fix(embedder): report configured provider in slow-call logs (#1403)

qin-ptr · qin-ctx · web-flow · commit 044e25916335 · 2026-04-13T11:44:52.000+08:00
* fix(embedder): report configured provider in slow-call logs

Preserve the configured embedding provider in slow-call warnings so OpenAI-compatible backends like Ollama do not show up as unknown or as their transport mode.

* fix log

---------

Co-authored-by: qin-ctx &lt;qinhaojie.exe@bytedance.com&gt;
diff --git a/openviking/models/embedder/cohere_embedders.py b/openviking/models/embedder/cohere_embedders.py
@@ -53,6 +53,7 @@ def __init__(
         config: Optional[Dict[str, Any]] = None,
     ):
         super().__init__(model_name, config)
+        self.provider = "cohere"
 
         self.api_key = api_key
         self.api_base = (api_base or "https://api.cohere.com").rstrip("/")
diff --git a/openviking/models/embedder/gemini_embedders.py b/openviking/models/embedder/gemini_embedders.py
@@ -131,6 +131,7 @@ def __init__(
         config: Optional[Dict[str, Any]] = None,
     ):
         super().__init__(model_name, config)
+        self.provider = "gemini"
         if not api_key:
             raise ValueError("Gemini provider requires api_key")
         if task_type and task_type not in _VALID_TASK_TYPES:
diff --git a/openviking/models/embedder/jina_embedders.py b/openviking/models/embedder/jina_embedders.py
@@ -99,6 +99,7 @@ def __init__(
             ValueError: If api_key is not provided
         """
         super().__init__(model_name, config)
+        self.provider = "jina"
 
         self.api_key = api_key
         self.api_base = api_base or "https://api.jina.ai/v1"
diff --git a/openviking/models/embedder/litellm_embedders.py b/openviking/models/embedder/litellm_embedders.py
@@ -68,6 +68,7 @@ def __init__(
             config: Additional configuration dict.
         """
         super().__init__(model_name, config)
+        self.provider = "litellm"
 
         os.environ.setdefault("LITELLM_LOCAL_MODEL_COST_MAP", "True")
 
diff --git a/openviking/models/embedder/minimax_embedders.py b/openviking/models/embedder/minimax_embedders.py
@@ -56,6 +56,7 @@ def __init__(
             extra_headers: Extra headers, useful for passing GroupId for MiniMax API
         """
         super().__init__(model_name, config)
+        self.provider = "minimax"
 
         self.api_key = api_key
         self.api_base = api_base or self.DEFAULT_API_BASE
diff --git a/openviking/models/embedder/openai_embedders.py b/openviking/models/embedder/openai_embedders.py
@@ -75,6 +75,7 @@ def __init__(
         extra_headers: Optional[Dict[str, str]] = None,
         input_type: Optional[str] = None,
         provider: str = "openai",
+        configured_provider: Optional[str] = None,
     ):
         """Initialize OpenAI-Compatible Dense Embedder
 
@@ -118,6 +119,7 @@ def __init__(
         self.query_param = query_param
         self.document_param = document_param
         self._provider = provider.lower()
+        self.provider = (configured_provider or provider).lower()
         self._client_kwargs: Dict[str, Any] = {"api_key": self.api_key or "no-key"}
 
         # Allow missing api_key when api_base is set (e.g. local OpenAI-compatible servers)
diff --git a/openviking/models/embedder/vikingdb_embedders.py b/openviking/models/embedder/vikingdb_embedders.py
@@ -160,6 +160,7 @@ def __init__(
         config: Optional[Dict[str, Any]] = None,
     ):
         DenseEmbedderBase.__init__(self, model_name, config)
+        self.provider = "vikingdb"
         self._init_vikingdb_client(ak, sk, region, host)
         self.model_version = model_version
         self.dimension = dimension
@@ -299,6 +300,7 @@ def __init__(
         config: Optional[Dict[str, Any]] = None,
     ):
         SparseEmbedderBase.__init__(self, model_name, config)
+        self.provider = "vikingdb"
         self._init_vikingdb_client(ak, sk, region, host)
         self.model_version = model_version
         self.sparse_model = {
@@ -438,6 +440,7 @@ def __init__(
         config: Optional[Dict[str, Any]] = None,
     ):
         HybridEmbedderBase.__init__(self, model_name, config)
+        self.provider = "vikingdb"
         self._init_vikingdb_client(ak, sk, region, host)
         self.model_version = model_version
         self.dimension = dimension
diff --git a/openviking/models/embedder/volcengine_embedders.py b/openviking/models/embedder/volcengine_embedders.py
@@ -81,6 +81,7 @@ def __init__(
             ValueError: If api_key is not provided
         """
         super().__init__(model_name, config)
+        self.provider = "volcengine"
 
         self.api_key = api_key
         self.api_base = api_base or "https://ark.cn-beijing.volces.com/api/v3"
@@ -326,6 +327,7 @@ def __init__(
             ValueError: If api_key is not provided
         """
         super().__init__(model_name, config)
+        self.provider = "volcengine"
 
