Amazon EKS and ECS have introduced fully managed Model Context Protocol (MCP) servers in preview. These servers enhance AI applications with real-time insights about clusters, simplifying development and operations by eliminating local installation and maintenance. Developers can configure AI coding assistants, while operators gain access to a rich knowledge base for best practices and troubleshooting.

Amazon ECS now allows tmpfs mounts for Linux tasks on AWS Fargate and ECS Managed Instances. This enables the creation of in-memory file systems for tasks, which is useful for temporary data and sensitive information that shouldn’t persist after the task ends. Users can configure tmpfs mounts through task definitions in the ECS console or via AWS CLI and other tools.

AWS ECS tasks running on EC2 instances face weak task-level isolation, leading to potential security risks like credential theft. The article highlights the importance of hardening configurations, particularly by restricting access to the EC2 Instance Metadata Service (IMDS), and discusses various networking modes and methods to effectively block IMDS access for ECS tasks.

Migrating from AWS CodeDeploy to Amazon ECS for blue/green deployments offers organizations enhanced capabilities such as service discovery options, headless service support, and improved operational features. The article outlines key considerations for migration, including differences in API and console functionalities, and details the implementation process for ECS blue/green deployments compared to CodeDeploy. Organizations are encouraged to assess their deployment strategies and configurations to leverage the benefits of ECS blue/green deployments effectively.