Can timeouts be specified at the flow-level? Occasionally I’ve had issues with the dask cloudprovider process instantiating my Dask cluster on ECS, where it hangs indefinitely, preventing any tasks from actually starting. In this scenario the task timeout doesn’t help because it never actually gets to the point of running a task, it just hangs during the setup process. So it’d be helpful to be able to specify timeout at the overall flow level, if that is currently possible?

Hi @Andrew Hannigan, what Cloud plan are you on? You can do this through Automations.

@Kevin Kho We have a similar situation. I assume you are suggesting using Flow SLA Failure https://docs.prefect.io/orchestration/concepts/automations.html#events-reference. Is this limited to a specific cloud plan? If so, is that documented anywhere?

Standard+ plans get the Flow SLA Failure. Not quite documented yet I think

Starter plan at the moment @Kevin Kho

I’ll check to see what you can do and get back to you

Hey @Sean Harkins -- your best option is probably to implement your own timeout logic for dask cluster startup. I can help you get started with that if you want. Otherwise, unless you have a lot of task runs it's generally easy to fit into the free tier of Standard which gets you some of these "guarantees" from Cloud.

@Kevin Kho Just wanted to check in here on what the suggested approach is for those using Cloud?

The standard plan would allow you to use Automations to cancel the flow run after a certain amount of time. Otherwise, you would need the timeout logic for Dask as Michael suggested.