PingIDM 7.5.0

Metrics reference

IDM exposes a number of metrics. All metrics are available at both the openidm/metrics/api and openidm/metrics/prometheus endpoints. The actual metric names can vary, depending on the endpoint used. Also refer to Monitoring.

Metric types

Metrics are organized into the following types:

Timer

Timers provide a histogram of the duration of an event, along with a measure of the rate of occurrences. Timers can be monitored using the Dropwizard dashboard widget and the IDM Prometheus endpoint. Durations in timers are measured in milliseconds. Rates are reported in number of calls per second. The following example shows a Timer metric:

{
   "_id": "sync.source.perform-action",
   "count": 2,
   "max": 371.53391,
   "mean": 370.1752705,
   "min": 368.816631,
   "p50": 371.53391,
   "p75": 371.53391,
   "p95": 371.53391,
   "p98": 371.53391,
   "p99": 371.53391,
   "p999": 371.53391,
   "stddev": 1.3586395,
   "m15_rate": 0.393388581528647,
   "m1_rate": 0.311520313228562,
   "m5_rate": 0.3804917698002856,
   "mean_rate": 0.08572717156016606,
   "duration_units": "milliseconds",
   "rate_units": "calls/second",
   "total": 740.350541,
   "_type": "timer"
 }

Summary

Summaries are similar to Timers in that they measure a distribution of events. However, Summaries record values that aren’t units of time, such as user login counts. Summaries cannot be graphed in the Dropwizard dashboard widget, but are available through the Prometheus endpoint, and by querying the openidm/metrics/api endpoint directly. The following example shows a Summary metric:

{
  "_id": "audit.recon",
  "m15_rate": 0.786777163057294,
  "m1_rate": 0.623040626457124,
  "m5_rate": 0.7609835396005712,
  "mean_rate": 0.16977218861919927,
  "units": "events/second",
  "total": 4,
  "count": 4,
  "_type": "summary"
}

Gauge

Gauge metrics return a numerical value that can increase or decrease. The value for a gauge is calculated on request, and represents the state of the metric at that specific time. The following example shows a Gauge metric:

{
  "_id": "jvm.used-memory",
  "value": 2147483648,
  "_type": "gauge"
}