Skip to content

DEPLOYMENT.md

Latest commit

 

History

History
130 lines (93 loc) · 2.64 KB

DEPLOYMENT.md

File metadata and controls

130 lines (93 loc) · 2.64 KB

Deployment Guide

This document describes the current NUSMods deployment model.

Overview

NUSMods is deployed as two parts:

  1. Website + export endpoints on Vercel.
  2. Self-managed data pipeline (scraper + static API data hosting, plus search infra).

Website and Export (Vercel)

The website (nusmods.com) and export service are deployed via Vercel.

  • Website config: website/vercel.json
  • Export config: export/vercel.json

Operational flow:

  1. Push a branch / open a PR to get a preview deployment.
  2. Merge to the production branch configured in Vercel.
  3. Validate critical flows after deploy:
    • module search
    • timetable rendering
    • timetable export (PNG/PDF)
    • error reporting

For local production-like verification:

cd website
pnpm build
npx serve -s dist

Self-Managed Data/API Infrastructure

Use this section if you operate scraper/data infrastructure outside Vercel.

Recommended baseline

  • Ubuntu LTS
  • Node 22 LTS
  • pnpm
  • PM2
  • Nginx (for static API data serving)

Server bootstrap (Ubuntu)

sudo apt update && sudo apt upgrade -y
sudo apt install -y git nginx curl build-essential

Install Node 22 LTS and pnpm using your standard team method (nvm, fnm, or distro policy), then verify:

node -v
pnpm -v

Install PM2 globally:

pnpm add -g pm2
pm2 -v

Repository setup

git clone https://github.com/nusmodifications/nusmods.git
cd nusmods
pnpm install

Scraper setup and runtime

In scrapers/nus-v2:

cd scrapers/nus-v2
cp env.example.json env.json
# fill in env.json with required internal API credentials / endpoints
pnpm dev test | pnpm bunyan
pnpm build
pm2 start ecosystem.config.js
pm2 save
pm2 startup systemd

Notes:

  • Scraper production scripts are defined in scrapers/nus-v2/package.json.
  • PM2 config is in scrapers/nus-v2/ecosystem.config.js.
  • ElasticSearch config is optional for local development, but required for production search.

Static API data serving (Nginx)

The API static JSON data directory is typically served as api.nusmods.com.

  1. Place generated data under your chosen path (for example, /home/<user>/api.nusmods.com).
  2. Configure an Nginx server block to serve that directory.
  3. Enable and restart Nginx:
sudo systemctl enable nginx
sudo systemctl restart nginx
  1. Validate permissions for all parent directories and files.
  2. Point DNS to this host and verify public access.

Updating production data pipeline

cd /path/to/nusmods
git pull
pnpm install
cd scrapers/nus-v2
pnpm build
pm2 restart ecosystem.config.js

Run a scraper health check after restart:

pnpm dev test | pnpm bunyan