Azure ML

[JamboChen]

Describe the bug

ToolLoadError: Failed to load tool 'Vector Index Lookup' for node 'Vector_Index_Lookup_lomr' due to 'internal server error'.

What happened?

The following error occurs when running the Vector Index Lookup node

Stack trace

Error: Request failed with status code 400
    at Fz (https://ml.azure.com/assets/index-80b513df.js:261:13689)
    at Iz (https://ml.azure.com/assets/index-80b513df.js:261:13877)
    at XMLHttpRequest.E (https://ml.azure.com/assets/index-80b513df.js:262:1672)

Client request ID: f3d3b402-d476-4d83-87c3-3ee2cbf0c838
Trace ID: 723e3ce4-c011-4051-b4a9-88e8641d4c7d

What did you expect to happen?

The node can complete

How would you reproduce the bug?

1. Click Create under the vector index tab of the Prompt flow
2. Upload files, choose Faiss Index, then click Create
3. Clone Q&A with your own data using Faiss Index
4. Copy the Datastore URI of vector index, and paste it on the Vector Index Lookup node of Flows

Did this happen via Desktop and/or Smartphone?

Desktop

Operating System

No response

What browsers are you seeing the problem on?

Microsoft Edge

Version

115.0.1901.188

Device

No response

What is your current role?

Student Ambassador

What technical topic is this bug related to?

AI/Data Science/Machine Learning

Relevant log output

{
  "error": {
    "code": "UserError",
    "message": "Failed to load tool 'Vector Index Lookup' for node 'Vector_Index_Lookup_lomr' due to 'internal server error'.",
    "innerError": {
      "code": "ValidationError",
      "innerError": {
        "code": "ToolValidationError",
        "innerError": {
          "code": "ToolLoadError"
        }
      }
    },
    "additionalInfo": null,
    "details": []
  },
  "correlation": {
    "operation": "80fc27b75bea4b16a36e5fbea81f1af5",
    "request": "957e4bfe2f86c7a8"
  }
}

[github-actions]

Thank you for submitting this issue! The team will review your issue, tag with the appropriate tags, and comment with any additional questions on information needed. ✨

[Briwalkr]

I am getting the same error on an environment that I setup yesterday. New connection to AOAI, Local Dataset(AML storage), new compute/prompt flow runtime setup (promptflow-runtime:20230801.v1), new faiss index created this morning.

Using the sample flow that was created as an output of the faiss index creation, I get the following error.

Failed to load tool 'Vector Search' for node 'search_question_from_indexed_docs' due to 'internal server error'.

Retesting - pointing the path directly to blob index/ source and then again directly pointing at the MLIndex file, and both get the following error (with a different /tmp/runSpecificTempDir/MLIndex target). It appears the file is not being moved to the expected location for the Vector Search tool.

ToolLoadError: Failed to load tool 'Vector Search' for node 'search_question_from_indexed_docs' due to '[Errno 2] No such file or directory: '/tmp/tmpq2ysf5uj/MLIndex''.