Public service status page implementation

Implementing Status Page (Public Service Status Page)

Status Page is public page where users see current service state and incident history. Without it, during outage support gets flooded with "it's broken" tickets, Twitter gets angry posts, team spends time on communication instead of recovery.

What Status Page Should Contain

Current component status. Each key component displayed separately:

• Main application / website
• API
• Authentication
• Payment system
• CDN / file storage
• Email sends

Statuses: Operational, Degraded Performance, Partial Outage, Major Outage, Under Maintenance.

Availability history. Uptime graph over last 90 days — classic "green squares" by day.

Active incidents. If something broken — incident with update chronology.

Planned maintenance. Advance notice of maintenance window.

Update subscriptions. Email, Slack, webhook — user picks channel.

Implementation Options

Statuspage.io (Atlassian) — market leader. Integration with PagerDuty, Datadog, Jira. Automatic status update by metrics. Public and private pages. From $79/month.

Cachet — open source, self-hosted. PHP + MySQL/PostgreSQL. Full control, no monthly fee. API integration. Good if custom domain and branding needed.

Instatus — modern alternative. Faster UI, competitive pricing.

Self-built on static site — minimalist: GitHub Pages + YAML status file + GitHub Actions updates page. Good for small projects.

Automatic Status Updates

Status Page without automation is page forgotten during incident (when update matters most).

Datadog → Statuspage.io Integration:

import requests

def update_status_page(component_id: str, status: str):
    # status: operational | degraded_performance | partial_outage | major_outage
    requests.patch(
        f"https://api.statuspage.io/v1/pages/{PAGE_ID}/components/{component_id}",
        headers={"Authorization": f"OAuth {API_KEY}"},
        json={"component": {"status": status}}
    )

# Called from Datadog webhook on alert trigger
def on_alert_triggered(alert_data):
    component = map_alert_to_component(alert_data["alert_title"])
    update_status_page(component["id"], "major_outage")

Prometheus Alertmanager → webhook → update script → Status Page API.

Rule: status update should happen automatically within 2-3 minutes of problem detection, not 20 minutes later when someone remembers.

Incident Workflow on Status Page

Incident structure:

1. Investigating — problem detected, investigating
2. Identified — cause found, working on fix
3. Monitoring — fix applied, monitoring stabilization
4. Resolved — incident closed

Updates every 20-30 minutes during active incident. Users know they're not forgotten.

Technical Details of Self-Hosted Cachet

# docker-compose.yml
services:
  cachet:
    image: cachethq/docker:latest
    environment:
      APP_ENV: production
      APP_KEY: ${APP_KEY}
      DB_DRIVER: pgsql
      DB_HOST: db
      DB_DATABASE: cachet
    ports:
      - "80:8000"
  db:
    image: postgres:15
    environment:
      POSTGRES_DB: cachet
      POSTGRES_PASSWORD: ${DB_PASSWORD}

Nginx reverse proxy with SSL (Let's Encrypt). Custom domain like status.yourdomain.com.

Implementation Timeline

• Statuspage.io / Instatus (managed) — 1-2 days setup
• Cachet self-hosted — 2-3 days
• Automatic update from monitoring — 1-2 days
• Integration with incident management — 1 day