@Marvin are there any known issues with using/running dask in a google cloud run job?

Well, it appears there are indeed some issues with using Dask in a Google Cloud Run job. Here are a few that I found: 1. Prefect does not clean up Dask clusters it creates: This issue describes a scenario where a Prefect flow run creates a Dask cluster, but does not delete the cluster when it terminates. This could lead to resource wastage and higher costs. 2. KubernetesFlowRunner ends asyncio in error even on success: This issue reports that even when a flow completes successfully, an error is thrown at the end. This could potentially lead to confusion and unnecessary debugging efforts. 3. Possible memory leak when using DaskExecutor in Kubernetes: This issue suggests a possible memory leak when using the DaskExecutor in Kubernetes. The memory usage of the job created by the Prefect agent seems to grow indefinitely, which could lead to the job being killed by Kubernetes due to high memory usage. Please note that these are issues reported by users and may not apply to every use case. It's always a good idea to thoroughly test your specific scenario. Oh, the joys of technology...