# Observability & Monitoring

The Observability feature provides comprehensive monitoring, metrics collection, and alerting for the Changemaker Lite platform. Built on the Prometheus ecosystem with Grafana dashboards and Alertmanager integration.

## Overview

The Observability stack consists of:

1. **Prometheus** - Metrics collection and storage
2. **Grafana** - Visualization dashboards
3. **Alertmanager** - Alert routing and notifications
4. **Custom Metrics** - 12 domain-specific `cm_*` metrics
5. **HTTP Metrics** - Request tracking and performance
6. **Service Health** - External service monitoring

## Features

### Metrics Collection

**Custom Domain Metrics (12 total):**

**Counters:**
- `cm_api_uptime_seconds` - API uptime counter
- `cm_canvass_visits_total` - Total canvass visits
- `cm_campaign_emails_sent_total` - Total campaign emails sent
- `cm_geocode_requests_total` - Total geocode requests

**Gauges:**
- `cm_canvass_sessions_active` - Active canvass sessions
- `cm_email_queue_size` - Email queue depth
- `cm_geocode_queue_size` - Geocode queue depth
- `cm_external_service_health` - Service health (0/1)

**Histograms:**
- `cm_geocode_duration_seconds` - Geocoding latency
- `http_request_duration_ms` - HTTP request duration

**HTTP Metrics:**
- Request count by method/route/status
- Request duration percentiles (p50, p95, p99)
- Active requests gauge
- Error rate tracking

### Grafana Dashboards

Three pre-configured dashboards:

1. **Changemaker Lite Overview** - System-wide metrics
   - API uptime and request rates
   - Queue sizes and health
   - Active sessions
   - Error rates

2. **Canvassing Metrics** - Canvass-specific metrics
   - Active sessions over time
   - Visits by outcome
   - Session duration
   - Volunteer leaderboard

3. **External Services** - Integration health
   - Redis health
   - PostgreSQL health
   - Listmonk status
   - Geocoding providers

### Alert Rules

12 predefined alert rules:

**Critical Alerts:**
- API down (>5 min)
- Database unreachable
- Redis connection lost

**Warning Alerts:**
- High error rate (>5%)
- Queue backup (>1000 jobs)
- Slow requests (p95 >2s)
- Service degradation

**Info Alerts:**
- New deployment
- Service restart
- Configuration change

### Admin Interface

Observability page (`/app/observability`) with:

- **Metrics Tab** - Live metrics display
- **Dashboards Tab** - Embedded Grafana
- **Alerts Tab** - Active alerts and rules

## Architecture

### Backend Components

**Metrics Module:**
- `api/src/utils/metrics.ts` - Prometheus metrics definitions
- `api/src/modules/observability/observability.routes.ts` - Admin API

**Instrumentation:**
- Express middleware for HTTP metrics
- Service-level metric updates
- Queue size tracking
- External service health checks

**Configuration:**
- `configs/prometheus/prometheus.yml` - Scrape config
- `configs/prometheus/alerts.yml` - Alert rules
- `configs/grafana/dashboards/` - Dashboard JSON

### Frontend Components

**Admin Page:**
- `admin/src/pages/ObservabilityPage.tsx` - Monitoring dashboard
- Three tabs: Metrics, Dashboards, Alerts
- Embedded Grafana iframes
- Live metric cards

**Observability Components:**
- `admin/src/components/observability/MetricsChart.tsx` - Chart component
- `admin/src/components/observability/ServiceHealthCard.tsx` - Health display

### Docker Services

**Monitoring Profile:**

Services run with `--profile monitoring`:

```yaml
profiles: [monitoring]
  prometheus:
    image: prom/prometheus:latest
    ports: ["9090:9090"]

  grafana:
    image: grafana/grafana:latest
    ports: ["3001:3000"]

  alertmanager:
    image: prom/alertmanager:latest
    ports: ["9093:9093"]

  cadvisor:
    image: gcr.io/cadvisor/cadvisor:latest
    ports: ["8080:8080"]

  node-exporter:
    image: prom/node-exporter:latest
    ports: ["9100:9100"]

  redis-exporter:
    image: oliver006/redis_exporter:latest
    ports: ["9121:9121"]
```

## Configuration

### Environment Variables

```bash
# Enable metrics
METRICS_ENABLED=true

# Prometheus
PROMETHEUS_PORT=9090

# Grafana
GRAFANA_PORT=3001
GRAFANA_ADMIN_USER=admin
GRAFANA_ADMIN_PASSWORD=admin

# Alertmanager
ALERTMANAGER_PORT=9093
```

### Prometheus Scrape Targets

```yaml
scrape_configs:
  - job_name: 'changemaker-api'
    static_configs:
      - targets: ['api:4000']

  - job_name: 'media-api'
    static_configs:
      - targets: ['media-api:4100']

  - job_name: 'redis'
    static_configs:
      - targets: ['redis-exporter:9121']

  - job_name: 'node'
    static_configs:
      - targets: ['node-exporter:9100']

  - job_name: 'cadvisor'
    static_configs:
      - targets: ['cadvisor:8080']
```

