Pulsar Detector

The Pulsar detector monitors the health of Pulsar clusters. When you need to monitor the distribution latency from message sending to message consumption in one cluster or across clusters, you can use the Pulsar detector. The Pulsar detector enables you to easily integrate with existing monitoring tools like Prometheus, and check the critical metrics in Grafana.

Pulsar detector is enabled by default in StreamNative Platform. To check the health status of your Pulsar cluster, you can check it in your pulsar_detector dashboard in Grafana. The following is an example.

screenshot of pulsar_detector dashboard in Grafana

To disable the tool, you can set pulsar_detector to false in your ${PLATFORM_HOME}/pulsar/values.yaml file.

Metrics

Pulsar detector supports the following metrics. You can check the complete list of metrics in Grafana.

Name	Type	Description
pulsar_detector_e2e_latency_ms	Summary	The distribution latency from message sending to message consumption.
pulsar_detector_publish_latency_ms	Summary	The distribution latency of message sending.
pulsar_detector_geo_latency_ms	Summary	The distribution latency from message sending to message consumption across clusters.
pulsar_sla_messaging_up	Gauge	Whether the message publishing service is running.
pulsar_sla_webservice_up	Gauge	Whether the webservice is running. The value `1` means the webservice is running. The value `0` means the webservice is not running.