Quartify Deployment

Version française

Helm/Kubernetes configuration to deploy the Quartify web application on SSP Cloud (Onyxia).

Description

This repository contains the deployment configuration for Quartify, an R package that automatically converts R scripts into Quarto markdown documents (.qmd).

Docker Image: ddottaagr/quartify:latest

Application Features

The Quartify web application offers:

• Interactive Shiny interface (EN/FR)
• R Quarto (.qmd) conversion
• HTML rendering with Quarto (Quarto is installed in the Docker image)
• Support for callouts, tabsets, Mermaid diagrams
• Source line numbers for traceability
• Customizable Quarto themes (25+ themes)
• Download generated .qmd and .html files

Deployment on SSP Cloud

Prerequisites

• Access to SSP Cloud
• Active user account

Option 1: Via Web Interface (Recommended)

1. Log in to https://datalab.sspcloud.fr/
2. Go to My Lab Service Catalog
3. Search for "Shiny" or create a custom service
4. In the configuration:
   • Repository: https://github.com/ddotta/quartify_deployment
   • Chart: . (root of the repository)
   • Values file: values-sspcloud.yaml
5. Click Launch
6. Once the service is started, click on the provided URL

Option 2: Via Helm CLI

# Clone the repository
git clone https://github.com/ddotta/quartify_deployment.git
cd quartify_deployment

# Install with Helm
helm install quartify . -f values-sspcloud.yaml

# Get the service URL
kubectl get ingress

Option 3: Via kubectl

# Apply manifests directly
kubectl apply -f templates/

Configuration

Main Files

• Chart.yaml: Helm chart metadata
• values.yaml: Default configuration
• values-sspcloud.yaml: SSP Cloud specific configuration
• templates/: Kubernetes manifests (Deployment, Service, Ingress)

Customization

To modify the configuration, edit values-sspcloud.yaml:

image:
  repository: ddottaagr/quartify
  tag: latest
  pullPolicy: Always

resources:
  limits:
    memory: "2Gi"
    cpu: "2000m"
  requests:
    memory: "1Gi"
    cpu: "1000m"

service:
  type: ClusterIP
  port: 3838

ingress:
  enabled: true
  hosts:
    - host: quartify.lab.sspcloud.fr

Repository Structure

quartify_deployment/
 Chart.yaml                 # Helm metadata
 values.yaml                # Default configuration
 values-sspcloud.yaml       # SSP Cloud configuration
 templates/
    deployment.yaml        # Kubernetes Deployment
    service.yaml          # Kubernetes Service
    ingress.yaml          # Ingress for exposure
    _helpers.tpl          # Helm template helpers
 README.md                  # This file
 README_FR.md              # French version

Docker Image

The ddottaagr/quartify Docker image contains:

• R 4.4.1 (rocker/shiny)
• Shiny Server
• Quarto 1.4.549 (enables HTML rendering)
• quartify package (installed from GitHub)
• All required R dependencies

Building the Image

The image is automatically built via GitHub Actions on every push to the main branch of the quartify repository.

To build manually:

# Clone the main repository
git clone https://github.com/ddotta/quartify.git
cd quartify

# Build the image
docker build -t ddottaagr/quartify:latest .

# Test locally
docker run -p 3838:3838 ddottaagr/quartify:latest

Troubleshooting

Service Won't Start

# Check pod logs
kubectl get pods
kubectl logs <pod-name>

# Check events
kubectl describe pod <pod-name>

Resource Issues

If the pod is in OOMKilled state (out of memory), increase limits in values-sspcloud.yaml:

resources:
  limits:
    memory: "4Gi"  # Increase to 4GB

Application Not Responding

Verify the port is correct:

# Service should be on port 3838
kubectl get service quartify

Updates

To update the application with a new image version:

# Update the deployment
helm upgrade quartify . -f values-sspcloud.yaml

# Or force restart
kubectl rollout restart deployment/quartify

Useful Links

• Main Repository: https://github.com/ddotta/quartify
• Docker Image: https://hub.docker.com/r/ddottaagr/quartify
• Documentation: https://ddotta.github.io/quartify/
• SSP Cloud: https://datalab.sspcloud.fr/

Support

For any questions or issues:

• Open an issue on GitHub
• Consult the complete documentation

License

This deployment project follows the same license as the main Quartify project (MIT).