<@ULVA73B9P> i have a flow scheduled to run at 4am...
# ask-marvin
n
@Marvin i have a flow scheduled to run at 4am every day. today it failed to schedule or run it. What could cause this
m
The search didn't return relevant information about Prefect flow scheduling issues. I'll try searching the GitHub issues for potential similar problems users might have faced. Please hold on for a moment. The search didn’t bring any specific solutions to your issue, however, I did find some related issues on GitHub which may or may not be directly related to your problem. It seems there can be numerous reasons why a scheduled flow may fail to start, such as issues with the Kubernetes cluster being unable to schedule a pod in time, or late runs needing to be cleared after an agent goes down. Here are the relevant issues: 1. Handle failed Kubernetes scheduling events more gracefully: When a Kubernetes cluster is unable to schedule a pod in time, flow runs might be reported as crashed but may subsequently complete successfully. For details, see Issue #12071. 2. Multi-select flow runs and set state in bulk: There's a feature request for the ability to clear late runs, cancel, or delete multiple scheduled runs in bulk, which could potentially affect the normal scheduling of flows if late runs aren't managed properly. For more info, check Issue #7006. 3. Import of module '__prefect_loader__' failed when 'multiprocessing.Pool.apply_async' is used: Although not directly linked to scheduling, this issue could potentially affect the execution of your flow if you’re using multiprocessing in your code and deploying through an agent. Here's the Issue #9329. For your specific issue, you might consider checking the following things: - Verify if your Prefect agent was running at the scheduled time. - Check the logs of the Prefect agent and server/UI for any error messages that occurred around that time. - Review your flow and its dependencies to ensure they're all correct and can be found by the agent. - Ensure your system clock is accurate and in sync. In case you’ve verified the above points and there is still an issue, I would recommend reaching out to Prefect Support or opening a new issue in the Prefect GitHub repository with specific details of your problem.