test(db): harden the DB lane per code-review (R6 S3)

/code-review high on S3 — 3 low findings, 2 fixed, 1 no-change:
- _db_reachable() now probes via a throwaway engine with connect_timeout=3, so a
  slow/firewalled DATABASE_URL can't hang the whole suite at import (the app engine
  has no connect timeout).
- test_plex_facets sets the "configured" gate flag via monkeypatch (auto-restored)
  instead of poking state._configured_cache in a manual finally — no leak risk, and
  the now-unnecessary mark_configured DB write is dropped.
No change: test_plex_jsonb_normalize copying migration 0060's SQL is deliberate —
migrations must stay self-contained (no app imports) and a shipped migration is
immutable, so there's no real drift to guard against.

Gate: ruff clean; DB lane 176 passed; pure-logic 172 passed / 4 skipped (fast).

backend/tests/conftest.py

@@ -15,19 +15,26 @@ tests that COMMIT (the whole point of this lane — transaction-ownership behavi
 import pytest
 from alembic import command
 from alembic.config import Config
-from sqlalchemy import text
+from sqlalchemy import create_engine, text
 
 import app.models  # noqa: F401 — register every model on Base.metadata
+from app.config import settings
 from app.db import Base, SessionLocal, engine
 
 
 def _db_reachable() -> bool:
+    # A throwaway engine with a bounded connect_timeout so a slow/firewalled DATABASE_URL can't
+    # hang the whole suite at import (the app engine has no connect timeout). The compose lane
+    # gates on `service_healthy`, but this keeps a manual run honest too.
+    probe = create_engine(settings.database_url, connect_args={"connect_timeout": 3})
     try:
-        with engine.connect() as conn:
+        with probe.connect() as conn:
             conn.execute(text("SELECT 1"))
         return True
     except Exception:
         return False
+    finally:
+        probe.dispose()
 
 
 _DB_REACHABLE = _db_reachable()

backend/tests/test_plex_facets.py

@@ -15,16 +15,17 @@ from app.main import app
 from app.models import PlexLibrary, PlexShow, User
 
 
-def test_facets_survives_json_null_genres(db):
+def test_facets_survives_json_null_genres(db, monkeypatch):
     lib = PlexLibrary(plex_key="1", title="Shows", kind="show", enabled=True)
     db.add(lib)
     db.flush()
     db.add(PlexShow(rating_key="s1", library_id=lib.id, title="No Genre Show"))
     user = User(email="facets@example.com")
     db.add(user)
-    # The request gate 503s an unconfigured instance; mark this throwaway DB configured.
-    state.mark_configured(db)
     db.commit()
+    # The request gate 503s an unconfigured instance; flip the process-level "configured" flag the
+    # gate reads. monkeypatch auto-restores it, so it can't leak "configured" into another test.
+    monkeypatch.setattr(state, "_configured_cache", True)
     # Simulate a pre-0.53.1 row: genres persisted as a JSONB scalar `null` (what SQLAlchemy wrote
     # before none_as_null=True), NOT SQL NULL — this is the value jsonb_array_elements_text() chokes
     # on. A Python None now maps to SQL NULL, so we set the scalar directly.
@@ -38,7 +39,6 @@ def test_facets_survives_json_null_genres(db):
         resp = client.get("/api/plex/facets?scope=show")
     finally:
         app.dependency_overrides.clear()
-        state._configured_cache = False  # one-way process cache — don't leak "configured" to other tests
 
     assert resp.status_code == 200, resp.text
     # The scalar-null show contributes no genres — and, crucially, doesn't abort the whole endpoint.