ingester-prometheus-metrics

Export Metrics

Metrics monitoring is a critical component for ensuring optimal ingester performance and reliability. It provides:

• Real-time visibility into ingester health and performance
• Proactive monitoring capabilities for maintaining system stability
• Rapid troubleshooting and issue resolution

For detailed information about available metrics, including descriptions and usage guidelines, see our comprehensive metrics documentation.

Start Metrics Exporter

To start an HTTP server that serves Prometheus metrics, initialize the metrics exporter before starting the GreptimeDB ingester:

    MetricsExporter metricsExporter = new MetricsExporter(MetricsUtil.metricRegistry());
    metricsExporter.init(ExporterOptions.newDefault());
    // Start GreptimeDB ingester
    // ...

Get Metrics

You can check the output of curl http://<host>:<port>/metrics by getting the latest metrics of the ingester.

Export Metrics to Prometheus

The ingester supports exporting metrics to Prometheus. Before configuring export of metrics, you need to setup Prometheus by following their official documentation.

To scrape metrics from the ingester, write a Prometheus configuration file and save it as prometheus.yml:

global:
  scrape_interval: 15s

scrape_configs:
  - job_name: 'greptimedb-ingester'
    static_configs:
      # Assuming that the ingester is running locally.
      # The default HTTP port of 8090.
      - targets: ['localhost:8090']

Start Prometheus using the configuration file:

./prometheus --config.file=prometheus.yml

Grafana Dashboard

You can import dashboard via JSON model: greptimedb-ingester-dashboard.json