Hello, I have some questions. What is the state of async support for Kedro? I tried to create async node where the function is async since I created a custom dataset that returns an async iterator (for performance purpose :~) so I had to define async function since I have to await on the iterator. Issues are that, to test async support, I tried to create a fake node where I save the iterator values in a partitioned dataset. The error I have in this case is that the input data passed to the save function is actually a coroutine. I tried with or without the async flag, with all three runners.
Hi Yetunde, yes, I have. That is why I was confused on the state of async support because it seems you cannot define a dataset with either:
• explicit async _load function
• return an async iterator and therefore we have to define async node
I was curious about how the event loops and the async worked with kedro.