Gopher Tunnels

A local tunnel built ~~with~~by Gopher.

If ngrok is blocked in your network, this is an alternative one that you can deploy to your own AWS account in a few steps. This application runs on serverelss. I mean all servers will be managed by Amazon 100%. You pay nothing if there is no traffic coming to your service. However, Amazon may charge for configurations.

This project uses serverless framework. So, make sure that you're familiar with that framework a bit before giving it a go.

Requirements

You may want to install this library to your $GOPATH

This is optional and these libraries can be installed to vendor directory locally.

go get -u github.com/aws/aws-sdk-go
go get -u github.com/pkg/errors
go get -u github.com/rs/xid
go get -u golang.org/x/crypto/nacl/box

What are included in this project?

Server (Lambda functions)

Binary path: bin/server

gopher_register - the client program (gopher) has to call this function when it starts.
gopher_webhook - this function handle your webhook. The endpoint will be provided to you when you run the client.
gopher_respond - this function handle the response returning from the client.
gopher_cleanup - work in progress

Client

Binary path: bin/client

gopher - the client that you can distribute to your team or anyone you want.

Tools

Binary path: bin/tools

config - this tool will generate a client configuration file from application.yml.
echo_server - this tool is really useful for debugging an HTTP call.
It is an http server that returns exactly what it receives like an echo service. The nice part of this tool is that it dumps the request to the console as well. So, you can inspect the HTTP request such as path, method, headers, and body.

How to configure the project

Please copy application.template.yml to application.yml Edit that file to make sure that all attributes are correct. Note that you don't know baseApiEndpoint until you deploy the API it to API Gateway. This attribute can be left empty when you deploy your serverless because it's required by the client not the server.

How to build the project

Make sure that everything is installed to vendor directory
$ dep ensure

Compile the project
$ make

How to deploy the project

Deploying a server

You can deploy simply by running a simple command like the following. Thanks to serverless framework.
sls deploy

You can override some attributes such as stage and region via command line.
sls deploy --stage prod --region us-west-1

You can deploy individual function by providing the function name sls deploy function --function gopher_register

Packaging a client

Assume that your current directory is at the project root. Otherwise, you will get an error.
If you want to create a client for the default stage (dev), you simply run ./dist.sh.
dist.sh will create the dist directory, generate a client configuration from your application.yml, and copy all related binary files and drop them there. You can just zip this directory and distribute it.

However, you can specify the stage name by providing the stage name as the first argument. I'm a bash script noob and I don't care to improve it. So, contribution is very welcome :)

Note that gopher will read the configuration named gopher.yml if you don't provide any argument. Please see the usage section for more details.

Usage

You can override host, mode, and port by providing the arguments for those attributes. Otherwise, the values from a configuration file will be used.

$ ./gopher -h
Usage of ./gopher:
  -host string
    	target host
  -mode string
    	webhook mode: sync, async
  -port string
    	target port

If your configuration file is not gopher.yml, you have to provide the configuration file as the last argument as the following. $ ./gopher gopher.dev.yml

TailrecIO / gopher-tunnels