Refactor persistence config to use stores key with unified backends

- Add StoresConfig to group all store references under persistence.stores
- Use single 'default' backend instead of separate metadata_backend/inference_backend
- Update resolver to access persistence.stores.{metadata,inference,conversations}
- All SQLite distributions now use single store.db file with shared backend

llama_stack/distributions/nvidia/run.yaml

@@ -13,7 +13,7 @@ apis:
 - tool_runtime
 - vector_io
 providers:
-  inference:
+    inference:
   - provider_id: nvidia
     provider_type: remote::nvidia
     config:
@@ -88,16 +88,14 @@ providers:
         db_path: ${env.SQLITE_STORE_DIR:=~/.llama/distributions/nvidia}/files_metadata.db
 persistence:
   backends:
-    metadata_backend:
+    default:
       type: sqlite
       db_path: ${env.SQLITE_STORE_DIR:=~/.llama/distributions/nvidia}/registry.db
-    inference_backend:
-      type: sqlite
-      db_path: ${env.SQLITE_STORE_DIR:=~/.llama/distributions/nvidia}/inference_store.db
-  metadata:
-    backend: metadata_backend
-  inference:
-    backend: inference_backend
+  stores:
+    metadata:
+      backend: default
+    inference:
+      backend: default
 models: []
 shields: []
 vector_dbs: []