Initial Aurora Core 0.1 implementation

Production-grade media metadata backend per docs/aurora-core-spec.md: Aurora-owned UUIDv7 identity system, normalized movie/TV/people/credits schema, provider abstraction with a TMDb adapter, an idempotent async ingestion pipeline (normalize -> match -> upsert -> provenance), a Postgres FTS/trigram search layer, a stateless FastAPI v1 API with Redis caching/rate-limiting, arq workers, dev/prod Docker Compose topologies, unit + integration + load tests, CI, and architecture/ database/API/ERD documentation. Co-Authored-By: Claude Sonnet 5 <[email protected]>
2026-09-04 17:34:03 -05:00
commit c20df1c503
111 changed files with 10041 additions and 0 deletions
@@ -0,0 +1,103 @@
+"""Load test for Aurora Core 0.1 (section 30/2).
+
+Simulates the read-heavy interactive traffic Core 0.1 targets: searching,
+opening movie/series/person detail pages, and paging through search
+results. Run against a seeded environment (ingest a reasonable movie/series/
+person catalog first — an empty database will pass load tests trivially but
+tell you nothing about real query cost).
+
+Usage:
+
+    pip install locust
+    locust -f tests/load/locustfile.py --host http://localhost:8000
+
+Then open http://localhost:8089 and start a run with up to 500 users to
+verify the section 2 targets (search p95 < 500ms, other endpoints p95 <
+250ms, 5xx rate < 0.1%). Locust reports p50/p95/p99 latency and failure
+rate directly; compare them against docs/architecture.md's targets.
+
+See docs/architecture.md for how to seed sample data before running this.
+"""
+
+from __future__ import annotations
+
+import random
+
+from locust import HttpUser, between, task
+
+SEARCH_TERMS = [
+    "matrix",
+    "star wars",
+    "alien",
+    "the office",
+    "breaking bad",
+    "keanu reeves",
+    "christopher nolan",
+    "marvel",
+    "batman",
+    "inception",
+]
+
+
+class AuroraCoreUser(HttpUser):
+    """A typical interactive client: search, then drill into a result."""
+
+    wait_time = between(1, 3)
+
+    def on_start(self) -> None:
+        self._known_movie_ids: list[str] = []
+        self._known_series_ids: list[str] = []
+        self._known_person_ids: list[str] = []
+
+    @task(5)
+    def search(self) -> None:
+        query = random.choice(SEARCH_TERMS)
+        limit = random.choice([5, 10, 20])
+        with self.client.get(
+            f"/api/v1/search?q={query}&limit={limit}",
+            name="/api/v1/search",
+            catch_response=True,
+        ) as response:
+            if response.status_code != 200:
+                response.failure(f"search failed: {response.status_code}")
+                return
+            body = response.json()
+            self._known_movie_ids.extend(m["id"] for m in body.get("movies", []))
+            self._known_series_ids.extend(s["id"] for s in body.get("series", []))
+            self._known_person_ids.extend(p["id"] for p in body.get("people", []))
+            # Keep the pools bounded so memory doesn't grow unbounded over a
+            # long run.
+            self._known_movie_ids = self._known_movie_ids[-200:]
+            self._known_series_ids = self._known_series_ids[-200:]
+            self._known_person_ids = self._known_person_ids[-200:]
+
+    @task(4)
+    def open_movie(self) -> None:
+        if not self._known_movie_ids:
+            return
+        movie_id = random.choice(self._known_movie_ids)
+        self.client.get(f"/api/v1/movies/{movie_id}", name="/api/v1/movies/[id]")
+        self.client.get(
+            f"/api/v1/movies/{movie_id}/credits", name="/api/v1/movies/[id]/credits"
+        )
+
+    @task(2)
+    def open_series(self) -> None:
+        if not self._known_series_ids:
+            return
+        series_id = random.choice(self._known_series_ids)
+        self.client.get(f"/api/v1/series/{series_id}", name="/api/v1/series/[id]")
+        self.client.get(
+            f"/api/v1/series/{series_id}/seasons", name="/api/v1/series/[id]/seasons"
+        )
+
+    @task(2)
+    def open_person(self) -> None:
+        if not self._known_person_ids:
+            return
+        person_id = random.choice(self._known_person_ids)
+        self.client.get(f"/api/v1/people/{person_id}", name="/api/v1/people/[id]")
+
+    @task(1)
+    def health_probe(self) -> None:
+        self.client.get("/health", name="/health")