Amazon Elastic Container Service (ECS) enhanced its container instance management by continuously monitoring ECS agent connectivity, detecting impairments caused by infrastructure events like network failures or hardware degradation. For ECS Managed Instances and AWS Fargate, ECS automatically recovers by draining tasks and replacing instances, while ECS on EC2 surfaces these health changes for user-driven workflows.
This automated detection and repair capability aims to reduce unnoticed workload failures and improve application availability without manual intervention. It is available at no cost across all AWS Commercial and GovCloud Regions, supporting customers managing containers at scale with minimal manual effort.
Source & further reading
Amazon ECS now automatically detects and repairs container instances with impaired agent connectivity ↗Source published 2026-08-24. This report was prepared with AI assistance from the linked primary-source material. Claims and limitations should be checked against the original announcement or research.