BerriAI/litellm · error · RuntimeError

A2A send_message failed: no response received after retry…

Error message

A2A send_message failed: no response received after retry attempts.

What it means

_execute_a2a_send_with_retry loops the non-streaming send (retrying once after a localhost-card rewrite). If the loop ends without ever capturing a successful response — retries exhausted or every attempt failing in a way that produced no response object — this RuntimeError is raised. It is the aggregate 'attempted but never succeeded' error for message/send.

Solutions

Verify the agent is up and the URL is right: curl <api_base>/.well-known/agent-card.json
Check logs for a preceding 'Retrying with base URL' warning — if present, fix the agent's advertised card URL
Wrap asend_message with your own retry/backoff for transient outages
Inspect the original underlying exception in the traceback (the retry wrapper re-raises it) to classify network vs agent error

Defensive patterns

Strategy: retry

Validate before calling

import httpx

async def agent_ready(api_base: str) -> bool:
    try:
        async with httpx.AsyncClient(timeout=5) as hc:
            return hc and (await hc.get(f'{api_base.rstrip("/")}/.well-known/agent-card.json')).is_success
    except httpx.HTTPError:
        return False

if not await agent_ready(api_base):
    raise RuntimeError(f'A2A agent not reachable at {api_base}')

Try / catch

last = None
for attempt in range(3):
    try:
        resp = await asend_message(a2a_client=client, request=req, api_base=api_base)
        break
    except RuntimeError as e:
        if 'after retry attempts' not in str(e):
            raise
        last = e
        await asyncio.sleep(2 ** attempt)
else:
    raise RuntimeError(f'A2A agent unreachable after backoff: {api_base}') from last

Prevention

Validate api_base with a card fetch at startup and on config reload
Circuit-break agents that repeatedly fail instead of hard-failing user requests
Watch for the localhost-retry warning in logs; it usually precedes this error

When it happens

Trigger: api_base unreachable (agent down, wrong host/port) so every attempt errors; a localhost-card retry succeeded logistically but the resent request still failed; retry budget consumed by transient network errors.

Common situations: Pointing at a stopped agent container; wrong port in api_base; flaky network between litellm and the agent; agent restarting mid-request.

Related errors

AI-assisted analysis of BerriAI/litellm@77b7c6c40c (2026-08-18). Data as JSON: /api/errors/3393829dfb8150e7. Report an issue: GitHub.

Appendix: source

Thrown at litellm/a2a_protocol/main.py:311

                is_streaming=False,
            )
            card_url = get_agent_card_url(agent_card) if agent_card else None
        except Exception as e:
            try:
                map_a2a_exception(e, card_url, api_base, model=agent_name)
            except A2ALocalhostURLError as localhost_err:
                a2a_client = await handle_a2a_localhost_retry(
                    error=localhost_err,
                    agent_card=agent_card,
                    a2a_client=a2a_client,
                    is_streaming=False,
                )
                card_url = get_agent_card_url(agent_card) if agent_card else None
                continue
            except Exception:
                raise
    if a2a_response is None:
        raise RuntimeError("A2A send_message failed: no response received after retry attempts.")
    return a2a_response


async def _stream_messages(
    a2a_client: "A2AClientType", request: "SendStreamingMessageRequest"
) -> AsyncIterator["SendStreamingMessageResponse"]:
    """Stream message events via a2a-sdk 1.x and yield JSON-RPC chunks."""
    if _a2a_conversions is None:
        raise ImportError(
            "The 'a2a' package is required for A2A agent invocation. Install it with: pip install a2a-sdk"
        )

    pb_request: Final[CoreSendMessageRequest] = _a2a_conversions.to_core_send_message_request(request)
    async for event in a2a_client.send_message(pb_request, context=_get_a2a_call_context(a2a_client)):
        compat_chunk = _to_compat_stream_response(event, request_id=request.id)
        yield SendStreamingMessageResponse(root=compat_chunk)

View on GitHub (pinned to 77b7c6c40c)