Route Dagger build telemetry to Tempo

The Dagger engine's internal OTLP proxy returns 500 on /v1/metrics when
there's no real backend, causing ~9s retry warnings per pipeline step.
Point OTEL_EXPORTER_OTLP_ENDPOINT at Tempo to give it a real endpoint.
Also removes the stale os.environ workaround from main.py (the SDK
initializes telemetry before our module loads, so it had no effect).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

.forgejo/workflows/build-container.yaml

@@ -73,6 +73,11 @@ jobs:
     needs: detect
     if: needs.detect.outputs.dagger != '[]'
     runs-on: k8s
+    env:
+      # Send Dagger OTLP telemetry to Tempo. Without a real backend the
+      # engine's internal proxy returns 500 on /v1/metrics, causing noisy
+      # retry warnings in every build.
+      OTEL_EXPORTER_OTLP_ENDPOINT: http://tempo.tracing.svc.cluster.local:4318
     strategy:
       matrix:
         container: ${{ fromJson(needs.detect.outputs.dagger) }}

docs/changelog.d/+dagger-otel-metrics-fix.bugfix.md

@@ -1 +1 @@
-Fix OTEL metrics exporter warnings in Dagger builds by hard-overriding the env var before SDK init.
+Route Dagger build telemetry to Tempo, fixing OTEL metrics exporter warnings.

src/blumeops/main.py

@@ -1,13 +1,5 @@
-import os
 from pathlib import Path
 
-# Force-disable OTLP metrics exporter before the Dagger SDK initializes
-# OpenTelemetry. The engine shim may set OTEL_METRICS_EXPORTER=otlp before
-# our module loads, so setdefault won't work — we need a hard override.
-# Without this, the engine's local OTLP endpoint returns 500s on metrics,
-# causing ~9s retry cycles per pipeline step.
-os.environ["OTEL_METRICS_EXPORTER"] = "none"
-
 import dagger
 from dagger import dag, function, object_type