phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-12-12 20:12:33 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

group all server stores under storage.stores

Browse source
This commit is contained in:
Ashwin Bharambe 2025-10-19 08:06:19 -07:00
parent 636764c2a1
commit 2bba56a0a8
35 changed files with 806 additions and 503 deletions
Download patch file Download diff file Expand all files Collapse all files

31
llama_stack/core/configure.py
View file

@ -159,6 +159,37 @@ def upgrade_from_routing_table(
config_dict["apis"] = config_dict["apis_to_serve"]
config_dict.pop("apis_to_serve", None)
# Add default storage config if not present
if "storage" not in config_dict:
config_dict["storage"] = {
"backends": {
"kv_default": {
"type": "kv_sqlite",
"db_path": "~/.llama/kvstore.db",
},
"sql_default": {
"type": "sql_sqlite",
"db_path": "~/.llama/sql_store.db",
},
},
"stores": {
"metadata": {
"namespace": "registry",
"backend": "kv_default",
},
"inference": {
"table_name": "inference_store",
"backend": "sql_default",
"max_write_queue_size": 10000,
"num_writers": 4,
},
"conversations": {
"table_name": "openai_conversations",
"backend": "sql_default",
},
},
}
return config_dict

4
llama_stack/core/conversations/conversations.py
View file

@ -56,9 +56,9 @@ class ConversationServiceImpl(Conversations):
self.policy = config.policy
# Use conversations store reference from run config
conversations_ref = config.run_config.conversations_store
conversations_ref = config.run_config.storage.stores.conversations
if not conversations_ref:
raise ValueError("conversations_store must be configured in run config")
raise ValueError("storage.stores.conversations must be configured in run config")
base_sql_store = sqlstore_impl(conversations_ref)
self.sql_store = AuthorizedSqlStore(base_sql_store, self.policy)

28
llama_stack/core/datatypes.py
View file

@ -27,9 +27,7 @@ from llama_stack.apis.vector_dbs import VectorDB, VectorDBInput
from llama_stack.apis.vector_io import VectorIO
from llama_stack.core.access_control.datatypes import AccessRule
from llama_stack.core.storage.datatypes import (
InferenceStoreReference,
KVStoreReference,
SqlStoreReference,
StorageBackendType,
StorageConfig,
)
@ -470,19 +468,7 @@ can be instantiated multiple times (with different configs) if necessary.
""",
)
storage: StorageConfig = Field(
description="Catalog of named storage backends available to the stack",
)
metadata_store: KVStoreReference | None = Field(
default=None,
description="Reference to the KV store backend used by the distribution registry (kv_* backend).",
)
inference_store: InferenceStoreReference | None = Field(
default=None,
description="Reference to the SQL store backend used by the inference API (sql_* backend).",
)
conversations_store: SqlStoreReference | None = Field(
default=None,
description="Reference to the SQL store backend used by the conversations API (sql_* backend).",
description="Catalog of named storage backends and references available to the stack",
)
# registry of "resources" in the distribution
@ -523,8 +509,9 @@ can be instantiated multiple times (with different configs) if necessary.
return v
@model_validator(mode="after")
def validate_storage_references(self) -> "StackRunConfig":
backend_map = self.storage.backends if self.storage else {}
def validate_server_stores(self) -> "StackRunConfig":
backend_map = self.storage.backends
stores = self.storage.stores
kv_backends = {
name
for name, cfg in backend_map.items()
@ -558,9 +545,10 @@ can be instantiated multiple times (with different configs) if necessary.
f"{'kv_*' if expected_set is kv_backends else 'sql_*'} is required."
)
_ensure_backend(self.metadata_store, kv_backends, "metadata_store")
_ensure_backend(self.inference_store, sql_backends, "inference_store")
_ensure_backend(self.conversations_store, sql_backends, "conversations_store")
_ensure_backend(stores.metadata, kv_backends, "storage.stores.metadata")
_ensure_backend(stores.inference, sql_backends, "storage.stores.inference")
_ensure_backend(stores.conversations, sql_backends, "storage.stores.conversations")
_ensure_backend(stores.responses, sql_backends, "storage.stores.responses")
return self

