feat: Application status page with healthcheck and custom metrics

[rafaelrsantosti]

Description

Create a public status page for each application showing real-time health status and optional custom metrics.

Features

1. Status Page per Application

Each application gets a dedicated status page accessible via:

• /status/{app-slug} (internal)
• status.{app-url} or {app-url}/status (public, optional)

2. Healthcheck Monitoring

Display real-time status based on Kubernetes healthcheck probes:

Status	Condition
🟢 Operational	All pods healthy, readiness/liveness passing
🟡 Degraded	Some pods unhealthy or restarting
🔴 Outage	All pods down or failing healthchecks
⚪ Maintenance	Manually set by user

3. Uptime History

• 24h/7d/30d/90d uptime percentage
• Incident timeline showing past outages
• Response time graph (if HTTP healthcheck)

4. Custom Metrics (Optional)

Users can enable custom metrics to display on status page:

# Example: prometheus metrics to expose
statusPage:
  enabled: true
  public: true  # accessible without auth
  metrics:
    - name: "API Response Time"
      query: "histogram_quantile(0.95, http_request_duration_seconds)"
      unit: "ms"
    - name: "Request Rate"
      query: "rate(http_requests_total[5m])"
      unit: "req/s"
    - name: "Error Rate"
      query: "rate(http_requests_total{status=~'5..'}[5m])"
      unit: "%"

5. Components Status

For applications with multiple components:

MyApp Status Page
─────────────────────────────
🟢 Web Frontend      Operational
🟢 API Backend       Operational  
🟡 Worker            Degraded (1/3 pods)
🟢 Database          Operational
─────────────────────────────
Overall: Partially Degraded
Uptime (30d): 99.7%

6. Incident Management

• Auto-detect incidents from healthcheck failures
• Manual incident creation for planned maintenance
• Incident updates with timeline
• Subscriber notifications (email, webhook)

Data Model

CREATE TABLE status_checks (
    uuid UUID PRIMARY KEY,
    component_uuid UUID REFERENCES application_components(uuid),
    status VARCHAR(20), -- operational, degraded, outage, maintenance
    response_time_ms INTEGER,
    checked_at TIMESTAMP,
    details JSONB
);

CREATE TABLE incidents (
    uuid UUID PRIMARY KEY,
    application_uuid UUID REFERENCES applications(uuid),
    title VARCHAR(255),
    status VARCHAR(20), -- investigating, identified, monitoring, resolved
    impact VARCHAR(20), -- none, minor, major, critical
    started_at TIMESTAMP,
    resolved_at TIMESTAMP
);

CREATE TABLE incident_updates (
    uuid UUID PRIMARY KEY,
    incident_uuid UUID REFERENCES incidents(uuid),
    message TEXT,
    status VARCHAR(20),
    created_at TIMESTAMP
);

API Endpoints

GET  /api/applications/{uuid}/status          # Current status
GET  /api/applications/{uuid}/status/history  # Uptime history
GET  /api/applications/{uuid}/incidents       # Incident list
POST /api/applications/{uuid}/incidents       # Create incident
PUT  /api/applications/{uuid}/incidents/{id}  # Update incident

# Public endpoints (no auth if public=true)
GET  /public/status/{app-slug}                # Public status page data

Portal UI

Status Page View

• Real-time status indicators
• Uptime chart (line graph)
• Response time chart
• Component breakdown
• Recent incidents list

Status Page Settings

• Enable/disable public access
• Configure custom domain
• Select metrics to display
• Notification settings

Technical Implementation

1. Status Collector Service

   • Polls Kubernetes API for pod health
   • Executes HTTP healthchecks
   • Stores results in database
   • Calculates uptime percentages

2. Metrics Integration

   • Query Prometheus for custom metrics
   • Cache results (30s-1min TTL)
   • Aggregate for display

3. Real-time Updates

   • WebSocket for live status updates
   • Or polling every 30 seconds

MVP Scope

• Basic status page with healthcheck status
• Uptime percentage (24h, 7d, 30d)
• Component-level status
• Public/private toggle

Future Enhancements

• Custom metrics from Prometheus
• Incident management
• Email/webhook notifications
• Custom branding
• Status page embedding (iframe/widget)
• API for external monitoring tools

References

• Atlassian Statuspage
• Instatus
• Cachet (open source)