ERROR logs on pipeline shutdown when using Python Redpanda Connect plugin

Hi! I’m experimenting with the new Redpanda Connect Plugins framework. During pipeline shutdown, I see the following ERROR logs:

INFO[2026-01-23T07:44:07Z] Running main config from specified file       @service=redpanda-connect benthos_version=4.78.0 path=connect.yaml
INFO[2026-01-23T07:44:07Z] Successfully loaded Redpanda license          @service=redpanda-connect expires_at="2036-01-21T07:44:07Z" license_org="" license_type=open_source
INFO[2026-01-23T07:44:07Z] Listening for HTTP requests at: http://0.0.0.0:4195  @service=redpanda-connect
INFO[2026-01-23T07:44:08Z] Launching a Redpanda Connect instance, use CTRL+C to close  @service=redpanda-connect
INFO[2026-01-23T07:44:08Z] Output type file is now active                @service=redpanda-connect label="" path=root.output
INFO[2026-01-23T07:44:08Z] Input type generate is now active             @service=redpanda-connect label="" path=root.input
INFO[2026-01-23T07:44:09Z] ERROR:grpc._cython.cygrpc:Exception not handled by _handle_exceptions in servicer method [/redpanda.runtime.v1alpha1.BatchProcessorService/Close]  @service=redpanda-connect label="" path=root.pipeline.processors.0 source=stderr
INFO[2026-01-23T07:44:09Z] Traceback (most recent call last):            @service=redpanda-connect label="" path=root.pipeline.processors.0 source=stderr
INFO[2026-01-23T07:44:09Z]   File "src/python/grpcio/grpc/_cython/_cygrpc/aio/server.pyx.pxi", line 787, in grpc._cython.cygrpc._schedule_rpc_coro  @service=redpanda-connect label="" path=root.pipeline.processors.0 source=stderr
INFO[2026-01-23T07:44:09Z] asyncio.exceptions.CancelledError             @service=redpanda-connect label="" path=root.pipeline.processors.0 source=stderr
INFO[2026-01-23T07:44:09Z] Traceback (most recent call last):            @service=redpanda-connect label="" path=root.pipeline.processors.0 source=stderr
INFO[2026-01-23T07:44:09Z]   File "src/python/grpcio/grpc/_cython/_cygrpc/aio/server.pyx.pxi", line 787, in grpc._cython.cygrpc._schedule_rpc_coro  @service=redpanda-connect label="" path=root.pipeline.processors.0 source=stderr
INFO[2026-01-23T07:44:09Z] asyncio.exceptions.CancelledError             @service=redpanda-connect label="" path=root.pipeline.processors.0 source=stderr
INFO[2026-01-23T07:44:09Z] Successfully loaded Redpanda Connect RPC plugin  @service=redpanda-connect label="" path=root.pipeline.processors.0 source=stdout
INFO[2026-01-23T07:44:09Z] Pipeline has terminated. Shutting down the service  @service=redpanda-connect

The pipeline itself works as expected. Is this expected behavior, and is there a way to avoid these ERROR logs during shutdown?

Here’s the setup:

connect.yaml

input:
  generate:
    mapping: |
      root = {
        "message": fake("sentence")
      }
    interval: ""
    count: 1000

pipeline:
  processors:
    - yell: {}

output:
  file:
    codec: lines
    path: ./processed/output.jsonl

plugin.yaml

name: yell
summary: yells back what you give it
command: ["uv", "run", "--no-sync", "--frozen", "main.py"]
type: processor
fields: []

main.py

import asyncio
import logging
from datetime import datetime, timezone
from redpanda_connect import Message, processor, processor_main

logging.getLogger("grpc").setLevel(logging.ERROR)
logging.getLogger("grpc._cython.cygrpc").setLevel(logging.ERROR)
logging.getLogger("redpanda_connect._grpc").setLevel(logging.ERROR)

logger = logging.getLogger(__name__)


@processor
def yell(msg: Message) -> Message:
    try:
        data = msg.payload
        data["message"] = data["message"].upper()
        data["yelled_at"] = datetime.now(timezone.utc)
        return msg
    except Exception as err:
        logger.error(f"Error processing message: {err}", exc_info=True)
        return msg


if __name__ == "__main__":
    asyncio.run(processor_main(yell))

Reproduce the shutdown behavior with connect.yaml, plugin.yaml, and main.py, using the provided pipeline and Python processor. Start by tracing processor_main(yell) and the gRPC BatchProcessorService Close path; done means determining whether the CancelledError logs are expected and, if not, preventing them during normal shutdown.