How to run Python based grpc Service in production?

[vinitra...@gmail.com]

Hi There,

I am newbie to grpc services, and written couple of service in python-grpc. I was wondering how should I deploy thin on production, basically question is more towards is there a way to host both rest and grpc service in the same runtime may be using gunicorn or uwsgi. Or if the grpc.Server can be used in production as well and I should only use supervisor kind of thing to ensure its running state.

[Ken Payson]

On Tuesday, May 22, 2018 at 9:30:44 PM UTC-7, vinitra...@gmail.com wrote:

Hi There,

I am newbie to grpc services, and written couple of service in python-grpc. 

I was wondering how should I deploy thin on production, basically question is more towards is there a way to host both rest and grpc service in the same runtime may be using gunicorn or uwsgi.

There isn't a good way to serve both REST/gRPC on the same port, but there isn't anything preventing you from running a gRPC server side-by-side with a standard HTTP1.1 server on another port. 

Or if the grpc.Server can be used in production as well and I should only use supervisor kind of thing to ensure its running state.

The canonical approach to gRPC health checking is described here:

https://github.com/grpc/grpc/blob/master/doc/health-checking.md 

gRPC provides a Python implementation of the health checking service here

https://pypi.org/project/grpcio-health-checking/