-
Type:
Bug
-
Resolution: Done
-
Priority:
Major - P3
-
None
-
Affects Version/s: None
-
Component/s: None
-
2
-
Not Needed
-
🔵 Done
-
Not Needed
-
None
-
None
-
None
-
None
-
None
-
None
It was discovered that vector search indexes cannot be created in n8n despite it being supported. This is likely due to a missing "type" field in the implementation. Slack thread: https://mongodb.slack.com/archives/C05M73LQ5TN/p1758556145827119
Flow in n8n to reproduce
- Create new workflow and trigger
- Add MongoDB node to create search index
- Toggle to vector search index
- Specify index definition and execute node/workflow
- Error message: "attribute mappings missing"
Comment from the slack thread:
Not sure how the innards of n8n work too well, but looking at the function signature here, the missing part may be the "type" field. (I always confuse myself on unpacking in js, so I could be wrong)
When not supplied for vectorSearch definitions, it will assume you are attempting to create a $search index.Checked the node documentation to confirm vectorSearch index creation in supported (edited)Â
Documentation Requirements
- Please create DOCSP ticket to update docs to update any examples
- related to
-
NODE-6645 Implement MongoDB Vector Search Node in n8n.io
-
- Execution Blocked
-
- links to