all SubConns are in TransientFailure

[rhzs]

Hi guys,

I tried to run this repo examples folder but didn't work as expected.

Step to reproduce:

1. Install deps:

go get -u github.com/grpc-ecosystem/grpc-gateway/protoc-gen-grpc-gateway
go get -u github.com/grpc-ecosystem/grpc-gateway/protoc-gen-swagger
go get -u github.com/golang/protobuf/protoc-gen-go

2. cd $GOPATH/src/github.com/grpc-ecosystem/grpc-gateway/examples
3. go build && ./examples or go run main.go
4. In another terminal window:

➜ curl -X POST \
  http://localhost:8080/v1/example/echo_body \
  -H 'cache-control: no-cache' \
  -H 'content-type: application/json' \
  -d '{
        "id": "example"
}'
{"error":"all SubConns are in TransientFailure","code":14}%

I also got the same result with other endpoints POST http://localhost:8080/v1/example/echo/myid.

Any idea why I keep getting {"error":"all SubConns are in TransientFailure","code":14}%? How to make it work?

Thank you.

[rhzs]

Nvm, I haven't started GRPC server

[gusseleet]

Could you give some more information? How did you start it?

[philipithomas]

For anybody else getting this error - I have seen it on the client if the server panics and crashes. So, for a couple seconds, the server is down until kubernetes replaces the pod. It can be confusing to find the problem because you need the logs from the crashed container.

[Emixam23]

I have this issue on Mac OS X, any idea?

[johanbrandhorst]

This is just indicative of the client not being able to connect to the server. Check your setup.

[Emixam23]

Ok so I did check my setup and, I found this link on internet that, even if I have others problems now, helped me to get rid of this error : https://github.com/mycodesmells/golang-examples/blob/master/grpc/cmd/server/main.go

The thing I just don't get with this example is that, the guy initialize the grpc server, half in main, half in a go routine, then he serv the http part of its server and, the only difference I see is that he initialize a part of it before the go routine instead of initializing everything inside of this go routine (I hope I am clear, english isn't my first language sorry)

[vaikzs]

@johanbrandhorst I'm getting the same error now. I did no code change with the setup. Could you please help?

[johanbrandhorst]

@vaikzs I'm not sure what else I can add, this is an error returned from the gRPC client when it cannot connect to the server. Please ensure your gateway can connect to the gRPC backend.

[Emixam23]

@vaikzs Just as a side question, do you have any defer conn.close()? This might be where you issue comes from. Also, gRPC Gateway and gRPC server are two different instance (if I can tell so), make sure you're on different ports

[vaikzs]

@Emixam23 @johanbrandhorst Thanks for the responses. I guess my grpc server was not running for some reason. It seems to be fine now.

[lishuhao]

The README example just start Http server, not start rpc server
This can work

func run() error {
	ctx := context.Background()
	ctx, cancel := context.WithCancel(ctx)
	defer cancel()

	// Register gRPC server endpoint
	// Note: Make sure the gRPC server is running properly and accessible
	mux := runtime.NewServeMux()
	opts := []grpc.DialOption{grpc.WithInsecure()}
	err := gw.RegisterPushServiceHandlerFromEndpoint(ctx, mux, "localhost:8888", opts)
	if err != nil {
		return err
	}

	go func() {
		listener, err := net.Listen("tcp", conf.Config.Listen)
		if err != nil {
			log.Fatalln("listen failed:", err)
		}

		rpcServer := grpc.NewServer()

		gw.RegisterPushServiceServer(rpcServer, &service.PushServer{})
		if err := rpcServer.Serve(listener); err != nil {
			log.Fatalln("serve error:", err)
		}
	}()


	// Start HTTP server (and proxy calls to gRPC server endpoint)
	return http.ListenAndServe(":9999", mux)
}