[Bug] Service API cannot re-enable a segment with Elasticsearch after disabling it (regression of #10445) #39988

Self Checks

• I have read the Contributing Guide and Language Policy.
• This is only for bug report, if you would like to ask a question, please head to Discussions.
• I have searched for existing issues search for existing issues, including closed ones.
• I confirm that I am using English to submit this report, otherwise it will be closed.
• 【中文用户 & Non English User】请使用英语提交，否则会被关闭 ：）
• Please do not modify this template :) and fill in all the required fields.

Dify version

1.16.0

Cloud or Self Hosted

Self Hosted (Docker)

Steps to reproduce

1. Configure a self-hosted Dify instance to use Elasticsearch as the vector store.
2. Create a high-quality knowledge base with a paragraph-indexed document and wait until its segments are indexed successfully.
3. Disable an enabled segment through the Knowledge Service API:

POST /v1/datasets/{dataset_id}/documents/{document_id}/segments/{segment_id}
Authorization: Bearer <dataset-api-key>
Content-Type: application/json

{"segment":{"enabled":false}}

4. Wait for disable_segment_from_index_task to finish. The vector is removed successfully from Elasticsearch.
5. Re-enable the same segment through the Knowledge Service API without changing its content:

POST /v1/datasets/{dataset_id}/documents/{document_id}/segments/{segment_id}
Authorization: Bearer <dataset-api-key>
Content-Type: application/json

{"segment":{"enabled":true}}

6. The API responds with HTTP 200, but the returned/persisted segment is disabled and has status=error.

The API/worker logs show the following sanitized call chain:

update segment index failed
DatasetService.update_segment(...)
VectorService.update_segment_vector(...)
vector.delete_by_ids([segment.index_node_id])
ElasticsearchVector.delete_by_ids(...)
elasticsearch.NotFoundError: NotFoundError(404, 'not_found',
  '{"_index":"Vector_index_<redacted>_Node",
    "_id":"<index_node_id>","result":"not_found"}')

Re-enabling the same segment from the Dify Web console succeeds. The Web console uses update_segments_status() and enable_segments_to_index_task, which reloads the vector instead of first deleting the already-missing vector.

✔️ Expected Behavior

Re-enabling a segment through the Knowledge Service API should recreate/reload the segment vector and leave the segment with enabled=true and a non-error status.

Re-enabling should also be idempotent when the vector has already been removed by the preceding disable task. A missing vector during this flow should not abort re-indexing.

❌ Actual Behavior

In Dify 1.16.0, the unchanged-content branch of DatasetService.update_segment() calls VectorService.update_segment_vector() when args.enabled is true:

https://github.com/langgenius/dify/blob/1.16.0/api/services/dataset_service.py#L3599-L3602

update_segment_vector() unconditionally deletes the existing vector before adding it:

https://github.com/langgenius/dify/blob/1.16.0/api/services/vector_service.py#L131-L135

The Elasticsearch adapter calls client.delete() without ignoring a missing-document 404:

https://github.com/langgenius/dify/blob/1.16.0/api/providers/vdb/vdb-elasticsearch/src/dify_vdb_elasticsearch/elasticsearch_vector.py#L182-L186

Because the disable task has already removed the vector, the second delete returns 404 not_found. The exception prevents add_texts() from running. DatasetService.update_segment() then catches the exception, sets enabled=false and status=error, commits that state, and returns normally:

https://github.com/langgenius/dify/blob/1.16.0/api/services/dataset_service.py#L3787-L3797

This appears to be a regression of #10373 / #10445. PR #10445 handled re-enabling unchanged content by calling VectorService.create_segments_vector() directly. The Dify 1.16.0 path now calls update_segment_vector() instead, reintroducing the failure for Elasticsearch.

• Original issue: segment can not be used when enable by api after disable #10373
• Original fix: fix segment enable service api #10445