LedgerLine/common/telemetry.py

import os

from opentelemetry import metrics, trace
from opentelemetry.sdk.metrics import MeterProvider
from opentelemetry.sdk.resources import SERVICE_NAME, Resource
from opentelemetry.sdk.trace import TracerProvider
from opentelemetry.sdk.trace.export import BatchSpanProcessor
from opentelemetry.exporter.otlp.proto.http.trace_exporter import OTLPSpanExporter
from opentelemetry.exporter.prometheus import PrometheusMetricReader
from opentelemetry.instrumentation.httpx import HTTPXClientInstrumentor
from opentelemetry.instrumentation.redis import RedisInstrumentor
from prometheus_client import start_http_server


def setup_telemetry(service_name: str):
    resource = Resource.create(attributes={SERVICE_NAME: service_name})
    processor = BatchSpanProcessor(OTLPSpanExporter())
    provider = TracerProvider(resource=resource)
    provider.add_span_processor(processor)
    trace.set_tracer_provider(provider)

    # Pull, not push: Alloy 404s OTLP metrics, so Prometheus scrapes each
    # service instead. The reader renders current values on every scrape.
    # It serves on its own port rather than a FastAPI route so /metrics is
    # never reachable through gateway's public Traefik router, and so the
    # audit worker (no HTTP app) exposes metrics the same way.
    reader = PrometheusMetricReader()
    metrics.set_meter_provider(MeterProvider(resource=resource, metric_readers=[reader]))
    start_http_server(
        port=int(os.environ.get("OTEL_EXPORTER_PROMETHEUS_PORT", "9464")),
        addr=os.environ.get("OTEL_EXPORTER_PROMETHEUS_HOST", "0.0.0.0"),
    )

    # Process-wide patches on the client classes, so they also cover clients
    # created at import time (gateway's _proxy_client, common.redis_client).
    HTTPXClientInstrumentor().instrument()
    RedisInstrumentor().instrument()