llama-stack 0.3.2__tar.gz → 0.3.3__tar.gz

{llama_stack-0.3.2/llama_stack.egg-info → llama_stack-0.3.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: llama_stack
-Version: 0.3.2
+Version: 0.3.3
 Summary: Llama Stack
 Author-email: Meta Llama <llama-oss@meta.com>
 License: MIT
@@ -22,7 +22,7 @@ Requires-Dist: fire
 Requires-Dist: httpx
 Requires-Dist: jinja2>=3.1.6
 Requires-Dist: jsonschema
-Requires-Dist: llama-stack-client>=0.3.2
+Requires-Dist: llama-stack-client>=0.3.3
 Requires-Dist: openai>=1.107
 Requires-Dist: prompt-toolkit
 Requires-Dist: python-dotenv
@@ -44,7 +44,7 @@ Requires-Dist: sqlalchemy[asyncio]>=2.0.41
 Provides-Extra: ui
 Requires-Dist: streamlit; extra == "ui"
 Requires-Dist: pandas; extra == "ui"
-Requires-Dist: llama-stack-client>=0.3.2; extra == "ui"
+Requires-Dist: llama-stack-client>=0.3.3; extra == "ui"
 Requires-Dist: streamlit-option-menu; extra == "ui"
 Dynamic: license-file
 

{llama_stack-0.3.2 → llama_stack-0.3.3}/llama_stack/providers/inline/vector_io/faiss/faiss.py

@@ -223,7 +223,8 @@ class FaissVectorIOAdapter(OpenAIVectorStoreMixin, VectorIO, VectorStoresProtoco
             return HealthResponse(status=HealthStatus.ERROR, message=f"Health check failed: {str(e)}")
 
     async def register_vector_store(self, vector_store: VectorStore) -> None:
-        assert self.kvstore is not None
+        if self.kvstore is None:
+            raise RuntimeError("KVStore not initialized. Call initialize() before registering vector stores.")
 
         key = f"{VECTOR_DBS_PREFIX}{vector_store.identifier}"
         await self.kvstore.set(key=key, value=vector_store.model_dump_json())
@@ -239,7 +240,8 @@ class FaissVectorIOAdapter(OpenAIVectorStoreMixin, VectorIO, VectorStoresProtoco
         return [i.vector_store for i in self.cache.values()]
 
     async def unregister_vector_store(self, vector_store_id: str) -> None:
-        assert self.kvstore is not None
+        if self.kvstore is None:
+            raise RuntimeError("KVStore not initialized. Call initialize() before unregistering vector stores.")
 
         if vector_store_id not in self.cache:
             return
@@ -248,6 +250,27 @@ class FaissVectorIOAdapter(OpenAIVectorStoreMixin, VectorIO, VectorStoresProtoco
         del self.cache[vector_store_id]
         await self.kvstore.delete(f"{VECTOR_DBS_PREFIX}{vector_store_id}")
 
+    async def _get_and_cache_vector_store_index(self, vector_store_id: str) -> VectorStoreWithIndex | None:
+        if vector_store_id in self.cache:
+            return self.cache[vector_store_id]
+
+        if self.kvstore is None:
+            raise RuntimeError("KVStore not initialized. Call initialize() before using vector stores.")
+
+        key = f"{VECTOR_DBS_PREFIX}{vector_store_id}"
+        vector_store_data = await self.kvstore.get(key)
+        if not vector_store_data:
+            raise VectorStoreNotFoundError(vector_store_id)
+
+        vector_store = VectorStore.model_validate_json(vector_store_data)
+        index = VectorStoreWithIndex(
+            vector_store=vector_store,
+            index=await FaissIndex.create(vector_store.embedding_dimension, self.kvstore, vector_store.identifier),
+            inference_api=self.inference_api,
+        )
+        self.cache[vector_store_id] = index
+        return index
+
     async def insert_chunks(self, vector_db_id: str, chunks: list[Chunk], ttl_seconds: int | None = None) -> None:
         index = self.cache.get(vector_db_id)
         if index is None:

