langchain-ai/langchain · error · IndexingException

The delete operation to DocumentIndex failed.

Error message

The delete operation to DocumentIndex failed.

What it means

Raised by `_delete` in `langchain_core.indexing.api` when the destination is a `DocumentIndex` rather than a `VectorStore`. `DocumentIndex.delete(ids)` returns a dict response; if it contains `"num_failed"` greater than zero, some deletions failed on the index side and an `IndexingException` is raised during indexing cleanup.

Source

Thrown at libs/core/langchain_core/indexing/api.py:271

    Args:
        vector_store: The vector store or document index to delete from.
        ids: List of document IDs to delete.

    Raises:
        IndexingException: If the delete operation fails.
        TypeError: If the `vector_store` is neither a `VectorStore` nor a
            `DocumentIndex`.
    """
    if isinstance(vector_store, VectorStore):
        delete_ok = vector_store.delete(ids)
        if delete_ok is not None and delete_ok is False:
            msg = "The delete operation to VectorStore failed."
            raise IndexingException(msg)
    elif isinstance(vector_store, DocumentIndex):
        delete_response = vector_store.delete(ids)
        if "num_failed" in delete_response and delete_response["num_failed"] > 0:
            msg = "The delete operation to DocumentIndex failed."
            raise IndexingException(msg)
    else:
        msg = (  # type: ignore[unreachable]
            f"Vectorstore should be either a VectorStore or a DocumentIndex. "
            f"Got {type(vector_store)}."
        )
        raise TypeError(msg)


# PUBLIC API


class IndexingResult(TypedDict):
    """Return a detailed a breakdown of the result of the indexing operation."""

    num_added: int
    """Number of added documents."""
    num_updated: int
    """Number of updated documents because they were not up to date."""

View on GitHub (pinned to e32fa9a52e)

Solutions

1. Check the DocumentIndex backend logs/health for why deletes failed and retry the indexing run — indexing is resumable.
2. Verify connectivity and permissions to the DocumentIndex before running cleanup.
3. If failures are caused by already-missing IDs, clear the record manager state (`record_manager.delete_finished_records()` style reset) so stale IDs are not chased.

from langchain_core.indexing import IndexingException
import time

for attempt in range(3):
    try:
        index(doc_index, docs, rm, cleanup="full")
        break
    except IndexingException as e:
        if attempt == 2:
            raise
        time.sleep(2 ** attempt)  # backend transient delete failures often clear

When it happens

Trigger: Running `index()` with a DocumentIndex destination (e.g. a docstore retriever index) during full/incremental cleanup, where the backend reports partial delete failures via `num_failed > 0` in its response dict.

Common situations: DocumentIndex backends with eventual consistency where immediate deletes report failures; deleted/expired credentials mid-run; IDs already gone causing the backend to count them as failed.

Related errors

• The delete operation to VectorStore failed.
• source_id_key should be either None, a string or a callable.
• cleanup should be one of 'incremental', 'full', 'scoped_full
• Source id key is required when cleanup mode is incremental o
• Vectorstore has not implemented the delete method

AI-assisted analysis of langchain-ai/langchain@e32fa9a52e (2026-08-14). Data as JSON: /api/errors/c2da8e67db52eef8. Report an issue: GitHub.