CODEWITHSUNDEEP
dockertensorflowgrpctensorflow-servinggrpc-python
Mattia
Mattia

Reputation: 21

gRPC server responds with OS Error, grpc_status: 14

Using a basic gRPC client from the Tensorflow Serving examples to get predictions from a model running on docker I get this response:

grpc._channel._Rendezvous: <_Rendezvous of RPC that terminated with:
        status = StatusCode.UNAVAILABLE
        details = "OS Error"
        debug_error_string = "{"created":"@1580748231.250387313",
            "description":"Error received from peer",
            "file":"src/core/lib/surface/call.cc",
            "file_line":1017,"grpc_message":"OS Error","grpc_status":14}"

This is what my client currently looks like:

import grpc
import tensorflow as tf
import cv2

from tensorflow_serving.apis import predict_pb2
from tensorflow_serving.apis import prediction_service_pb2_grpc


def main():
    data = cv2.imread('/home/matt/Downloads/cat.jpg')

    channel = grpc.insecure_channel('localhost:8500')
    stub = prediction_service_pb2_grpc.PredictionServiceStub(channel)
    
    request = predict_pb2.PredictRequest()
    request.model_spec.name = 'model'
    request.model_spec.signature_name = 'serving_default'
 
    request.inputs['image_bytes'].CopyFrom(
        tf.make_tensor_proto(data, shape=[1, data.size]))
    result = stub.Predict(request, 10.0)  # 10 secs timeout
    print(result)

if __name__ == '__main__':
    main()

Thanks in advance for any help :)

Upvotes: 2

Views: 10550

Answers (1)

user11530462
user11530462

Reputation:

Providing the Solution here even though it is present in Comments Section for the benefit of the community.

The solution is that we need to Invoke the Tensorflow Model Server by running the Docker Container using the code given below, before executing the Client File:

docker run -t --rm -p 8501:8501 \
    -v "$TESTDATA/saved_model_half_plus_two_cpu:/models/half_plus_two" \
    -e MODEL_NAME=half_plus_two \
    tensorflow/serving &

In addition to invoking the Tensorflow Model Server,

  1. It will Map the local path of the Model with the path of the Model on the Server and
  2. It will map the Ports for communicating with the Tensorflow Model Server. (Port 8500 is exposed for gRPC and Port 8501 is exposed for REST API)

Upvotes: 0

Related Questions