Amazon CloudWatch now supports resource tags for monitoring vended metrics, allowing DevOps engineers to create dynamic monitoring views aligned with their organizational structure. This tag-based telemetry experience simplifies the management of alarms and metrics, enabling faster insights and reducing manual overhead after deployments. The feature is available in multiple AWS regions and can be enabled easily through the CloudWatch Settings or AWS CLI.

The Amazon Product Search team shares their journey of transitioning from traditional threshold-based monitoring to Service Level Objectives (SLO) monitoring using CloudWatch Application Signals. Part 1 focuses on the limitations of conventional monitoring methods and the benefits of SLOs in detecting significant issues while reducing false alarms, leading to improved system observability and reliability.

The article discusses how to monitor agentic AI applications using Amazon CloudWatch, highlighting the importance of observability for ensuring reliability and performance. It details the setup of a sample Weather Forecaster application built with Strands Agents SDK, which utilizes CloudWatch to collect telemetry data, including metrics, traces, and logs, for comprehensive analysis. Additionally, it provides a step-by-step guide for deploying the application and analyzing the generated telemetry data in the CloudWatch console.