{llama_stack-0.3.2 → llama_stack-0.3.3}/llama_stack/providers/inline/vector_io/sqlite_vec/sqlite_vec.py

@@ -412,6 +412,14 @@ class SQLiteVecVectorIOAdapter(OpenAIVectorStoreMixin, VectorIO, VectorStoresPro
         return [v.vector_store for v in self.cache.values()]
 
     async def register_vector_store(self, vector_store: VectorStore) -> None:
+        if self.kvstore is None:
+            raise RuntimeError("KVStore not initialized. Call initialize() before registering vector stores.")
+
+        # Save to kvstore for persistence
+        key = f"{VECTOR_DBS_PREFIX}{vector_store.identifier}"
+        await self.kvstore.set(key=key, value=vector_store.model_dump_json())
+
+        # Create and cache the index
         index = await SQLiteVecIndex.create(
             vector_store.embedding_dimension, self.config.db_path, vector_store.identifier
         )
@@ -421,13 +429,16 @@ class SQLiteVecVectorIOAdapter(OpenAIVectorStoreMixin, VectorIO, VectorStoresPro
         if vector_store_id in self.cache:
             return self.cache[vector_store_id]
 
-        if self.vector_store_table is None:
-            raise VectorStoreNotFoundError(vector_store_id)
+        # Try to load from kvstore
+        if self.kvstore is None:
+            raise RuntimeError("KVStore not initialized. Call initialize() before using vector stores.")
 
-        vector_store = self.vector_store_table.get_vector_store(vector_store_id)
-        if not vector_store:
+        key = f"{VECTOR_DBS_PREFIX}{vector_store_id}"
+        vector_store_data = await self.kvstore.get(key)
+        if not vector_store_data:
             raise VectorStoreNotFoundError(vector_store_id)
 
+        vector_store = VectorStore.model_validate_json(vector_store_data)
         index = VectorStoreWithIndex(
             vector_store=vector_store,
             index=SQLiteVecIndex(

{llama_stack-0.3.2 → llama_stack-0.3.3}/llama_stack/providers/remote/inference/vertexai/vertexai.py

@@ -4,6 +4,7 @@
 # This source code is licensed under the terms described in the LICENSE file in
 # the root directory of this source tree.
 
+from collections.abc import Iterable
 
 import google.auth.transport.requests
 from google.auth import default
@@ -42,3 +43,12 @@ class VertexAIInferenceAdapter(OpenAIMixin):
         Source: https://cloud.google.com/vertex-ai/generative-ai/docs/start/openai
         """
         return f"https://{self.config.location}-aiplatform.googleapis.com/v1/projects/{self.config.project}/locations/{self.config.location}/endpoints/openapi"
+
+    async def list_provider_model_ids(self) -> Iterable[str]:
+        """
+        VertexAI doesn't currently offer a way to query a list of available models from Google's Model Garden
+        For now we return a hardcoded version of the available models
+
+        :return: An iterable of model IDs
+        """
+        return ["google/gemini-2.0-flash", "google/gemini-2.5-flash", "google/gemini-2.5-pro"]

{llama_stack-0.3.2 → llama_stack-0.3.3}/llama_stack/providers/remote/vector_io/chroma/chroma.py

@@ -131,7 +131,6 @@ class ChromaVectorIOAdapter(OpenAIVectorStoreMixin, VectorIO, VectorStoresProtoc
 
     async def initialize(self) -> None:
         self.kvstore = await kvstore_impl(self.config.persistence)
-        self.vector_store_table = self.kvstore
 
         if isinstance(self.config, RemoteChromaVectorIOConfig):
             log.info(f"Connecting to Chroma server at: {self.config.url}")
@@ -190,9 +189,16 @@ class ChromaVectorIOAdapter(OpenAIVectorStoreMixin, VectorIO, VectorStoresProtoc
         if vector_store_id in self.cache:
             return self.cache[vector_store_id]
 
