Based on the GitHub issues, I should also mention that if you're running flows with Dask on Kubernetes, you might want to ensure proper cleanup of resources when canceling flow runs, as there's been some discussion about orphaned pods in those scenarios.
The most reliable way to see execution details is through a combination of the UI and CLI tools I mentioned earlier. If you need more specific details about a particular aspect of the execution, let me know and I can provide more targeted information.