GRPC server is down when running the model.ipynb

[geek-yang]

The GRPC server is down when running the Hype model (see the code and the error message below).

model.initialize(cfg_file)

---------------------------------------------------------------------------
_InactiveRpcError                         Traceback (most recent call last)
Cell In[8], line 1
----> 1 model.initialize(cfg_file)

File /opt/conda/envs/ewatercycle/lib/python3.10/site-packages/ewatercycle/base/model.py:168, in eWaterCycleModel.initialize(self, config_file)
    162 def initialize(self, config_file: str) -> None:
    163     """Initialize the model.
    164 
    165     Args:
    166         config_file: Name of initialization file.
    167     """
--> 168     self._bmi.initialize(config_file)

File /opt/conda/envs/ewatercycle/lib/python3.10/site-packages/grpc4bmi/bmi_optionaldest.py:49, in OptionalDestBmi.initialize(self, config_file)
     48 def initialize(self, config_file: Optional[str]) -> None:
---> 49     return self.origin.initialize(config_file)

File /opt/conda/envs/ewatercycle/lib/python3.10/site-packages/grpc4bmi/bmi_memoized.py:57, in MemoizedBmi.initialize(self, filename)
     55 def initialize(self, filename):
     56     self.cache = dict()
---> 57     return self.origin.initialize(filename)

File /opt/conda/envs/ewatercycle/lib/python3.10/site-packages/grpc4bmi/bmi_grpc_client.py:101, in BmiClient.initialize(self, filename)
     99     return self.stub.initialize(bmi_pb2.InitializeRequest(config_file=fname))
    100 except grpc.RpcError as e:
--> 101     handle_error(e)

File /opt/conda/envs/ewatercycle/lib/python3.10/site-packages/grpc4bmi/bmi_grpc_client.py:99, in BmiClient.initialize(self, filename)
     97 fname = "" if filename is None else filename
     98 try:
---> 99     return self.stub.initialize(bmi_pb2.InitializeRequest(config_file=fname))
    100 except grpc.RpcError as e:
    101     handle_error(e)

File /opt/conda/envs/ewatercycle/lib/python3.10/site-packages/grpc/_channel.py:1161, in _UnaryUnaryMultiCallable.__call__(self, request, timeout, metadata, credentials, wait_for_ready, compression)
   1146 def __call__(
   1147     self,
   1148     request: Any,
   (...)
   1153     compression: Optional[grpc.Compression] = None,
   1154 ) -> Any:
   1155     (
   1156         state,
   1157         call,
   1158     ) = self._blocking(
   1159         request, timeout, metadata, credentials, wait_for_ready, compression
   1160     )
-> 1161     return _end_unary_response_blocking(state, call, False, None)

File /opt/conda/envs/ewatercycle/lib/python3.10/site-packages/grpc/_channel.py:1004, in _end_unary_response_blocking(state, call, with_call, deadline)
   1002         return state.response
   1003 else:
-> 1004     raise _InactiveRpcError(state)

_InactiveRpcError: <_InactiveRpcError of RPC that terminated with:
	status = StatusCode.UNAVAILABLE
	details = "Socket closed"
	debug_error_string = "UNKNOWN:Error received from peer  {created_time:"2023-11-03T10:12:17.896994836+01:00", grpc_status:14, grpc_message:"Socket closed"}"
>

[BSchilperoort]

Seems like the image is broken. It has to be fixed. The Dockerfile is in https://github.com/eWaterCycle/hype-bmi/, but it's very rotten.

[BSchilperoort]

The apr2020 image does run, but I don't seem to be able to connect to it, even with the basic gprc4bmi functions

[BSchilperoort]

The apr2020 image does run, but I don't seem to be able to connect to it, even with the basic gprc4bmi functions

The reason why eWaterCycle doesn't connect, is that the ewatercycle/hype-grpc4bmi:apr2020 image uses the port 50051. eWC 2.0 uses 55555 by default.

When overwriting the _make_bmi_instance method, making use of the correct port, we can connect to the model successfully.
get_component_name() returns hype-5.6.1.

However, upon initialization with a config file the docker container fails with:

_InactiveRpcError: <_InactiveRpcError of RPC that terminated with:
	status = StatusCode.UNAVAILABLE
	details = "Socket closed"
	debug_error_string = "UNKNOWN:Error received from peer  {created_time:"2023-12-21T13:17:02.851491892+01:00", grpc_status:14, grpc_message:"Socket closed"}"
>

---

Perhaps this has to do with the following comment in the Dockerfiles:

# The input data root directory should be mounted as /data
# The filedir.txt in the input directory should only contain the single line: /data/

eWaterCycle 2.0 does not mount the input data root directory as /data.

[BSchilperoort]

The entire log of the container is:

BMI grpc server attached to server address 0.0.0.0:50051
 ---------------------------------------------
    HYPE version 5.6.2                        
 ---------------------------------------------
 New versions of HYPE at the HYPE Open Source 
 Community website (hypecode.smhi.se), as well
 as information on up-coming courses.         
 ---------------------------------------------
STOP 1

Not helpful.