-        vector_store = await self.vector_store_table.get_vector_store(vector_store_id)
-        if not vector_store:
+        # Try to load from kvstore
+        if self.kvstore is None:
+            raise RuntimeError("KVStore not initialized. Call initialize() before using vector stores.")
+
+        key = f"{VECTOR_DBS_PREFIX}{vector_store_id}"
+        vector_store_data = await self.kvstore.get(key)
+        if not vector_store_data:
             raise ValueError(f"Vector DB {vector_store_id} not found in Llama Stack")
+
+        vector_store = VectorStore.model_validate_json(vector_store_data)
         collection = await maybe_await(self.client.get_collection(vector_store_id))
         if not collection:
             raise ValueError(f"Vector DB {vector_store_id} not found in Chroma")

{llama_stack-0.3.2 → llama_stack-0.3.3}/llama_stack/providers/remote/vector_io/milvus/milvus.py

@@ -328,13 +328,16 @@ class MilvusVectorIOAdapter(OpenAIVectorStoreMixin, VectorIO, VectorStoresProtoc
         if vector_store_id in self.cache:
             return self.cache[vector_store_id]
 
-        if self.vector_store_table is None:
-            raise VectorStoreNotFoundError(vector_store_id)
+        # Try to load from kvstore
+        if self.kvstore is None:
+            raise RuntimeError("KVStore not initialized. Call initialize() before using vector stores.")
 
-        vector_store = await self.vector_store_table.get_vector_store(vector_store_id)
-        if not vector_store:
+        key = f"{VECTOR_DBS_PREFIX}{vector_store_id}"
+        vector_store_data = await self.kvstore.get(key)
+        if not vector_store_data:
             raise VectorStoreNotFoundError(vector_store_id)
 
+        vector_store = VectorStore.model_validate_json(vector_store_data)
         index = VectorStoreWithIndex(
             vector_store=vector_store,
             index=MilvusIndex(client=self.client, collection_name=vector_store.identifier, kvstore=self.kvstore),

{llama_stack-0.3.2 → llama_stack-0.3.3}/llama_stack/providers/remote/vector_io/pgvector/pgvector.py

@@ -368,6 +368,22 @@ class PGVectorVectorIOAdapter(OpenAIVectorStoreMixin, VectorIO, VectorStoresProt
             log.exception("Could not connect to PGVector database server")
             raise RuntimeError("Could not connect to PGVector database server") from e
 
+        # Load existing vector stores from KV store into cache
+        start_key = VECTOR_DBS_PREFIX
+        end_key = f"{VECTOR_DBS_PREFIX}\xff"
+        stored_vector_stores = await self.kvstore.values_in_range(start_key, end_key)
+        for vector_store_data in stored_vector_stores:
+            vector_store = VectorStore.model_validate_json(vector_store_data)
+            pgvector_index = PGVectorIndex(
+                vector_store=vector_store,
+                dimension=vector_store.embedding_dimension,
+                conn=self.conn,
+                kvstore=self.kvstore,
+            )
+            await pgvector_index.initialize()
+            index = VectorStoreWithIndex(vector_store, index=pgvector_index, inference_api=self.inference_api)
+            self.cache[vector_store.identifier] = index
+
     async def shutdown(self) -> None:
         if self.conn is not None:
             self.conn.close()
@@ -377,7 +393,13 @@ class PGVectorVectorIOAdapter(OpenAIVectorStoreMixin, VectorIO, VectorStoresProt
 
     async def register_vector_store(self, vector_store: VectorStore) -> None:
         # Persist vector DB metadata in the KV store
-        assert self.kvstore is not None
+        if self.kvstore is None:
+            raise RuntimeError("KVStore not initialized. Call initialize() before registering vector stores.")
+
+        # Save to kvstore for persistence
+        key = f"{VECTOR_DBS_PREFIX}{vector_store.identifier}"
+        await self.kvstore.set(key=key, value=vector_store.model_dump_json())
+
         # Upsert model metadata in Postgres
         upsert_models(self.conn, [(vector_store.identifier, vector_store)])
 
