[BUG] Upserting to Vector Store fails in Document Sore - Request to Ollama server failed

[paintbot]

When im upserting two PDFs (655 and 371 chunks; 9105590 and 479470 chars) to a Faiss Vector Store with Ollama Embeddings I get the following error:

I used text splitter with 2000 chunk size and 200 overlap) and updated Flowise and Ollama

Im using Ollama Embedding with nomic-embed-text embedding model

The same process works before with smaller PDFs

Oh snap! The following error occured when loading this page. Status: 500 Error: documentStoreServices.insertIntoVectorStore - Error: documentStoreServices._insertIntoVectorStoreWorkerThread - Error: Request to Ollama server failed: 404 Not Found Please retry after some time. If the issue persists, reach out to us on our Discord server. Alternatively, you can raise an issue on Github.

[paintbot]

Also have the same problem when trying to insert from the Chat Flow instead of using Document Store.

Are the Ollama Embeddings the problem? Can I use other embeddings with ollama LLMs as well?

[banatus]

Hi,I was encountering the same things after upsert at Flowise even I was using Ubuntu 24.04 and having /home/myname/vector path" 0 Error: documentStoreServices.insertIntoVectorStore - Error: documentStoreServices._insertIntoVectorStoreWorkerThread - Error: Request to Ollama server failed: 404 Not Found". Banatus Soiraya

[OpeyemiSanusi]

having the same issues running on a mac any solution?

Error: documentStoreServices.insertIntoVectorStore - Error: documentStoreServices._insertIntoVectorStoreWorkerThread - Error: Request to Ollama server failed: 404 Not Found

[paintbot]

I found that it works with smaller amounts of text. I guess rather Faiss or Ollama hit some capacity limit. Did anyone try a different vector store that is also able to run locally like ChromaDB? Or is the amount of data you can have in a vector store also somehow limited by flowise?