Google Cloud deepens OpenTelemetry integration to simplify metrics ingestion

For years, observability teams have chased consistency. They instrument applications one way, collect metrics another way, and then spend hours translating formats before they can even troubleshoot performance. Now, Google Cloud is trying to reduce that friction by expanding OpenTelemetry Protocol support inside Cloud Monitoring.

With this update, developers can send metrics in OTLP format directly into Cloud Monitoring, alongside traces and logs. Previously, Google Cloud supported OTLP traces, but metrics ingestion required more adaptation. Now, it is possible to create telemetry data with standard OpenTelemetry SDKs and send it directly to Cloud Monitoring or to an OpenTelemetry Collector. In this way, there is vendor neutrality in the instrumentation, while at the same time ingesting the data natively within Google’s stack.

This is important because many organizations are now moving in this direction and want to have more portability. They do not want to lock in to proprietary agents and want to use open standards that give them flexibility across platforms. Therefore, this move by Google Cloud aligns with the direction that the industry is moving in. At the same time, the company maintains compatibility with existing Monitoring tools, since it stores OTLP metrics similarly to Prometheus formatted data.

The update also introduces practical improvements. Delta metrics reduce memory pressure by reporting incremental changes rather than full counters. Exponential histograms support dynamic bucket sizing, which helps teams capture performance patterns more precisely. In addition, expanded naming conventions allow dots and slashes, which better reflect common OpenTelemetry semantic structures.

Meanwhile, Google Cloud launched Managed OpenTelemetry support for Google Kubernetes Engine. This managed pipeline deploys and scales collectors automatically, so platform teams do not need to maintain that infrastructure themselves. As a result, Kubernetes workloads can use standardized telemetry with reduced operational burdens.

Across the cloud landscape, cloud providers are increasingly competing on integration rather than features.  Other observability platforms offer support for the OTLP protocol in various capacities, but Google Cloud has the protocol built directly into the underlying monitoring stack.

The support for OTLP metrics remains in preview and requires OpenTelemetry version 0.140.0 or higher. The trend, however, is obvious. Standardized telemetry is becoming more and more a non-negotiable aspect of cloud operations.