Mount health router and fix client ownership

Add an async handler that returns StatusCode::OK and attach it via a
separate health_router. Clone the Postgres client for the main app
router to avoid ownership/move issues when reusing the pool.

src/views/mod.rs

@@ -46,6 +46,10 @@ async fn auth(req: Request<axum::body::Body>, next: Next) -> Result<Response, St
     Ok(next.run(req).await)
 }
 
+async fn health_check() -> StatusCode {
+    StatusCode::OK
+}
+
 pub async fn get_router() -> Router {
     let client = get_postgres_pool().await;
 
@@ -58,15 +62,23 @@ pub async fn get_router() -> Router {
         .nest("/api/v1/books", get_books_router().await)
         .nest("/api/v1/sequences", get_sequences_router().await)
         .layer(middleware::from_fn(auth))
-        .layer(Extension(client))
+        .layer(Extension(client.clone()))
         .layer(prometheus_layer);
 
+    let health_router = Router::new()
+        .route("/health", get(health_check))
+        .layer(Extension(client));
+
     let metric_router =
         Router::new().route("/metrics", get(|| async move { metric_handle.render() }));
 
-    Router::new().merge(app_router).merge(metric_router).layer(
-        TraceLayer::new_for_http()
-            .make_span_with(trace::DefaultMakeSpan::new().level(Level::INFO))
-            .on_response(trace::DefaultOnResponse::new().level(Level::INFO)),
-    )
+    Router::new()
+        .merge(app_router)
+        .merge(health_router)
+        .merge(metric_router)
+        .layer(
+            TraceLayer::new_for_http()
+                .make_span_with(trace::DefaultMakeSpan::new().level(Level::INFO))
+                .on_response(trace::DefaultOnResponse::new().level(Level::INFO)),
+        )
 }