@@ -396,7 +418,8 @@ class PGVectorVectorIOAdapter(OpenAIVectorStoreMixin, VectorIO, VectorStoresProt
             del self.cache[vector_store_id]
 
         # Delete vector DB metadata from KV store
-        assert self.kvstore is not None
+        if self.kvstore is None:
+            raise RuntimeError("KVStore not initialized. Call initialize() before unregistering vector stores.")
         await self.kvstore.delete(key=f"{VECTOR_DBS_PREFIX}{vector_store_id}")
 
     async def insert_chunks(self, vector_db_id: str, chunks: list[Chunk], ttl_seconds: int | None = None) -> None:
@@ -413,13 +436,16 @@ class PGVectorVectorIOAdapter(OpenAIVectorStoreMixin, VectorIO, VectorStoresProt
         if vector_store_id in self.cache:
             return self.cache[vector_store_id]
 
-        if self.vector_store_table is None:
-            raise VectorStoreNotFoundError(vector_store_id)
+        # Try to load from kvstore
+        if self.kvstore is None:
+            raise RuntimeError("KVStore not initialized. Call initialize() before using vector stores.")
 
-        vector_store = await self.vector_store_table.get_vector_store(vector_store_id)
-        if not vector_store:
+        key = f"{VECTOR_DBS_PREFIX}{vector_store_id}"
+        vector_store_data = await self.kvstore.get(key)
+        if not vector_store_data:
             raise VectorStoreNotFoundError(vector_store_id)
 
+        vector_store = VectorStore.model_validate_json(vector_store_data)
         index = PGVectorIndex(vector_store, vector_store.embedding_dimension, self.conn)
         await index.initialize()
         self.cache[vector_store_id] = VectorStoreWithIndex(vector_store, index, self.inference_api)

{llama_stack-0.3.2 → llama_stack-0.3.3}/llama_stack/providers/remote/vector_io/qdrant/qdrant.py

@@ -183,7 +183,8 @@ class QdrantVectorIOAdapter(OpenAIVectorStoreMixin, VectorIO, VectorStoresProtoc
         await super().shutdown()
 
     async def register_vector_store(self, vector_store: VectorStore) -> None:
-        assert self.kvstore is not None
+        if self.kvstore is None:
+            raise RuntimeError("KVStore not initialized. Call initialize() before registering vector stores.")
         key = f"{VECTOR_DBS_PREFIX}{vector_store.identifier}"
         await self.kvstore.set(key=key, value=vector_store.model_dump_json())
 
@@ -200,20 +201,24 @@ class QdrantVectorIOAdapter(OpenAIVectorStoreMixin, VectorIO, VectorStoresProtoc
             await self.cache[vector_store_id].index.delete()
             del self.cache[vector_store_id]
 
-        assert self.kvstore is not None
+        if self.kvstore is None:
+            raise RuntimeError("KVStore not initialized. Call initialize() before using vector stores.")
         await self.kvstore.delete(f"{VECTOR_DBS_PREFIX}{vector_store_id}")
 
     async def _get_and_cache_vector_store_index(self, vector_store_id: str) -> VectorStoreWithIndex | None:
         if vector_store_id in self.cache:
             return self.cache[vector_store_id]
 
-        if self.vector_store_table is None:
-            raise ValueError(f"Vector DB not found {vector_store_id}")
+        # Try to load from kvstore
+        if self.kvstore is None:
+            raise RuntimeError("KVStore not initialized. Call initialize() before using vector stores.")
 
