@Marvin Prefect loses control over Fargate tasks after 24 hours. How to correct it?

It seems the issue of Prefect losing control over Fargate tasks after 24 hours can be related to lifecycle management and the configurations set on Fargate. Here are a few steps you can take to correct this issue: 1. Check

requiresCompatibilities

and `networkMode`: Ensure that your task definition for Fargate specifies

requiresCompatibilities

as

FARGATE

and uses the

awsvpc

network mode. For more information, refer here. 2. Pool limitations: If you are using managed execution, ensure you are within the limitations of your Prefect account tier. The limits for free and pro tiers can impact the control over tasks. Details available here. 3. Scalability and Task Queue Management: Ensure that your architecture can effectively manage and monitor tasks over an extended period. This involves robust queue management, observability, centralized logging, and error handling as suggested here. For more specific troubleshooting or if the issue persists, you may want to search or raise an issue on GitHub under the Prefect repository.