Skip to content

Monitoring

Terminal window
curl http://localhost:8080/health

Returns the health status of the Alpha node.

The dashboard provides real-time monitoring for your databases:

  • Query count — queries per minute/hour/day
  • Storage usage — current storage consumption and billing estimate
  • Latency — p50, p95, p99 query latency
  • Materialization stats — inferred edges per mutation
  • Error rate — failed queries and rejected mutations

OWLGraph exposes Prometheus-compatible metrics at /debug/prometheus_metrics:

Terminal window
curl http://localhost:8080/debug/prometheus_metrics

Key metrics:

MetricDescription
dgraph_num_queries_totalTotal query count
dgraph_num_mutations_totalTotal mutation count
dgraph_latency_secondsQuery/mutation latency histogram
dgraph_pending_queriesCurrently executing queries
dgraph_disk_usage_bytesStorage consumption

For self-hosted deployments, connect Prometheus to Grafana for dashboards and alerting. Sample Grafana dashboards are available in the OWLGraph repository.

Paid accounts include configurable alerts for:

  • Storage approaching limit
  • Query volume approaching limit
  • Latency exceeding thresholds
  • Instance health degradation

Configure alerts in your monitoring stack (Prometheus + Alertmanager recommended).