Skip to content
This repository has been archived by the owner on Jun 11, 2020. It is now read-only.


Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?

Latest commit


Git stats


Failed to load latest commit information.
Latest commit message
Commit time

Status: Archived

status: inactive

This repository has been archived and is no longer maintained.

We recommend you use:

Unofficial Native Go Runtime for Google Cloud Functions

Disclaimer: This is not an official Google product. It is not and will not be maintained by Google, and is not part of Google Cloud Functions project. There is no guarantee of any kind, including that it will work or continue to work, or that it will supported in any way.


When this doc was first written, Google Cloud Functions only supported a Node.js runtime. Since then, new official runtimes have been added. Using an official runtime (such as the official Go runtime) is the recommended approach for almost all users.

But if you're feeling adventurous, you can run non-JavaScript code by using native Node modules and running subprocesses. Node will always be handling the requests. Wrapping Go code in a native Node module will require writing complicated foreign function interfaces (FFIs). FFIs are notoriously difficult to get right and are an easy way to introduce memory leaks as well as security and stability issues. Using a subprocess is much easier to get right, but it introduces the overhead of interprocess communication (IPC) and context switches. In both cases, the requests are proxied between Node and Go which typically involves extra copies and translations.

100% Pure Go Request Path

After an initial bootstrap step, the only thing running is a pure Go binary. This project works by completely replacing the Node process and lets a Go process directly handle requests. There is no cgo and no proxying. This means that it is faster and potentially more secure than other projects which typically to use both.

How It Works

A native module calls execve(2) without the normal preceding fork(2) call when the user module is imported. Open socket FDs are preserved and handled by the new Go process to ensure a smooth transition. The new Go process then pretends to be Node.


There are four supported environments:


For advanced users who don't like VirtualBox.

  • Linux or macOS development environment (The native Node module depends on Linux/macOS specific behavior.)
  • Go 1.5 or above
  • Node.js v0.10 or above, npm and node-gyp
  • Make and GCC
  • If you are using macOS then as of right now you need the FULL xcode installed. Command line tools alone will not work.


Compatible with Windows, macOS and Linux and easier than native development.

In the cloud-functions-go directory, run vagrant up to start the environment and vagrant ssh to get a shell. Support for other Vagrant providers besides VirtualBox is in the works.

Windows with Cygwin

  • Go for Windows
  • Node.js for Windows
  • Cygwin with make and zip (you may also want git and an editor like vim or nano)

Use the Cygwin Terminal to run the commands as described below. Note that make test won't work under Windows.

Windows with Powershell 5.0

  • Go 1.5 or above
  • Node.js (optional)

The commands described below may be run as-is using Command Prompt, or prefixed with ./ using Windows PowerShell (i.e. ./make or ./make godev). Note that make test won't work under Windows.

External Dependencies

The events sub-package depends on the following libraries:


Hello, world!

A demo hello world example is included. To try it out, simply skip to the Deployment section.

Making Changes

For normal development, it should only be necessary to modify the main.go file. Although this file probably shouldn't live in your GOPATH, feel free to import libraries from your GOPATH (including libraries downloaded with go get).

Local Testing

Run make test to compile your code and start the test server. Open http://localhost:8080/execute in your browser. The page should display User function is ready. Refresh the page to talk to your code.

When using a go-only or non-unix environment, run make godev instead.


The logger may be used directly:

nodego.InfoLogger.Println("Hello World!")
nodego.ErrorLogger.Println("Something went wrong!")

or via the log package after calling nodego.OverrideLogger():

func init() {


log.Println("Hello World!")

Use the nodego.WithLogger() or nodego.WithLoggerFunc() middleware for your logs to reference the specific execution:

http.WithLoggerFunc(func(w http.ResponseWriter, r *http.Request) {
	log.Println("Hello World!")

A full example is included in examples/logging.go.


Run make to compile and package your code. Upload the generated file to Google Cloud Functions as an HTTP trigger function.


Run vagrant up to start the envirement. Run vagrant ssh to connect to the envirement. Run cd /vagrant to access the respority files. The instructions in Local Testing and Deployment should now work.


Some versions of Node.js (especially those packaged for Ubuntu) name their main binary nodejs instead of node. The symptom of this problem is an error about the node binary not being found in the path even though Node.js is installed. This can be fixed with sudo ln -s $(which nodejs) /usr/local/bin/node. There's also a package called nodejs-legacy that can be installed in some Debian and Ubuntu distros using apt that creates a symlink node in /usr/bin/

It may also be helpful to understand the nodejs environment that your Google Cloud Function runs in (e.g. how does regular console.log send logs to Stackdriver, whereas fmt.Println doesn't?). You can find more details in the file.


Copyright 2017, Google, Inc.

Licensed under the Apache License, Version 2.0