Read StatsD plugin - Netuitive/collectdwin GitHub Wiki

Read StatsD plugin

The Read StatsD plugin implements the StatsD network protocol. It listens on a UDP port and receives metrics from applications. This plugin supports the following four event types:

Gauges

A gauge is an instantaneous measurement of a value, like the gas gauge in a car. It differs from a counter by being calculated at the client rather than the server.

<metric name>:<value>|g

Counters

A counter is a gauge calculated at the server. Metrics sent by the client increment or decrement the value of the gauge rather than giving its current value. Counters may also have an associated sample rate, given as a decimal of the number of samples per event count. For example, a sample rate of 1/10 would be exported as 0.1. Valid counter values are in the range (-2^63^, 2^63^).

<metric name>:<value>|c[|@<sample rate>]

Timers

A timer is a measure of the number of milliseconds elapsed between a start and end time, for example the time to complete rendering of a web page for a user.

<metric name>:<value>|ms

Sets

Unique occurrences of events within an interval.

uniques:765|s

Configuration

<ReadStatsd Host="localhost" Port="8125">
   <DeleteCache Counters="true" Timers="true" Gauges="true" Sets="true" />
   <Timer Lower="true" Upper="true" Sum="true" Count="true">
      <Percentiles>
         <Percentile Value="90.0" />
         <Percentile Value="95.0" />
      </Percentiles>
   </Timer>
</ReadStatsd>

Properties

Name Description
Statsd.Host Host name for StatsD UDP listener
Statsd.Port Port number for StatsD UDP listener
DeleteCache.Counters, DeleteCache.Timers, DeleteCache.Gauges, DeleteCache.Sets This option controls the behavior when metrics are not updated in an interval. If set to false, metrics are dispatched unchanged. If set to true, metrics are not dispatched and removed from the internal cache.
Timer.Lower, Timer.Upper, Timer.Sum, Timer.Count This option is to enable aggregate functions for timer metrics. If enabled, for each timer metric, its equivalent aggregate metric is also dispatched with appropriate extensions ("-lower", "-upper", "-sum", "-count")
Percentiles.Percentile Calculate and dispatch the configured percentile, i.e. compute the latency, so that Percent of all reported timers are smaller than or equal to the computed latency. If enabled, for each timer metric, its equivalent percentile metric is also dispatched with appropriate extensions ("-percentile-90", "-percentile-95"). If not configured, no percentile is calculated / dispatched.
⚠️ **GitHub.com Fallback** ⚠️