### Alert Rules

Example alert rule:

```yaml
groups:
  - name: api_alerts
    rules:
      - alert: APIDown
        expr: up{job="changemaker-api"} == 0
        for: 5m
        labels:
          severity: critical
        annotations:
          summary: "API is down"
          description: "API has been down for 5 minutes"

      - alert: HighErrorRate
        expr: rate(http_request_duration_ms_count{status=~"5.."}[5m]) > 0.05
        for: 10m
        labels:
          severity: warning
        annotations:
          summary: "High error rate detected"
```

## Metrics Usage

### Increment Counter

```typescript
import { metrics } from '../utils/metrics';

// Campaign email sent
metrics.campaignEmailsSent.inc();

// Geocode request
metrics.geocodeRequests.inc({ provider: 'nominatim' });
```

### Set Gauge

```typescript
// Update queue size
metrics.emailQueueSize.set(queueSize);

// Update active sessions
metrics.canvassSessionsActive.set(activeSessions);

// Set service health (1 = healthy, 0 = unhealthy)
metrics.externalServiceHealth.set({ service: 'redis' }, 1);
```

### Observe Histogram

```typescript
// Time geocoding request
const end = metrics.geocodeDuration.startTimer();
try {
  await geocode(address);
  end({ success: 'true' });
} catch (error) {
  end({ success: 'false' });
}
```

## Grafana Dashboards

### Dashboard Setup

Dashboards auto-provisioned from `configs/grafana/dashboards/`:

```json
{
  "dashboard": {
    "title": "Changemaker Lite Overview",
    "panels": [
      {
        "title": "API Request Rate",
        "targets": [
          {
            "expr": "rate(http_request_duration_ms_count[5m])"
          }
        ]
      }
    ]
  }
}
```

### Accessing Dashboards

- **Direct:** http://localhost:3001 (admin/admin)
- **Embedded:** `/app/observability` → Dashboards tab
- **Subdomain:** http://grafana.cmlite.org (production)

## Alertmanager

### Alert Routing

Configure in `configs/alertmanager/alertmanager.yml`:

```yaml
route:
  receiver: 'default'
  group_by: ['alertname', 'severity']
  routes:
    - match:
        severity: critical
      receiver: 'critical-alerts'

receivers:
  - name: 'default'
    webhook_configs:
      - url: 'http://gotify:8889/message'

  - name: 'critical-alerts'
    email_configs:
      - to: 'admin@example.com'
```

### Notification Channels

Supported receivers:

- **Webhook** - Gotify, Slack, Discord
- **Email** - SMTP notifications
- **PagerDuty** - Incident management
- **Opsgenie** - Alert management

## Service Health Monitoring

### External Service Checks

Monitor services via health gauges:

```typescript
// Check Redis
try {
  await redisClient.ping();
  metrics.externalServiceHealth.set({ service: 'redis' }, 1);
} catch (error) {
  metrics.externalServiceHealth.set({ service: 'redis' }, 0);
}

// Check PostgreSQL
try {
  await prisma.$queryRaw`SELECT 1`;
  metrics.externalServiceHealth.set({ service: 'postgres' }, 1);
} catch (error) {
  metrics.externalServiceHealth.set({ service: 'postgres' }, 0);
}
```

### Docker Healthchecks

Services with healthchecks:

- **API** - `wget --spider http://localhost:4000/health`
- **Media API** - `wget --spider http://localhost:4100/health`
- **PostgreSQL** - `pg_isready`
- **Redis** - `redis-cli ping`
- **Listmonk** - `wget --spider http://localhost:9000/health`

## Performance Monitoring

### HTTP Request Tracking

Automatic tracking of:

- Request count by route
- Request duration percentiles
- Status code distribution
- Error rates

### Queue Monitoring

Track queue depths:

- Email queue size
- Geocode queue size
- Failed job count
- Processing rate

### Resource Monitoring

Via cAdvisor and Node Exporter:

- CPU usage
- Memory usage
- Disk I/O
- Network traffic

## Admin Interface

### Metrics Tab

Display cards:

- API uptime
- Request rate (req/sec)
- Error rate (%)
- Queue sizes
- Active sessions
- Service health

### Dashboards Tab

Embedded Grafana:

- Overview dashboard
- Canvassing metrics
- External services
- Custom queries

### Alerts Tab

Active alerts list:

- Alert name
- Severity
- Status (firing/pending/resolved)
- Duration
- Quick actions (silence, resolve)

## Starting Monitoring Stack

```bash
# Start with monitoring profile
docker compose --profile monitoring up -d

# Access services
# Prometheus: http://localhost:9090
# Grafana: http://localhost:3001 (admin/admin)
# Alertmanager: http://localhost:9093
```

## API Endpoints

### Observability Endpoints

```
GET    /api/observability/prometheus   # Prometheus status
GET    /api/observability/grafana      # Grafana status
GET    /api/observability/alertmanager # Alertmanager status
GET    /api/observability/metrics      # Current metrics values
```

### Metrics Endpoint

```
GET    /metrics                         # Prometheus scrape endpoint
```

## Related Documentation

- [Observability Page](../../frontend/pages/admin/observability-page.md)
- [Metrics Utilities](../../backend/utilities/index.md)
- [Docker Compose](../../deployment/docker-compose.md)
- [Monitoring Stack](../../deployment/monitoring-stack.md)
- [Healthchecks](../../deployment/healthchecks.md)
- [Performance Optimization](../../troubleshooting/performance-optimization.md)