This page summarizes some errors that may occur when using the Pinecone Vector Store Provider.
Executing any operation against the Vector Store the following error occurs:
PineconeAuthorizationError: The API key you provided was rejected while calling https://api.pinecone.io/indexes/.
Please check your configuration values and try again.
You can find the configuration values for your project in the Pinecone developer console at https://app.pinecone.io
Check your Pinecone configuration credentials, in particular the following container values:
PINECONE_API_KEY
PINECONE_INDEX_NAME
This information must be correctly set on the associated deployed containers.
Check connectivity and credentials with the following cURL sample, using the correct values for IndexName and api_key:
curl --location 'https://api.pinecone.io/indexes/IndexName' \
--header 'Api-Key: api_key' \
--header 'Content-Type: application/json' \
--data ''
RAG Assistants configured with an embedding dimension value other than 768, 1024, 1536 cannot be deleted. This limitation arises due to chanking dimensions needs to create a new index. Attempting to remove such an assistant results in a server error caused by the inability to locate a matching index with the expected configuration.
Error message:
Error: StatusCode:500 Http:0 Internal Server Error Error:100 Failed to get Pinecone index:
PineconeNotFoundError: A call to https://api.pinecone.io/indexes/indexnameDimensions returned HTTP status 404.
Contact us to review the creation of an specific index for your purposes. If using an on-premises installation check the PINECONE_CREATE_INDEX environment variable configuration.
To delete the assistant configure dimensions to 1536 and delete it.