Skip to Content
GlossaryOpenTelemetry - Glossary

OpenTelemetry

Back to GlossaryDevelopment Tools

An open observability standard for collecting and exporting traces, metrics, and logs, used by the Rhesis SDK to integrate with external monitoring tools.

Also known as: OTel, Open Telemetry

Overview

OpenTelemetry (OTel) is a vendor-neutral, open-source observability framework maintained by the Cloud Native Computing Foundation (CNCF). It provides a standard way to instrument applications, collect telemetry data, and export it to any compatible backend.

Rhesis uses OpenTelemetry internally for its own tracing and also exposes OTel-compatible instrumentation through the SDK, allowing you to send traces to your own observability stack.

Core Concepts

Traces: A trace represents a single end-to-end operation (such as a test execution). It is composed of spans.

Spans: Individual units of work within a trace (function calls, LLM invocations, tool calls). Each span has a name, start time, duration, status, and optional attributes.

Exporters: Components that send telemetry data to a backend. OTel supports exporters for Jaeger, Grafana Tempo, Datadog, Honeycomb, and many others.

Using OpenTelemetry with Rhesis

The Rhesis SDK emits OTel-compatible spans when you use the and decorators. You can configure an exporter to send these spans to your preferred backend:

python
from opentelemetry.sdk.trace import TracerProvider
from opentelemetry.exporter.otlp.proto.grpc.trace_exporter import OTLPSpanExporter
from opentelemetry.sdk.trace.export import BatchSpanProcessor

provider = TracerProvider()
processor = BatchSpanProcessor(
      OTLPSpanExporter(endpoint="http://your-collector:4317")
)
provider.add_span_processor(processor)

MCP and OpenTelemetry

The Rhesis SDK also supports OpenTelemetry tracing for MCP tool calls, enabling end-to-end observability across both direct function calls and MCP-mediated tool invocations.

Best Practices

  • Start with the default Rhesis tracing before adding an external OTel exporter to avoid duplicate telemetry
  • Use sampling in high-volume production environments to control telemetry storage costs
  • Align OTel attribute names with your organization's observability conventions for consistent dashboards
  • Verify that your OTel exporter is configured before running test suites to ensure traces are captured end-to-end

Related Terms

TermsAcknowledgments