Skip to content
Ishvana Ishvana Docs

Reindex Documents

POST
/api/discovery/enhanced-search/reindex-documents

Reindex all documents with enhanced embeddings in background task.

Request Body:

  • model_name: Embedding model reported in the response (default: “multilingual-e5-large-instruct”)
  • force: Force reindexing even if embeddings exist (default: False)

Query Parameters:

  • force_reindex: Override force flag via query parameter
  • batch_size: Batch size for reindexing operations (1-1000, default: 100)

Returns:

  • status: “started”, “skipped”, or “completed”
  • message: Operation status message
  • document_count: Number of documents to be indexed
  • model: Embedding model being used

Raises:

  • 500: Failed to retrieve documents or start reindexing

Note: Long-running operation executed in background task.

Parameters

Section titled “ Parameters ”

Query Parameters

Section titled “Query Parameters ”
force_reindex
Force Reindex
boolean
batch_size
Batch Size
integer
default: 100

Request Body required

Section titled “Request Body required ”
ReindexRequest
object
modelName
Modelname

Embedding model identifier reported back in the response. Reindexing always runs against the bundled E5 ONNX runtime.

string
default: multilingual-e5-large-instruct
force
Force

Force reindexing even if embeddings exist

boolean

Responses

Section titled “ Responses ”

200

Section titled “200 ”

Successful Response

SearchReindexResponse

Search reindex response model.

Provides confirmation and status for document reindexing operations including operation status, document count, and embedding model information for enhanced semantic search.

Fields:

  • status: Reindex operation status (“started”, “skipped”, “completed”, “error”)
  • message: Human-readable status message describing the operation
  • document_count: Number of documents to be indexed or indexed (non-negative integer, optional)
  • model: Embedding model name used for reindexing (optional)

Status Values:

  • started: Reindexing operation initiated in background
  • skipped: Reindexing skipped (embeddings already exist, use force=true)
  • completed: Reindexing completed successfully
  • error: Reindexing encountered an error

Usage: POST /api/discovery/enhanced-search/reindex-documents returns this response model.

JSON Example:

{
  "status": "started",
  "message": "Reindexing 150 documents in background",
  "documentCount": 150,
  "model": "multilingual-e5-large-instruct"
}
object
status
required
Status

Reindex operation status

string
message
required
Message

Status message

string
documentCount
Any of:
integer
model
Any of:
string

422

Section titled “422 ”

Validation Error

HTTPValidationError
object
detail
Detail
Array<object>
ValidationError
object
loc
required
Location
Array
msg
required
Message
string
type
required
Error Type
string
input
Input
ctx
Context
object