Add production condition retrieval smoke test

2026-08-11 14:58:28 +07:00
parent 59e6ad2d0d
commit 7ebbe1f309
38 changed files with 3752 additions and 121 deletions
@@ -51,6 +51,13 @@ class Settings(BaseSettings):
    # never uses it.
    rerank_enabled: bool = False
    metrics_enabled: bool = True
+    # Optional bearer token for `GET /metrics`. Empty by default so the
+    # current Compose scrape and local development keep working: the endpoint
+    # is not reachable from the internet today (Caddy proxies only `web`, and
+    # ai-service publishes no host port). Set it wherever the service is
+    # exposed through an Ingress, which the Helm chart now allows — metrics
+    # carry query volumes, provider failure counts and abstain reasons.
+    metrics_token: str = ""
    # OpenTelemetry is opt-in so the existing EC2 Compose deployment keeps
    # answering when no collector is present. Docker/Kubernetes observability
    # profiles enable it and point OTLP/HTTP at their local collector Service.