-        vector_store = await self.vector_store_table.get_vector_store(vector_store_id)
-        if not vector_store:
+        key = f"{VECTOR_DBS_PREFIX}{vector_store_id}"
+        vector_store_data = await self.kvstore.get(key)
+        if not vector_store_data:
             raise VectorStoreNotFoundError(vector_store_id)
 
+        vector_store = VectorStore.model_validate_json(vector_store_data)
         index = VectorStoreWithIndex(
             vector_store=vector_store,
             index=QdrantIndex(client=self.client, collection_name=vector_store.identifier),

{llama_stack-0.3.2 → llama_stack-0.3.3}/llama_stack/providers/remote/vector_io/weaviate/weaviate.py

@@ -346,13 +346,16 @@ class WeaviateVectorIOAdapter(OpenAIVectorStoreMixin, VectorIO, NeedsRequestProv
         if vector_store_id in self.cache:
             return self.cache[vector_store_id]
 
-        if self.vector_store_table is None:
-            raise VectorStoreNotFoundError(vector_store_id)
+        # Try to load from kvstore
+        if self.kvstore is None:
+            raise RuntimeError("KVStore not initialized. Call initialize() before using vector stores.")
 
-        vector_store = await self.vector_store_table.get_vector_store(vector_store_id)
-        if not vector_store:
+        key = f"{VECTOR_DBS_PREFIX}{vector_store_id}"
+        vector_store_data = await self.kvstore.get(key)
+        if not vector_store_data:
             raise VectorStoreNotFoundError(vector_store_id)
 
+        vector_store = VectorStore.model_validate_json(vector_store_data)
         client = self._get_client()
         sanitized_collection_name = sanitize_collection_name(vector_store.identifier, weaviate_format=True)
         if not client.collections.exists(sanitized_collection_name):

{llama_stack-0.3.2 → llama_stack-0.3.3}/llama_stack/providers/utils/inference/inference_store.py

@@ -35,6 +35,7 @@ class InferenceStore:
         self.reference = reference
         self.sql_store = None
         self.policy = policy
+        self.enable_write_queue = True
 
         # Async write queue and worker control
         self._queue: asyncio.Queue[tuple[OpenAIChatCompletion, list[OpenAIMessageParam]]] | None = None
@@ -47,14 +48,13 @@ class InferenceStore:
         base_store = sqlstore_impl(self.reference)
         self.sql_store = AuthorizedSqlStore(base_store, self.policy)
 
-        # Disable write queue for SQLite to avoid concurrency issues
-        backend_name = self.reference.backend
-        backend_config = _SQLSTORE_BACKENDS.get(backend_name)
-        if backend_config is None:
-            raise ValueError(
-                f"Unregistered SQL backend '{backend_name}'. Registered backends: {sorted(_SQLSTORE_BACKENDS)}"
-            )
-        self.enable_write_queue = backend_config.type != StorageBackendType.SQL_SQLITE
+        # Disable write queue for SQLite since WAL mode handles concurrency
+        # Keep it enabled for other backends (like Postgres) for performance
+        backend_config = _SQLSTORE_BACKENDS.get(self.reference.backend)
+        if backend_config and backend_config.type == StorageBackendType.SQL_SQLITE:
+            self.enable_write_queue = False
+            logger.debug("Write queue disabled for SQLite (WAL mode handles concurrency)")
+
         await self.sql_store.create_table(
             "chat_completions",
             {
@@ -66,6 +66,14 @@ class InferenceStore:
             },
         )
 
+        if self.enable_write_queue:
+            self._queue = asyncio.Queue(maxsize=self._max_write_queue_size)
+            for _ in range(self._num_writers):
+                self._worker_tasks.append(asyncio.create_task(self._worker_loop()))
+            logger.debug(
+                f"Inference store write queue enabled with {self._num_writers} writers, max queue size {self._max_write_queue_size}"
+            )
+
     async def shutdown(self) -> None:
         if not self._worker_tasks:
             return

{llama_stack-0.3.2 → llama_stack-0.3.3}/llama_stack/providers/utils/inference/model_registry.py

