Default execution for single_file.py`` fails when connecting to nats

[piotrm-nvidia]

The single_file.py fails with default execution:

root@ulalegionbuntu:/workspace/examples/hello_world# python3 single_file.py 
Starting Workers
20:23:42 deployment.py:115[triton_distributed.worker.deployment] INFO: 

Starting Worker:

	Config:
	WorkerConfig(request_plane=<class 'triton_distributed.icp.nats_request_plane.NatsRequestPlane'>,
             data_plane=<function UcpDataPlane at 0x714b1c759080>,
             request_plane_args=(['nats://localhost:4223'], {}),
             data_plane_args=([], {}),
             log_level=1,
             operators=[OperatorConfig(name='encoder',
                                       implementation=<class 'triton_distributed.worker.triton_core_operator.TritonCoreOperator'>,
                                       repository='/workspace/examples/hello_world/operators/triton_core_models',
                                       version=1,
                                       max_inflight_requests=1,
                                       parameters={'config': {'instance_group': [{'count': 1,
                                                                                  'kind': 'KIND_CPU'}],
                                                              'parameters': {'delay': {'string_value': '0'}}}},
                                       log_level=None)],
             triton_log_path=None,
             name='encoder.0',
             log_dir='/workspace/examples/hello_world/logs',
             metrics_port=50000)
	<SpawnProcess name='encoder.0' parent=130 initial>

20:23:42 deployment.py:115[triton_distributed.worker.deployment] INFO: 

Starting Worker:

	Config:
	WorkerConfig(request_plane=<class 'triton_distributed.icp.nats_request_plane.NatsRequestPlane'>,
             data_plane=<function UcpDataPlane at 0x714b1c759080>,
             request_plane_args=(['nats://localhost:4223'], {}),
             data_plane_args=([], {}),
             log_level=1,
             operators=[OperatorConfig(name='decoder',
                                       implementation=<class 'triton_distributed.worker.triton_core_operator.TritonCoreOperator'>,
                                       repository='/workspace/examples/hello_world/operators/triton_core_models',
                                       version=1,
                                       max_inflight_requests=1,
                                       parameters={'config': {'instance_group': [{'count': 1,
                                                                                  'kind': 'KIND_GPU'}],
                                                              'parameters': {'delay': {'string_value': '0'}}}},
                                       log_level=None)],
             triton_log_path=None,
             name='decoder.0',
             log_dir='/workspace/examples/hello_world/logs',
             metrics_port=50001)
	<SpawnProcess name='decoder.0' parent=130 initial>

20:23:42 deployment.py:115[triton_distributed.worker.deployment] INFO: 

Starting Worker:

	Config:
	WorkerConfig(request_plane=<class 'triton_distributed.icp.nats_request_plane.NatsRequestPlane'>,
             data_plane=<function UcpDataPlane at 0x714b1c759080>,
             request_plane_args=(['nats://localhost:4223'], {}),
             data_plane_args=([], {}),
             log_level=1,
             operators=[OperatorConfig(name='encoder_decoder',
                                       implementation=<class '__main__.EncodeDecodeOperator'>,
                                       repository=None,
                                       version=1,
                                       max_inflight_requests=100,
                                       parameters={},
                                       log_level=None)],
             triton_log_path=None,
             name='encoder_decoder.0',
             log_dir='/workspace/examples/hello_world/logs',
             metrics_port=50002)
	<SpawnProcess name='encoder_decoder.0' parent=130 initial>

Sending Requests
nats: encountered error
Traceback (most recent call last):
  File "/usr/local/lib/python3.12/dist-packages/nats/aio/client.py", line 1353, in _select_next_server
    await self._transport.connect(
  File "/usr/local/lib/python3.12/dist-packages/nats/aio/transport.py", line 121, in connect
    r, w = await asyncio.wait_for(
           ^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/lib/python3.12/asyncio/tasks.py", line 520, in wait_for
    return await fut
           ^^^^^^^^^
  File "/usr/lib/python3.12/asyncio/streams.py", line 48, in open_connection
    transport, _ = await loop.create_connection(
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/lib/python3.12/asyncio/base_events.py", line 1122, in create_connection
    raise exceptions[0]
  File "/usr/lib/python3.12/asyncio/base_events.py", line 1104, in create_connection
    sock = await self._connect_sock(
           ^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/lib/python3.12/asyncio/base_events.py", line 1007, in _connect_sock
    await self.sock_connect(sock, address)
  File "/usr/lib/python3.12/asyncio/selector_events.py", line 651, in sock_connect
    return await fut
           ^^^^^^^^^
  File "/usr/lib/python3.12/asyncio/selector_events.py", line 691, in _sock_connect_cb
    raise OSError(err, f'Connect call failed {address}')
ConnectionRefusedError: [Errno 111] Connect call failed ('127.0.0.1', 4223)
Sending Requests:   0%|                                                                                                                | 0/100 [00:00<?, ?request/s]

[piotrm-nvidia]

#34

The control plane implementation doesn't support other address for NATS.io than localhost. My machine is configured in such a way that default bind for 0.0.0.0 used in NATS.io is not binding to 127.0.0.1 but to real WiFi port. I can't pass host to control plane so I can't run example.

[nnshah1]

@piotrm-nvidia - is there a way we can detect the binding from the NatsServer class directly and then use that to encapsulate the uri?

For the single_file example - would like to see if we can avoid any arguments .... maybe we could use environment variables - but goal for single file is to avoid any customization outside of the file - maybe we can add an arg in the file itself - or in the NatsServer class to toggle between localhost and others ...