Hey everyone! I’m having trouble setting up N8N with Qdrant. I’m trying to follow the first workflow from their blog post, but I keep getting this error:
Problem in node 'Qdrant Vector Store'
400 Bad Request: Wrong input: Not existing vector name error:
I’m not sure what I’m doing wrong. Does anyone have experience with this? I’ve set up my Qdrant collection, but maybe I missed something? Here’s what I did:
- Created a collection in the Qdrant dashboard
- Set up the N8N workflow using GitHub Document Loader, OpenAI Embeddings, and Qdrant Vector Store nodes
- Tried executing the workflow, but got the error
Is there a missing step or a specific way to configure the Qdrant collection for N8N? Any tips or working examples would be super helpful!
If anyone has a sample workflow file they could share, that would be awesome. I’ve searched online but couldn’t find much information on this setup.
Thanks in advance for your help!
I’ve actually faced a similar issue when integrating N8N with Qdrant. The error you’re getting suggests that the vector name specified in your N8N workflow doesn’t match what’s in your Qdrant collection.
Here’s what worked for me:
Double-check your Qdrant collection schema to ensure the vector field name exactly matches what you’re using in N8N. In the Qdrant Vector Store node, verify that the ‘Vector Name’ field is set correctly. If you’re using default settings, try explicitly using ‘vector’ in both Qdrant and N8N. Also, confirm your Qdrant connection details, including URL and API key if applicable. Remember, Qdrant is case-sensitive.
If these steps don’t resolve the issue, consider creating a simple test workflow with just the Qdrant Vector Store node to isolate the problem.
Hope this helps troubleshoot the integration.
hey, i had similar probs with n8n and qdrant. make sure ur vector name in n8n matches exactly whats in qdrant (case matters!). also, check if ur collection is set up right - sometimes it exists but doesnt have the right schema. u could try using qdrant’s api directly to test stuff out before going back to n8n. good luck!
I’ve encountered this exact problem before. The key is in the vector name configuration. Make sure the vector name in your N8N Qdrant Vector Store node matches the one in your Qdrant collection exactly. It’s case-sensitive, so ‘vector’ and ‘Vector’ are different.
Also, check if you’ve initialized the collection properly. Sometimes, the collection exists but hasn’t been set up with the correct schema. Try using the Qdrant API or CLI to verify the collection’s structure.
One thing that helped me was to use the Qdrant HTTP API node in N8N instead of the Vector Store node initially. This allowed me to debug the requests more easily. Once I got that working, switching to the Vector Store node was smoother.
Lastly, ensure your Qdrant instance is accessible from where N8N is running. Network issues can sometimes manifest as cryptic errors.