@@ -20,7 +20,7 @@ logger = get_logger(name=__name__, category="providers::utils")
 
 
 class RemoteInferenceProviderConfig(BaseModel):
-    allowed_models: list[str] | None = Field(  # TODO: make this non-optional and give a list() default
+    allowed_models: list[str] | None = Field(
         default=None,
         description="List of models that should be registered with the model registry. If None, all models are allowed.",
     )

{llama_stack-0.3.2 → llama_stack-0.3.3}/llama_stack/providers/utils/inference/openai_mixin.py

@@ -82,9 +82,6 @@ class OpenAIMixin(NeedsRequestProviderData, ABC, BaseModel):
     # This is set in list_models() and used in check_model_availability()
     _model_cache: dict[str, Model] = {}
 
-    # List of allowed models for this provider, if empty all models allowed
-    allowed_models: list[str] = []
-
     # Optional field name in provider data to look for API key, which takes precedence
     provider_data_api_key_field: str | None = None
 
@@ -191,6 +188,19 @@ class OpenAIMixin(NeedsRequestProviderData, ABC, BaseModel):
 
         return api_key
 
+    def _validate_model_allowed(self, provider_model_id: str) -> None:
+        """
+        Validate that the model is in the allowed_models list if configured.
+
+        :param provider_model_id: The provider-specific model ID to validate
+        :raises ValueError: If the model is not in the allowed_models list
+        """
+        if self.config.allowed_models is not None and provider_model_id not in self.config.allowed_models:
+            raise ValueError(
+                f"Model '{provider_model_id}' is not in the allowed models list. "
+                f"Allowed models: {self.config.allowed_models}"
+            )
+
     async def _get_provider_model_id(self, model: str) -> str:
         """
         Get the provider-specific model ID from the model store.
@@ -237,8 +247,11 @@ class OpenAIMixin(NeedsRequestProviderData, ABC, BaseModel):
         Direct OpenAI completion API call.
         """
         # TODO: fix openai_completion to return type compatible with OpenAI's API response
+        provider_model_id = await self._get_provider_model_id(params.model)
+        self._validate_model_allowed(provider_model_id)
+
         completion_kwargs = await prepare_openai_completion_params(
-            model=await self._get_provider_model_id(params.model),
+            model=provider_model_id,
             prompt=params.prompt,
             best_of=params.best_of,
             echo=params.echo,
@@ -270,6 +283,9 @@ class OpenAIMixin(NeedsRequestProviderData, ABC, BaseModel):
         """
         Direct OpenAI chat completion API call.
         """
+        provider_model_id = await self._get_provider_model_id(params.model)
+        self._validate_model_allowed(provider_model_id)
+
         messages = params.messages
 
         if self.download_images:
@@ -291,7 +307,7 @@ class OpenAIMixin(NeedsRequestProviderData, ABC, BaseModel):
             messages = [await _localize_image_url(m) for m in messages]
 
         request_params = await prepare_openai_completion_params(
-            model=await self._get_provider_model_id(params.model),
+            model=provider_model_id,
             messages=messages,
             frequency_penalty=params.frequency_penalty,
             function_call=params.function_call,
@@ -329,9 +345,13 @@ class OpenAIMixin(NeedsRequestProviderData, ABC, BaseModel):
         """
         Direct OpenAI embeddings API call.
         """
-        # Prepare request parameters
-        request_params = {
-            "model": await self._get_provider_model_id(params.model),
+        provider_model_id = await self._get_provider_model_id(params.model)
+        self._validate_model_allowed(provider_model_id)
+
+        # Build request params conditionally to avoid NotGiven/Omit type mismatch
+        # The OpenAI SDK uses Omit in signatures but NOT_GIVEN has type NotGiven
+        request_params: dict[str, Any] = {
+            "model": provider_model_id,
             "input": params.input,
             "encoding_format": params.encoding_format if params.encoding_format is not None else NOT_GIVEN,
             "dimensions": params.dimensions if params.dimensions is not None else NOT_GIVEN,
@@ -416,7 +436,7 @@ class OpenAIMixin(NeedsRequestProviderData, ABC, BaseModel):
         for provider_model_id in provider_models_ids:
             if not isinstance(provider_model_id, str):
                 raise ValueError(f"Model ID {provider_model_id} from list_provider_model_ids() is not a string")
-            if self.allowed_models and provider_model_id not in self.allowed_models:
+            if self.config.allowed_models is not None and provider_model_id not in self.config.allowed_models:
                 logger.info(f"Skipping model {provider_model_id} as it is not in the allowed models list")
                 continue
             if metadata := self.embedding_model_metadata.get(provider_model_id):

