OpenTelemetry Runtime Metrics
このページは日本語には対応しておりません。随時翻訳に取り組んでいます。翻訳に関してご質問やご意見ございましたら、お気軽にご連絡ください。
Overview
Runtime metrics are application metrics about memory usage, garbage collection, or parallelization. Datadog tracing libraries provide runtime metrics collection for each supported language, but in addition, OpenTelemetry (OTel) collects runtime metrics, which can be sent to Datadog through the OpenTelemetry SDKs.
Datadog collects OpenTelemetry runtime metrics in the following languages:
Metric naming conventions
Runtime metrics follow different naming conventions depending on their source: OpenTelemetry Collector Datadog Exporter, Datadog Agent OTLP Ingestion, or Datadog tracing libraries. When using OpenTelemetry runtime metrics with Datadog, you receive both the original OpenTelemetry runtime metrics as well as mapped Datadog runtime metrics for equivalent metrics. Runtime metrics have the following prefixes which indicate their source:
| OTel Collector Datadog Exporter | Datadog Agent OTLP Ingest | Datadog tracing library |
|---|
otel.process.runtime.* | process.runtime.* | runtime.<LANG>.* |
Note: OpenTelemetry runtime metrics are mapped to Datadog by metric name. Don’t do mapping renaming of host metrics for OpenTelemetry runtime metrics or it will break.
For details about host and container metrics mapping, read OpenTelemetry Metrics Mapping.
Setup
Select your language to see instructions for setting up and configuring the OpenTelemetry SDK to send runtime metrics:
View runtime metric dashboards
After setup is complete, see your runtime metrics in the a service’s details page (see Java example below), the flame graph metrics tab, and in default runtime dashboards.
