When inserting embeddings into Pinecone, the following error occurred to me.
PineconeError: index is full, cannot accept data
What does that mean? Plus, could someone tell me how to deal with the snag?
your indexes are limited by the pod types and number of them. The space you are given on any tier is not limitless, so do check what is your pod_fullnes (either in Dashboard or via code).
For reference: Indexes
- Each s1 pod has enough capacity for around 5M vectors of 768 dimensions.
- Each p1 pod has enough capacity for around 1M vectors of 768 dimensions.
- Each p2 pod has enough capacity for around 1M vectors of 768 dimensions. However, capacity may vary with dimensionality.
- When using the starter plan, you can create one pod with enough resources to support approximately 100,000 vectors with 1536-dimensional embeddings and metadata; the capacity is proportional for other dimensions.
So depending on which type of pods you are using, you might have reached that pod limit.
A solution: increase the number of pods you are using from x1 to x2/x4/x8 if you are on a paid tier. Otherwise there is not much you can do on Free tier 
Hope this helps
I see where you’re coming from. I appreciate your reply.
Can this error (coming back from the server) include which index is full? We encountered this from a shared application that uses several different Pinecone indexes across several different projects.
From just looking at the exception thrown by the Pinecone TS client in our Bugsnag logs, it’s not possible to tell which index is the one that is full.