domains / / awslabs/agentcore-samples

Use OTEL Env Overrides

When configuring OTLP tracing exporters in observability code, rely on the runtime-provided `OTEL_EXPORTER_OTLP_ENDPOINT` and `OTEL_EXPORTER_OTLP_HEADERS` environment variables, rather than hardcoding or redundantly passing `endpoint=` and `headers=` into `OTLPSpanExporter`. This prevents configuration drift and duplicated settings.

raw .md Observability Other

When configuring OTLP tracing exporters in observability code, rely on the runtime-provided OTEL_EXPORTER_OTLP_ENDPOINT and OTEL_EXPORTER_OTLP_HEADERS environment variables, rather than hardcoding or redundantly passing endpoint= and headers= into OTLPSpanExporter. This prevents configuration drift and duplicated settings.

Example (omit redundant exporter overrides):

import os
from opentelemetry.sdk.trace import TracerProvider
from opentelemetry.sdk.resources import Resource
from opentelemetry.sdk.trace.export import BatchSpanProcessor
from opentelemetry.exporter.otlp.proto.grpc.trace_exporter import OTLPSpanExporter

# Configure only resource/service metadata in code
resource = Resource.create({
    "model_id": "agentcore-strands-agent",
})

provider = TracerProvider(resource=resource)

# Do NOT set endpoint= / headers= here if the launch environment already sets:
#   OTEL_EXPORTER_OTLP_ENDPOINT
#   OTEL_EXPORTER_OTLP_HEADERS
exporter = OTLPSpanExporter()
provider.add_span_processor(BatchSpanProcessor(exporter))

Apply this by: (1) removing in-code endpoint/headers overrides, (2) documenting which OTEL env vars must be supplied to the process/container/agent launch, and (3) ensuring CI/CD or deployment manifests provide those variables consistently.