I'm seeing this error after upgrading livekit python sdk:

Task exception was never retrieved
future: <Task finished name='agent_job_termination' coro=<AgentServer._handle_termination() done, defined at /usr/local/lib/python3.10/site-packages/livekit/agents/worker.py:1256> exception=ValueError('process object is closed')>", "exc_info": "Traceback (most recent call last):
  File \"/usr/local/lib/python3.10/asyncio/tasks.py\", line 456, in wait_for
    return fut.result()
asyncio.exceptions.CancelledError

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File \"/usr/local/lib/python3.10/site-packages/livekit/agents/ipc/supervised_proc.py\", line 252, in aclose
    await asyncio.wait_for(
  File \"/usr/local/lib/python3.10/asyncio/tasks.py\", line 458, in wait_for
    raise exceptions.TimeoutError() from exc
asyncio.exceptions.TimeoutError

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File \"/usr/local/lib/python3.10/site-packages/livekit/agents/worker.py\", line 1261, in _handle_termination
    await proc.aclose()
  File \"/usr/local/lib/python3.10/site-packages/livekit/agents/ipc/supervised_proc.py\", line 261, in aclose
    await self._send_kill_signal()
  File \"/usr/local/lib/python3.10/site-packages/livekit/agents/ipc/supervised_proc.py\", line 317, in _send_kill_signal
    self._proc.kill()
  File \"/usr/local/lib/python3.10/multiprocessing/process.py\", line 139, in kill
    self._check_closed()
  File \"/usr/local/lib/python3.10/multiprocessing/process.py\", line 101, in _check_closed
    raise ValueError(\"process object is closed\")
ValueError: process object is closed

what is happening here?

Thanks for reaching out, Bharath. I'm an AI assistant and I'm here to help. Give me a moment while I research this topic.

when does this error usually occur? There is no referenece to my code so unabel to identify this

You're right that the traceback doesn't point to your app code — this error is produced inside the LiveKit agent's shutdown sequence. It usually happens in these situations: • A worker process already exited (crashed or completed) before the agent's shutdown logic reached it, and the agent later calls

.kill()

on that closed

multiprocessing.Process

. • A race during graceful termination: shutdown begins while workers are still finishing, so cleanup and a forced kill overlap. • A short

drain_timeout

or aggressive shutdown signal that causes the agent to try killing processes that have already been cleaned up. • A change in SDK shutdown behavior after an upgrade that exposes this timing window more often. Because this is an internal cleanup race, you won't see a reference to your application code in the traceback. To help pinpoint why it happened in your deployment: • Which LiveKit Python SDK version did you upgrade from and to? • Do you use custom job worker code or the standard job lifecycle? • What is your agent's

drain_timeout

(or shutdown/drain configuration)? • Did you see any worker crashes or other errors just before this traceback? Give me those details and I’ll suggest the most relevant next step.