googleCloudRunner

As easy as possible R scripts in the cloud, via Cloud Run, Cloud Build and Cloud Scheduler. Continuous Development and Integration tools on Google Cloud Platform.

Not an official Google product.

Ambition

Select an R file, and have it scheduled in the cloud with a couple of clicks.

Deploy your plumber API code automatically on Cloud Run to scale from 0 (no cost) to millions (auto-scaling)

Integrate R inputs and outputs with other languages in a serverless cloud environment.

Have R code react to events such as GitHub pushes, pub/sub messages and Cloud Storage file events.

Install

Get the CRAN stable version via

install.packages("googleCloudRunner")

Or the development version via:

remotes::install_github("MarkEdmondson1234/googleCloudRunner")

The dev version of the website is here which may include documentation of features not yet on CRAN.

Usage

Browse the googleCloudRunner website for topics on how to use:

• Setup
• R APIs using Cloud Run
• Serverless R scripts using Cloud Build
• Scheduled R in GCP using Cloud Scheduler
• Use Cases
• Function Reference

Diagrams

Some overviews of the function dependencies are below:

Scheduling using Cloud Scheduler

Building using Cloud Build

Helper deployment functions calling lower-level functions

Cloud Run functionality

library(targets) integration functions

Sponsor

Resources

Read the blog introducing the googleCloudRunner package that goes into some background.

A talk at R's 20th anniversary was performed at celebRation in Copenhagen on 29th Feb, 2020. The slides from the talk are here and a video of it is here:

A googleCloudRunner YouTube playlist of the demos is here.

If you blog anything interesting about the package let me know and I will list it here.

• Ander has a guide on using googleCloudRunner to schedule an R script on GCP [Spanish]
• Ander also writes how to use googleCloudRunner to productionise your R plumber API [Spanish]
• Arben documents his experience on how he got started scheduling BigQuery uploads using Docker and R
• Michał Ludwicki was a huge help in mentoring Arben for the post above - he also has created a GitHub repo of some useful example files for googleCloudRunner scripts
• Sam Terfa has a guide on how to use googleCloudRunner to create an R API you can use GoogleSheets to create its front end
• Louis shows how to create and deploy your own Python optimisation API using googleCloudRunner