Skip to main content

Monitoring

Monitor your Synjar deployment for uptime, performance, and issues.

Health endpoint​

curl http://localhost:6200/health

Response:

{
  "status": "ok",
  "timestamp": "2025-01-01T12:00:00Z"
}

Use this endpoint for:

  • Load balancer health checks
  • Uptime monitoring
  • Container orchestration liveness probes

Metrics​

Prometheus metrics​

If enabled, metrics are available at /metrics:

curl http://localhost:6200/metrics

Key metrics:

  • http_requests_total - Total HTTP requests
  • http_request_duration_seconds - Request latency
  • db_queries_total - Database queries
  • document_processing_duration - Processing time

Grafana dashboard​

Import the Synjar dashboard from our Grafana catalog (ID: 12345).

Logging​

Log levels​

Set via LOG_LEVEL environment variable:

  • error - Only errors
  • warn - Errors and warnings
  • info - Standard logging (default)
  • debug - Verbose debugging

Log format​

JSON format for easy parsing:

{
  "level": "info",
  "message": "Document processed",
  "documentId": "abc123",
  "duration": 1234,
  "timestamp": "2025-01-01T12:00:00Z"
}

Viewing logs​

# Docker Compose
docker-compose logs -f synjar

# Docker
docker logs -f synjar

# Kubernetes
kubectl logs -f deployment/synjar

Log aggregation​

Send logs to centralized logging:

Loki/Grafana:

# docker-compose.override.yml
services:
  synjar:
    logging:
      driver: loki
      options:
        loki-url: "http://loki:3100/loki/api/v1/push"

Elasticsearch:

services:
  filebeat:
    image: elastic/filebeat
    volumes:
      - /var/lib/docker/containers:/var/lib/docker/containers:ro

Alerting​

AlertConditionSeverity
Service downHealth check fails 3xCritical
High latencyp95 > 2s for 5 minWarning
High error rate>1% errors for 5 minWarning
Database connectionConnection errorsCritical

Example Prometheus rules​

groups:
  - name: synjar
    rules:
      - alert: SynjarDown
        expr: up{job="synjar"} == 0
        for: 1m
        labels:
          severity: critical
        annotations:
          summary: "Synjar is down"

      - alert: HighLatency
        expr: histogram_quantile(0.95, http_request_duration_seconds) > 2
        for: 5m
        labels:
          severity: warning
        annotations:
          summary: "High request latency"

Uptime monitoring​

External monitoring services:

  • UptimeRobot
  • Pingdom
  • StatusCake
  • Better Uptime

Monitor:

  • /health endpoint
  • Main application URL
  • Critical API endpoints

Performance optimization​

Database​

Check slow queries:

SELECT query, calls, mean_time
FROM pg_stat_statements
ORDER BY mean_time DESC
LIMIT 10;

Application​

Monitor for:

  • Memory usage trends
  • CPU utilization
  • Request queue depth

See also​