{llama_stack-0.3.2 → llama_stack-0.3.3}/llama_stack/providers/utils/responses/responses_store.py

@@ -3,6 +3,7 @@
 #
 # This source code is licensed under the terms described in the LICENSE file in
 # the root directory of this source tree.
+import asyncio
 
 from llama_stack.apis.agents import (
     Order,
@@ -17,12 +18,12 @@ from llama_stack.apis.agents.openai_responses import (
 )
 from llama_stack.apis.inference import OpenAIMessageParam
 from llama_stack.core.datatypes import AccessRule
-from llama_stack.core.storage.datatypes import ResponsesStoreReference, SqlStoreReference
+from llama_stack.core.storage.datatypes import ResponsesStoreReference, SqlStoreReference, StorageBackendType
 from llama_stack.log import get_logger
 
 from ..sqlstore.api import ColumnDefinition, ColumnType
 from ..sqlstore.authorized_sqlstore import AuthorizedSqlStore
-from ..sqlstore.sqlstore import sqlstore_impl
+from ..sqlstore.sqlstore import _SQLSTORE_BACKENDS, sqlstore_impl
 
 logger = get_logger(name=__name__, category="openai_responses")
 
@@ -59,6 +60,13 @@ class ResponsesStore:
         base_store = sqlstore_impl(self.reference)
         self.sql_store = AuthorizedSqlStore(base_store, self.policy)
 
+        # Disable write queue for SQLite since WAL mode handles concurrency
+        # Keep it enabled for other backends (like Postgres) for performance
+        backend_config = _SQLSTORE_BACKENDS.get(self.reference.backend)
+        if backend_config and backend_config.type == StorageBackendType.SQL_SQLITE:
+            self.enable_write_queue = False
+            logger.debug("Write queue disabled for SQLite (WAL mode handles concurrency)")
+
         await self.sql_store.create_table(
             "openai_responses",
             {
@@ -77,6 +85,14 @@ class ResponsesStore:
             },
         )
 
+        if self.enable_write_queue:
+            self._queue = asyncio.Queue(maxsize=self._max_write_queue_size)
+            for _ in range(self._num_writers):
+                self._worker_tasks.append(asyncio.create_task(self._worker_loop()))
+            logger.debug(
+                f"Responses store write queue enabled with {self._num_writers} writers, max queue size {self._max_write_queue_size}"
+            )
+
     async def shutdown(self) -> None:
         return
 

{llama_stack-0.3.2 → llama_stack-0.3.3}/llama_stack/providers/utils/sqlstore/sqlalchemy_sqlstore.py

@@ -17,6 +17,7 @@ from sqlalchemy import (
     String,
     Table,
     Text,
+    event,
     inspect,
     select,
     text,
@@ -75,7 +76,36 @@ class SqlAlchemySqlStoreImpl(SqlStore):
         self.metadata = MetaData()
 
