temporalio.contrib.opentelemetry.

module documentation

(source)

Provider-neutral OpenTelemetry helpers shared by serverless integrations.

This is a private module. It is intentionally not re-exported from temporalio.contrib.opentelemetry so that importing that package gains no new imports. The OTLP gRPC span exporter is imported lazily inside build_otlp_span_processor, so importing this module does not require the opentelemetry-exporter-otlp-proto-grpc dependency.

The AWS Lambda and GCP Cloud Run integrations layer their provider-specific policy (ID generators, default service names, environment fallbacks) on top of these helpers. Empty environment/argument values are skipped with a plain truthiness check and are not stripped of whitespace, matching the pre-existing AWS Lambda resolution behavior.

Function	`build_metrics_telemetry_config`	Build Core telemetry configuration for OTLP metrics export.
Function	`build_otlp_span_processor`	Build a batch span processor backed by the OTLP gRPC exporter.
Function	`resolve_endpoint`	Resolve the OTLP collector endpoint.
Function	`resolve_service_name`	Resolve the OpenTelemetry service name.

def build_metrics_telemetry_config(*, endpoint: str, service_name: str, metric_periodicity: timedelta | None) -> TelemetryConfig: (source) ¶

Build Core telemetry configuration for OTLP metrics export.

Parameters
endpoint:`str`	OTLP collector endpoint. Falls back to `DEFAULT_OTLP_ENDPOINT` when empty.
service_name:`str`	Service name added as the `service_name` global tag. When empty, no global tag is added.
metric_periodicity:`timedelta \| None`	Metric export interval, passed through unchanged.

Returns
`TelemetryConfig`	A `temporalio.runtime.TelemetryConfig` with metrics pointed at the collector.

def build_otlp_span_processor(endpoint: str, *, insecure: bool = True) -> BatchSpanProcessor: (source) ¶

Build a batch span processor backed by the OTLP gRPC exporter.

The exporter is imported lazily so that importing this module does not require opentelemetry-exporter-otlp-proto-grpc.

Parameters
endpoint:`str`	OTLP collector endpoint.
insecure:`bool`	Whether to use an insecure (non-TLS) gRPC channel.

Returns
`BatchSpanProcessor`	A batch span processor that exports to the OTLP collector.

Raises
`ImportError`	If the OTLP gRPC exporter is not installed. The caller decides whether to warn and continue or re-raise.

def resolve_endpoint(explicit: str | None, *, env: Mapping[str, str] = os.environ, default: str = DEFAULT_OTLP_ENDPOINT) -> str: (source) ¶

Resolve the OTLP collector endpoint.

Resolution order: explicit -> OTEL_EXPORTER_OTLP_ENDPOINT -> default. Empty values are skipped with a truthiness check, without stripping whitespace.

Parameters
explicit:`str \| None`	Endpoint supplied directly by the caller, if any.
env:`Mapping[str, str]`	Environment mapping. Defaults to the live process environment.
default:`str`	Endpoint used when nothing else is provided.

Returns
`str`	The resolved endpoint.

def resolve_service_name(explicit: str | None, fallback_env_vars: Sequence[str], default: str, *, env: Mapping[str, str] = os.environ) -> str: (source) ¶

Resolve the OpenTelemetry service name.

Resolution order: explicit -> OTEL_SERVICE_NAME -> each name in fallback_env_vars in order -> default. Empty values are skipped with a truthiness check, without stripping whitespace.

Parameters
explicit:`str \| None`	Service name supplied directly by the caller, if any.
fallback_env_vars:`Sequence[str]`	Provider-specific environment variable names checked, in order, after `OTEL_SERVICE_NAME`.
default:`str`	Service name used when nothing else is provided.
env:`Mapping[str, str]`	Environment mapping. Defaults to the live process environment.

Returns
`str`	The resolved service name.

temporalio.contrib.opentelemetry._serverless

`temporalio.contrib.opentelemetry._serverless`