@Marvin I'm getting this issue running flows on AWS ECS: RuntimeError: Failed to run ECS task, cluster 'default' not found. Confirm that the cluster is configured in your region.

Ah, the joys of human error. It seems you're not the only one who's had this issue. There are a couple of things that could be causing this. Firstly, it could be related to a bug that was reported on Prefect's GitHub page. The issue was that the agent was occasionally unable to retrieve jobs to run from Prefect Cloud. This was particularly prevalent after upgrading from version 2.4.0 to 2.6.5. You can find more details about this issue here. Secondly, there was a pull request that added the ability to use capacity providers in ECS Agent and ECSRun when set in

run_task_kwargs

. This allows for the use of Fargate Spot and ASG's. You can find more details about this here. In your case, the error message suggests that the 'default' cluster is not found. You might want to check that the cluster is correctly configured in your region. If you're still having trouble, you might want to consider reaching out to the Prefect community for further assistance. I hope this helps, or at least makes your day slightly less dreadful.

Thanks! Now I'm getting "Flow could not be retrieved from deployment" instead @Marvin

Ah, the endless cycle of issues. It seems like you're not alone in experiencing this problem. There have been several reports of similar issues on Prefect's GitHub page. One such issue is titled "Retry Cloning Of Remote Repository Upon Failure". It describes a situation where a scheduled flow run fails to clone from GitHub (or other repository) and returns the error: "Flow could not be retrieved from deployment." Another issue titled "Failed to pull from remote (RPC failed; HTTP 500 curl...)" also seems to be related to your problem. It describes an error where the flow could not be retrieved from deployment due to a failure to pull from the remote repository. There's also an issue titled "agent fails to get flow from storage-block with path set as relative path" that might be relevant. It describes a situation where the flow run fails in the agent due to a FileNotFoundError. Finally, there's an issue titled "`get_directory` failure log exposes GitHub Private Access Token" that describes a security concern where the GitHub access token is exposed in the error log. I would recommend checking out these issues for potential solutions or workarounds. If you're still having trouble, you might want to consider reaching out to the Prefect community for further assistance. I hope this helps, or at least makes your day slightly less dreadful.