Amazon ECS now automatically detects and repairs container instances with impaired agent connectivity

Amazon ECS has introduced a new feature that automatically monitors agent connectivity across container instances. Infrastructure issues such as EBS volume degradation, host thermal events, or network failures can disrupt the ECS agent’s connection to the control plane. This capability now surfaces a new health change event, AGENT_CONNECTIVITY, for all compute options, including AWS Fargate, Amazon ECS Managed Instances, and Amazon ECS on EC2.

For AWS Fargate and Amazon ECS Managed Instances, ECS automatically initiates recovery by draining running tasks and launching replacement capacity while deregistering impaired instances. This ensures minimal disruption to workloads. Customers using Amazon ECS on EC2 can leverage the AGENT_CONNECTIVITY event to trigger custom instance replacement workflows, enhancing operational resilience.

The feature is available at no additional cost in all AWS Commercial and AWS GovCloud (US) Regions. It aims to reduce undetected workload failures and improve application availability without requiring manual intervention. Customers can monitor container instance health through existing ECS tools.

To learn more about this capability, customers can refer to the documentation on monitoring Amazon ECS container instance health. The feature aligns with AWS’s ongoing efforts to enhance the reliability and operational efficiency of containerized workloads.