Amazon CloudWatch introduces OpenTelemetry metrics support in public preview

Amazon CloudWatch has announced the public preview of its support for native OpenTelemetry (OTel) metrics, allowing users to send metrics directly using the OpenTelemetry Protocol (OTLP) without the need for custom conversion processes or extra tools. This integration enables users to seamlessly combine their custom OpenTelemetry metrics with AWS-provided metrics from more than 70 services, and query them using PromQL, all without the necessity of additional agents or code adjustments.

With this native support, teams operating microservices on Amazon EKS and on-premises servers can now directly transmit OTel metrics from both environments to CloudWatch. This capability allows them to correlate application-level metrics, such as order processing latency from their on-premises services, with metrics like EKS pod CPU utilization and Application Load Balancer request counts. They can then utilize PromQL to create comprehensive dashboards and alarms that cover their entire infrastructure.

CloudWatch’s anomaly detection feature is compatible with OTel metrics, automatically identifying unusual patterns without the need for predefined static thresholds. Additionally, a new console experience, Query Studio, has been introduced for PromQL. This feature allows users to write queries, explore metrics, create alarms, and construct dashboards directly within the CloudWatch console.

The native OpenTelemetry metrics support is currently available in public preview in several regions, including US East (N. Virginia), US West (Oregon), Asia Pacific (Sydney), Asia Pacific (Singapore), and Europe (Ireland). During the preview period, there is no cost associated with using OpenTelemetry metrics or querying them. For more information, users can refer to the Amazon CloudWatch OpenTelemetry documentation.