Are they up?

Real-time service status pages powered by user outage reports and baseline-aware scoring.

Live site: aretheyup.com

What this project does

aretheyup helps answer a simple question: is a service down for everyone, or just me?

It combines:

• Crowdsourced user reports ("I am having issues")
• Historical baselines by time bucket
• Optional probe-failure signal (when probe data exists)

The result is a clear status for each service:

• Operational
• Issues Detected

Key features

• SEO-friendly status pages (/, /:slug)
• Fast, server-rendered UI with HTMX
• Report endpoint with abuse protection via rate limiting
• Background workers for baseline refresh and incident tracking
• Service management TUI for admins (manage mode)

Tech stack

• Go 1.25+
• Gin + GORM
• PostgreSQL + Redis
• HTMX + Tailwind + DaisyUI

Quick start (local development)

1) Prerequisites

• Go 1.25+
• Docker + Docker Compose

2) Clone and configure

git clone https://github.com/novembersoftware/aretheyup.git
cd aretheyup
cp .env.example .env.local

Update values in .env.local as needed.

3) Start local dependencies

docker compose -f docker-compose.dev.yml up -d

4) Run the app

go run main.go

Server runs on http://localhost:8080 by default.

Run modes

The binary supports multiple modes:

# API server (default)
go run main.go
# equivalent:
go run main.go api

# Manage mode (service admin TUI)
go run main.go manage

# Seed mode
go run main.go seed --count 25 --clear

Seed flags:

• --count number of services to seed (default 10)
• --clear clear existing data before seeding

Environment variables

See .env.example for full list. Most important values:

• ENV (dev or prod)
• API_PORT
• DB_DSN
• REDIS_URL
• SITE_BASE_URL
• ALLOWED_PAGE_ORIGINS (comma-separated)
• TRUSTED_PROXIES (comma-separated CIDRs/IPs, optional)
• REPORT_RATE_LIMIT_MAX_REQUESTS
• REPORT_RATE_LIMIT_WINDOW_SECONDS

Status algorithm (high level)

Status is calculated from a rolling 30-minute report window and compared to historical norms.

• Baselines are stored per service and hour-of-week bucket (0..167, UTC)
• Cold start behavior uses conservative hard thresholds
• Mature behavior uses z-score style anomaly checks
• Probe failures (if available) can independently trigger Issues Detected

Reference implementation:

• algorithm/status.go
• workers/baseline.go
• workers/incidents.go

SEO behavior

• GET /robots.txt allows public crawling and disallows /api/
• GET /sitemap.xml includes / and active /:slug pages
• Absolute sitemap URLs are derived from SITE_BASE_URL

Security notes

• Set TRUSTED_PROXIES to only your real edge/ingress proxy CIDRs
• Keep SITE_BASE_URL set in production
• /api/* routes require allowed website origin/referer checks

Stopping local services

docker compose -f docker-compose.dev.yml down

Contributing

Issues and PRs are welcome. Please open an issue before you start working on larger changes.

License

MIT. See LICENSE.