         self.api_key = api_key
         self.api_base = api_base
@@ -512,6 +514,7 @@ def __init__(
             ValueError: If api_key is not provided
         """
         super().__init__(model_name, config)
+        self.provider = "volcengine"
         self.api_key = api_key
         self.api_base = api_base
         self.dimension = dimension
diff --git a/openviking/models/embedder/voyage_embedders.py b/openviking/models/embedder/voyage_embedders.py
@@ -60,6 +60,7 @@ def __init__(
         config: Optional[Dict[str, Any]] = None,
     ):
         super().__init__(model_name, config)
+        self.provider = "voyage"
 
         self.api_key = api_key
         self.api_base = api_base or "https://api.voyageai.com/v1"
diff --git a/openviking_cli/utils/config/embedding_config.py b/openviking_cli/utils/config/embedding_config.py
@@ -376,6 +376,7 @@ def _create_embedder(
                     "api_version": cfg.api_version,
                     "dimension": cfg.dimension,
                     "provider": "openai",
+                    "configured_provider": "openai",
                     "config": dict(runtime_config),
                     **({"query_param": cfg.query_param} if cfg.query_param else {}),
                     **({"document_param": cfg.document_param} if cfg.document_param else {}),
@@ -391,6 +392,7 @@ def _create_embedder(
                     "api_version": cfg.api_version,
                     "dimension": cfg.dimension,
                     "provider": "azure",
+                    "configured_provider": "azure",
                     "config": dict(runtime_config),
                     **({"query_param": cfg.query_param} if cfg.query_param else {}),
                     **({"document_param": cfg.document_param} if cfg.document_param else {}),
@@ -500,6 +502,7 @@ def _create_embedder(
                     or "no-key",  # Ollama ignores the key, but client requires non-empty
                     "api_base": cfg.api_base or "http://localhost:11434/v1",
                     "dimension": cfg.dimension,
+                    "configured_provider": "ollama",
                     "config": dict(runtime_config),
                 },
             ),
diff --git a/tests/test_session_async_commit.py b/tests/test_session_async_commit.py
@@ -18,6 +18,7 @@
 from openviking.service.task_tracker import TaskStatus, get_task_tracker, reset_task_tracker
 
 
+
 @pytest_asyncio.fixture
 async def api_client(temp_dir) -> AsyncGenerator[Tuple[httpx.AsyncClient, OpenVikingService], None]:
     """Create in-process HTTP client for API endpoint tests."""
diff --git a/tests/unit/test_extra_headers_embedding.py b/tests/unit/test_extra_headers_embedding.py
@@ -9,7 +9,7 @@
   4. api_key dead-code bug fix: no raise when api_base is set without api_key
 """
 
-from unittest.mock import MagicMock, patch
+from unittest.mock import AsyncMock, MagicMock, patch
 
 import pytest
 
@@ -115,6 +115,49 @@ def test_factory_injects_embedding_max_retries(self, mock_openai_class):
 
         assert embedder.max_retries == 0
 
+    @pytest.mark.asyncio
+    @patch("openviking.models.embedder.openai_embedders.openai.AsyncOpenAI")
+    @patch("openviking.models.embedder.openai_embedders.openai.OpenAI")
+    async def test_factory_uses_configured_provider_for_slow_call_logging(
+        self,
+        mock_openai_class,
+        mock_async_openai_class,
+    ):
+        """Slow-call warnings should log the configured provider, not the transport client mode."""
+        mock_openai_class.return_value = _make_mock_client()
+
+        async_response = MagicMock(
+            data=[MagicMock(embedding=[0.1] * 8)],
+            usage=None,
+        )
+        mock_async_client = MagicMock()
+        mock_async_client.embeddings.create = AsyncMock(return_value=async_response)
+        mock_async_openai_class.return_value = mock_async_client
+
+        cfg = EmbeddingModelConfig(
+            provider="ollama",
+            model="nomic-embed-text",
+            api_base="http://localhost:11434/v1",
+            dimension=8,
+        )
+        embedder = EmbeddingConfig(dense=cfg)._create_embedder("ollama", "dense", cfg)
+
+        with (
+            patch(
+                "openviking.models.embedder.openai_embedders.logger.warning"
+            ) as mock_warning,
+            patch(
+                "openviking.models.embedder.base.time.monotonic",
+                side_effect=[0.0, 0.0, 0.0, 1.2],
+            ),
+        ):
+            await embedder.embed_async("hello")
+
+        mock_warning.assert_called_once()
+        call_args = mock_warning.call_args.args
+        assert call_args[1] == "OpenAI async embedding"
+        assert call_args[2] == "ollama"
+
 
 class TestEmbeddingModelConfigExtraHeaders:
     """Test that EmbeddingModelConfig accepts and stores the extra_headers field."""
