v0.15.3,When upload documents to the knowledge base through the document api create_by_file, there may be situations where the documents are available without being indexed

[ccooder]

Self Checks

• [x] This is only for bug report, if you would like to ask a question, please head to Discussions.
• [x] I have searched for existing issues search for existing issues, including closed ones.
• [x] I confirm that I am using English to submit this report (我已阅读并同意 Language Policy).
• [x] [FOR CHINESE USERS] 请务必使用英文提交 Issue，否则会被关闭。谢谢！:）
• [x] Please do not modify this template :) and fill in all the required fields.

Dify version

0.15.3

Cloud or Self Hosted

Self Hosted (Docker)

Steps to reproduce

Just use document's api create_by_file At first, it's worked, but then the situation shown in the title will appear

✔️ Expected Behavior

The uploaded file will Indexing and avavilable

❌ Actual Behavior

Although the status is available, the file does not have any segmentation because it becomes available immediately after the interface is called

[dosubot[bot]]

I found two similar unsolved discussions that might be relevant to your issue:

1. Knowledge base file upload text segmentation and cleaning error ^[1].
2. Unable to automatically perform embedding processing after uploading documents using the API? ^[2].

Regarding your issue, there are known limitations with the create_by_file API in version 0.15.3 that might be related to your problem. The API supports only single file uploads, and files must be 15MB or smaller. Additionally, the self-hosted Docker version does not support PPTX files, even though the official API does ^[3].

The Dify platform determines when a document is fully indexed by checking the document's indexing status. You can verify the indexing status using the following API call:

curl --location --request GET 'https://api.dify.ai/v1/datasets/{dataset_id}/documents/{batch}/indexing-status' \
--header 'Authorization: Bearer {api_key}'

This will provide the current status of the document indexing process. The possible indexing_status values include waiting, parsing, cleaning, splitting, indexing, paused, error, and completed ^[4][5].

If a document is marked as available but not indexed, it might be due to an error in the indexing process. Specific error messages such as document_indexing, unsupported_file_type, or file_too_large can indicate issues ^[4][6].

You might want to check the indexing status and error messages to diagnose the issue further.

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Bug Report | Other}

[ccooder]

Does a self-hosted service also have limitations?

[crazywoola]

Does a self-hosted service also have limitations?

No, there are some bugs in this process.

[disTance7777]

You can use the browser console to grab the parameters when you upload files using webui, and then write your api request body Just like the request body I wrote has parameters that are not declared in the interface document You will find that the mode parameter is inconsistent with what is declared in the interface specification

{
  "indexing_technique": "high_quality",
  "doc_form": "hierarchical_model",
  "process_rule": {
    "mode": "hierarchical",
    "rules": {
      "parent_mode": "full-doc",
      "pre_processing_rules": [
        {
          "id": "remove_extra_spaces",
          "enabled": true
        },
        {
          "id": "remove_urls_emails",
          "enabled": false
        }
      ],
      "segmentation": {
        "separator": "\n\n",
        "max_tokens": 500
      },
      "subchunk_segmentation": {
        "separator": "\n",
        "max_tokens": 4000
      }
    }
  }
}