BerriAI/litellm · error · Exception

Failed to poll operation: {response.text}

Error message

Failed to poll operation: {response.text}

What it means

Raised inside the long-running-operation polling loop when a GET on the Vertex AI operation returns a non-200 status. It signals a transient or auth failure while checking corpus-creation progress, not a completed-but-failed operation.

Source

Thrown at litellm/rag/ingestion/vertex_ai_ingestion.py:258

        url: Final = f"{base_url}/v1beta1/{operation_name}"

        client: Final = get_async_httpx_client(
            llm_provider=httpxSpecialProvider.RAG,
            params={"timeout": 60.0},
        )

        for attempt in range(max_retries):
            response = await client.get(
                url,
                headers={
                    "Authorization": f"Bearer {access_token}",
                },
            )

            if response.status_code != 200:
                error_msg = f"Failed to poll operation: {response.text}"
                verbose_logger.error(error_msg)
                raise Exception(error_msg)

            operation_data = response.json()

            if operation_data.get("done"):
                # Check for errors
                if "error" in operation_data:
                    error = operation_data["error"]
                    raise Exception(f"Operation failed: {error}")

                # Extract corpus name from response
                corpus_name = operation_data.get("response", {}).get("name", "")
                if corpus_name:
                    return corpus_name
                else:
                    raise Exception(f"No corpus name in operation response: {operation_data}")

            verbose_logger.debug("Operation not done yet, attempt %s/%s", attempt + 1, max_retries)
            await asyncio.sleep(retry_delay)

View on GitHub (pinned to 77b7c6c40c)

Solutions

1. Check the response text; verify credentials and that the operation name is valid, then retry polling.

When it happens

Trigger: Thrown at litellm/rag/ingestion/vertex_ai_ingestion.py:258 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of BerriAI/litellm@77b7c6c40c (2026-08-18). Data as JSON: /api/errors/c5e8ad9184ad3e6c. Report an issue: GitHub.