The purpose of this article is very simple. Just run a gRPC example in different Docker containers. Since it really costed me a lot of time, I’d like to keep records here.
Docker is a very useful container tool, which can be considered as a light virtualization technique comparing to traditional virtual machine tools such as Virtualbox, VMware, etc. For more details about the differences between them, as well as the Docker’s basics, please refer this Stack Overflow question, the Digital Ocean tutorial, or the Docker official introduction.
The official tutorial is good enough for Ubuntu. This alternative installation tutorial from Digital Ocean tutorials also introduces some basic commands and tips, such as executing the docker command without sudo. I tested them at an Ubuntu 16.04 x64 server, and everything went well.
Install commands for lazy guys who also use Ubuntu x64 release:
sudo apt-get update sudo apt-get install \ apt-transport-https \ ca-certificates \ curl \ software-properties-common curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo apt-key add - sudo add-apt-repository \ "deb [arch=amd64] https://download.docker.com/linux/ubuntu \ $(lsb_release -cs) \ stable" sudo apt-get update sudo apt-get install docker-ce
Install gRPC at local host OS
Actually this step is not needed if you want to do all the work inside containers. But with local installed package, we can test host-to-container communications through gRPC API.
Here, I explicitly use python3.
python3 -m pip install grpcio
pip3 install grpcio
Get grpc/python Docker image
Here since I did not build my own Docker image but directly use the official one.
To get latest Python 3 environment with gRPC, run
docker pull grpc/python:1.4
Note here “1.4” tag is necessary, otherwise it will pull “latest” one, which is Python 2.7 environment.
Get gRPC examples
Follow the gRPC official guide, we can get the example source code.
# Clone the repository to get the example code: git clone -b v1.6.x https://github.com/grpc/grpc # Navigate to the "hello, world" Python example: cd grpc/examples/python/helloworld
Run in containers
Run the RPC server in docker container
docker run -itd --rm --name my-running-server \ -v "$PWD":/usr/src/myapp -w /usr/src/myapp \ grpc/python:1.4 python3 greeter_server.py
Here “d” in
-itd means “detach”. So we can let the container there and run other commands.
We can use
docker attach my-running-server to see what is going on for server container, and
CTRL-q to detach again.
Then, we can check docker default network using
docker network inspect bridge
From the output, the server container should have the IP “18.104.22.168” at network “bridge”.
Small modify from “localhost” to “22.214.171.124” is required.
def run(): # channel = grpc.insecure_channel('localhost:50051') channel = grpc.insecure_channel('126.96.36.199:50051') ...
In same terminal (because I was using ssh connection to a server), run the client
docker run -it --rm --name my-running-client \ -v "$PWD":/usr/src/myapp -w /usr/src/myapp \ grpc/python:1.4 python3 greeter_client.py
We can get correct output.
Also we can run client locally.
The output should be same.