Containerizing Shiny Applications with shiny2docker
In shiny2docker: Generate Dockerfiles for 'Shiny' Applications

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

Introduction

shiny2docker is an R package designed to streamline the process of containerizing Shiny applications using Docker. This vignette demonstrates how to generate a Dockerfile for your Shiny application, customize it, and configure GitLab CI to build and deploy your Docker image.

Generating a Dockerfile

The main function, shiny2docker(), automates the creation of a Dockerfile. It does the following:

Utilizes a renv.lock file to capture your R package dependencies.
Generates a .dockerignore file to exclude unnecessary files, reducing build time and image size.
Creates a Dockerfile that includes instructions to install system dependencies, R packages, and launch your Shiny app.

To generate a Dockerfile in your current directory, simply run:

library(shiny2docker)

# Generate a Dockerfile in the current directory
shiny2docker(path = ".")

Alternatively, you can specify custom paths for your Shiny application, the renv.lock file, and the output Dockerfile:

library(shiny2docker)

# Generate a Dockerfile with specific paths
shiny2docker(
  path = "path/to/your/app",
  lockfile = "path/to/your/app/renv.lock",
  output = "path/to/your/app/Dockerfile"
)

Customizing the Dockerfile

The shiny2docker() function returns an object of class dockerfiler. This object can be further manipulated to add custom instructions before writing the Dockerfile to disk. For example, you may want to set an environment variable:

# Create the dockerfiler object
docker_obj <- shiny2docker()

# Add environment variables
docker_obj$add_after(  cmd = "ENV ARROW_WITH_S3=1",  after = 1)
docker_obj$add_after(  cmd = "ENV ARROW_S3=ON",after = 1)


# Write the updated Dockerfile to disk
docker_obj$write("Dockerfile")

Configuring CI/CD for Docker Builds

GitLab CI for Docker Builds

The package provides the set_gitlab_ci() function to simplify the process of configuring a GitLab CI pipeline. This pipeline is designed to build your Docker image and push it to the GitLab container registry.

# Copy the .gitlab-ci.yml file to the current directory
set_gitlab_ci(path = ".")

Once the .gitlab-ci.yml file is in place, you can integrate it with GitLab CI/CD to automate your Docker image build and deployment process.

GitHub Actions for Docker Builds

The package also provides the set_github_action() function, which simplifies the process of configuring a GitHub Actions pipeline. This pipeline is designed to build your Docker image and push the created image to the GitHub Container Registry.

# Copy the docker-build.yml file to the .github/workflows/ directory
set_github_action(path = ".")

Once the docker-build.yml file is in place, you can integrate it with GitHub Actions to automate the Docker image build and deployment process, allowing you to continuously build and deploy your containerized Shiny applications.

Conclusion

shiny2docker simplifies containerizing Shiny applications by automating Dockerfile creation and CI configuration. By integrating with renv for dependency management and providing a customizable dockerfiler object, it offers a flexible workflow for deploying Shiny apps in containerized environments.

For further details, please refer to the package documentation and visit the dockerfiler GitHub repository.