BerriAI/litellm · error · HTTPException
Unable to validate vector store access
Error message
Unable to validate vector store access
What it means
Raised when resolving a vector-store id from the shared cache (e.g. Redis) threw an unexpected exception: the registry lookup failed, then the cache fallback also errored, so access cannot be verified and the endpoint fails closed with 500 rather than guessing. The original exception is chained and logged as a warning.
Source
Thrown at litellm/proxy/vector_store_endpoints/utils.py:213
LiteLLM-managed multi-tenant stores.
"""
if not vector_store_id:
return None
if litellm.vector_store_registry is not None:
try:
vector_store: Final = litellm.vector_store_registry.get_litellm_managed_vector_store_from_registry(
vector_store_id=vector_store_id
)
if vector_store is not None:
return _normalize_litellm_params(vector_store)
except Exception as e:
verbose_proxy_logger.warning(
"Failed to resolve vector store id=%s from registry: %s",
vector_store_id,
e,
)
raise HTTPException(
status_code=500,
detail="Unable to validate vector store access",
) from e
try:
from litellm.proxy.auth.auth_checks import (
get_managed_vector_store_rows_by_uuids,
)
from litellm.proxy.proxy_server import (
prisma_client,
proxy_logging_obj,
user_api_key_cache,
)
if prisma_client is None:
return None
rows: Final = await get_managed_vector_store_rows_by_uuids(
uuids=[vector_store_id],View on GitHub (pinned to 77b7c6c40c)
Solutions
- Ensure the vector store exists and the database is connected so access can be validated; check logs for details.
Defensive patterns
Strategy: try-catch
When it happens
Trigger: Thrown at litellm/proxy/vector_store_endpoints/utils.py:213 when the library encounters an invalid state.
Common situations: See trigger scenarios.
AI-assisted analysis of BerriAI/litellm@77b7c6c40c (2026-08-18).
Data as JSON: /api/errors/09c6b9abc361e32b.
Report an issue: GitHub.