Successfully Ran Getting Started on Windows

[Steve-O-Murphy]

Hey Flyte team,

Just FYI, I successfully ran the '1. Implement' step on Windows (10). Wanted to let you know because you had asked for feedback.

[EngHabu]

This is awesome! thank you for letting us know. Do you have any questions or feedback about it?

[cgrisham]

I have not been able to run the Flyte getting started on windows. When I try to execute the workflow in the demo cluster I get this error

"[1/1] currentAttempt done. Last Error: USER::Pod failed. No message received from kubernetes.
[feb1d6fc03e6c4a87aa0-n0-0] terminated with exit code (137). Reason [OOMKilled]. Message:
tar: Removing leading `/' from member names"

[kumare3]

@cgrisham this is because the emulated version of docker are consuming more memory for some reason in the getting started. We are just releasing a fix, but until then if you can update the getting started code as follows

@task(limits=Resources(mem="500Mi"))
def generate_normal_df(n: int, mean: float, sigma: float) -> pd.DataFrame:
    return pd.DataFrame({"numbers": np.random.normal(mean, sigma, size=n)})


@task(limits=Resources(mem="500Mi"))
def compute_stats(df: pd.DataFrame) -> typing.Tuple[float, float]:
    return float(df["numbers"].mean()), float(df["numbers"].std())

As done on this issue that would be great - #2376

[katrogan]

hi @cgrisham alternatively, our latest Flyte beta release should also work if you don't want to muck with task resources

flytectl demo start --pre

and this should have increased resources for all tasks applied by default.

[cgrisham]

@kumare3 @katrogan Thank you both so much for your responsiveness. Both of these methods work to resolve the issue !