     def create_engine(self) -> AsyncEngine:
-        return create_async_engine(self.config.engine_str, pool_pre_ping=True)
+        # Configure connection args for better concurrency support
+        connect_args = {}
+        if "sqlite" in self.config.engine_str:
+            # SQLite-specific optimizations for concurrent access
+            # With WAL mode, most locks resolve in milliseconds, but allow up to 5s for edge cases
+            connect_args["timeout"] = 5.0
+            connect_args["check_same_thread"] = False  # Allow usage across asyncio tasks
+
+        engine = create_async_engine(
+            self.config.engine_str,
+            pool_pre_ping=True,
+            connect_args=connect_args,
+        )
+
+        # Enable WAL mode for SQLite to support concurrent readers and writers
+        if "sqlite" in self.config.engine_str:
+
+            @event.listens_for(engine.sync_engine, "connect")
+            def set_sqlite_pragma(dbapi_conn, connection_record):
+                cursor = dbapi_conn.cursor()
+                # Enable Write-Ahead Logging for better concurrency
+                cursor.execute("PRAGMA journal_mode=WAL")
+                # Set busy timeout to 5 seconds (retry instead of immediate failure)
+                # With WAL mode, locks should be brief; if we hit 5s there's a bigger issue
+                cursor.execute("PRAGMA busy_timeout=5000")
+                # Use NORMAL synchronous mode for better performance (still safe with WAL)
+                cursor.execute("PRAGMA synchronous=NORMAL")
+                cursor.close()
+
+        return engine
 
     async def create_table(
         self,

{llama_stack-0.3.2 → llama_stack-0.3.3/llama_stack.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: llama_stack
-Version: 0.3.2
+Version: 0.3.3
 Summary: Llama Stack
 Author-email: Meta Llama <llama-oss@meta.com>
 License: MIT
@@ -22,7 +22,7 @@ Requires-Dist: fire
 Requires-Dist: httpx
 Requires-Dist: jinja2>=3.1.6
 Requires-Dist: jsonschema
-Requires-Dist: llama-stack-client>=0.3.2
+Requires-Dist: llama-stack-client>=0.3.3
 Requires-Dist: openai>=1.107
 Requires-Dist: prompt-toolkit
 Requires-Dist: python-dotenv
@@ -44,7 +44,7 @@ Requires-Dist: sqlalchemy[asyncio]>=2.0.41
 Provides-Extra: ui
 Requires-Dist: streamlit; extra == "ui"
 Requires-Dist: pandas; extra == "ui"
-Requires-Dist: llama-stack-client>=0.3.2; extra == "ui"
+Requires-Dist: llama-stack-client>=0.3.3; extra == "ui"
 Requires-Dist: streamlit-option-menu; extra == "ui"
 Dynamic: license-file
 

{llama_stack-0.3.2 → llama_stack-0.3.3}/llama_stack.egg-info/requires.txt

@@ -4,7 +4,7 @@ fire
 httpx
 jinja2>=3.1.6
 jsonschema
-llama-stack-client>=0.3.2
+llama-stack-client>=0.3.3
 openai>=1.107
 prompt-toolkit
 python-dotenv
@@ -27,5 +27,5 @@ sqlalchemy[asyncio]>=2.0.41
 [ui]
 streamlit
 pandas
-llama-stack-client>=0.3.2
+llama-stack-client>=0.3.3
 streamlit-option-menu

{llama_stack-0.3.2 → llama_stack-0.3.3}/pyproject.toml

@@ -7,7 +7,7 @@ required-version = ">=0.7.0"
 
 [project]
 name = "llama_stack"
-version = "0.3.2"
+version = "0.3.3"
 authors = [{ name = "Meta Llama", email = "llama-oss@meta.com" }]
 description = "Llama Stack"
 readme = "README.md"
@@ -30,7 +30,7 @@ dependencies = [
     "httpx",
     "jinja2>=3.1.6",
     "jsonschema",
-    "llama-stack-client>=0.3.2",
+    "llama-stack-client>=0.3.3",
     "openai>=1.107",                                  # for expires_after support
     "prompt-toolkit",
     "python-dotenv",
@@ -55,7 +55,7 @@ dependencies = [
 ui = [
     "streamlit",
     "pandas",
-    "llama-stack-client>=0.3.2",
+    "llama-stack-client>=0.3.3",
     "streamlit-option-menu",
 ]