4
llama_stack/core/prompts/prompts.py
View file

@ -41,9 +41,9 @@ class PromptServiceImpl(Prompts):
async def initialize(self) -> None:
# Use metadata store backend with prompts-specific namespace
metadata_ref = self.config.run_config.metadata_store
metadata_ref = self.config.run_config.storage.stores.metadata
if not metadata_ref:
raise ValueError("metadata_store must be configured in run config")
raise ValueError("storage.stores.metadata must be configured in run config")
prompts_ref = KVStoreReference(namespace="prompts", backend=metadata_ref.backend)
self.kvstore = await kvstore_impl(prompts_ref)

4
llama_stack/core/routers/__init__.py
View file

@ -80,9 +80,9 @@ async def get_auto_router_impl(
# TODO: move pass configs to routers instead
if api == Api.inference:
inference_ref = run_config.inference_store
inference_ref = run_config.storage.stores.inference
if not inference_ref:
raise ValueError("inference_store must be configured in run config")
raise ValueError("storage.stores.inference must be configured in run config")
inference_store = InferenceStore(
reference=inference_ref,

7
llama_stack/core/stack.py
View file

@ -368,9 +368,10 @@ class Stack:
logger.info(f"API recording enabled: mode={os.environ.get('LLAMA_STACK_TEST_INFERENCE_MODE')}")
_initialize_storage(self.run_config)
if not self.run_config.metadata_store:
raise ValueError("metadata_store must be configured with a kv_* backend")
dist_registry, _ = await create_dist_registry(self.run_config.metadata_store, self.run_config.image_name)
stores = self.run_config.storage.stores
if not stores.metadata:
raise ValueError("storage.stores.metadata must be configured with a kv_* backend")
dist_registry, _ = await create_dist_registry(stores.metadata, self.run_config.image_name)
policy = self.run_config.server.auth.access_policy if self.run_config.server.auth else []
internal_impls = {}

27
llama_stack/core/storage/datatypes.py
View file

@ -72,7 +72,7 @@ class SqliteKVStoreConfig(CommonConfig):
class PostgresKVStoreConfig(CommonConfig):
type: Literal[StorageBackendType.KV_POSTGRES] = StorageBackendType.KV_POSTGRES
host: str = "localhost"
port: int = 5432
port: int | str = 5432
db: str = "llamastack"
user: str
password: str | None = None
@ -175,7 +175,7 @@ class SqliteSqlStoreConfig(SqlAlchemySqlStoreConfig):
class PostgresSqlStoreConfig(SqlAlchemySqlStoreConfig):
type: Literal[StorageBackendType.SQL_POSTGRES] = StorageBackendType.SQL_POSTGRES
host: str = "localhost"
port: int = 5432
port: int | str = 5432
db: str = "llamastack"
user: str
password: str | None = None
@ -254,7 +254,30 @@ class ResponsesStoreReference(InferenceStoreReference):
"""Responses store configuration with queue tuning."""
class ServerStoresConfig(BaseModel):
metadata: KVStoreReference | None = Field(
default=None,
description="Metadata store configuration (uses KV backend)",
)
inference: InferenceStoreReference | None = Field(
default=None,
description="Inference store configuration (uses SQL backend)",
)
conversations: SqlStoreReference | None = Field(
default=None,
description="Conversations store configuration (uses SQL backend)",
)
responses: ResponsesStoreReference | None = Field(
default=None,
description="Responses store configuration (uses SQL backend)",
)
class StorageConfig(BaseModel):
backends: dict[str, StorageBackendConfig] = Field(
description="Named backend configurations (e.g., 'default', 'cache')",
)
stores: ServerStoresConfig = Field(
default_factory=lambda: ServerStoresConfig(),
description="Named references to storage backends used by the stack core",
)