gt gt List DeviceDataFilePath redacted time now = datetime datetime utcnow logger = get run logger device data = for device id in device ids for day in range num days day time = time now datetime timedelta days=day year = day time year month = day time month day = day time day filter pattern = f lt redacted gt <http logger info|logger info> f Filter pattern filter pattern filtered objs = aws utils list bucket bucket=input bucket prefix=filter pattern objs = list filtered objs objs = sorted objs key=lambda obj obj last modified reverse=True lt redacted gt return device data```

Cole Murray · Answer

Calling flow ``` flow def hello world lt redacted gt try bucket = input bucket device paths = get object paths lt redacted gt ```

Anna Geller · Answer

it looks like your dataclass object is not pickleable with cloudpickle to troubleshoot you could try to cloudpickle it by hand not sure how helpful this info is but your stack trace seems to refer to this function in the cpython codebase <https github com python cpython blob main Lib dataclasses py L413> having said that we are working on more flexible result config functionality which will give you more control over what results will be persisted and how

Cole Murray · Answer

Yep not totally clear why the object isn t pickleable I ended up swapping the dataclass model into a pydantic model and resolved the issue