Class Frequencies

All Implemented Interfaces:
CompoundStat, Measurable, MeasurableStat, MetricValueProvider<Double>, Stat

public class Frequencies extends SampledStat implements CompoundStat
A CompoundStat that represents a normalized distribution with a Frequency metric for each bucketed value. The values of the Frequency metrics specify the frequency of the center value appearing relative to the total number of values recorded.

For example, consider a component that records failure or success of an operation using boolean values, with one metric to capture the percentage of operations that failed another to capture the percentage of operations that succeeded.

This can be accomplish by created a Sensor to record the values, with 0.0 for false and 1.0 for true. Then, create a single Frequencies object that has two Frequency metrics: one centered around 0.0 and another centered around 1.0. The Frequencies object is a CompoundStat, and so it can be added directly to a Sensor so the metrics are created automatically.

  • Constructor Details

    • Frequencies

      public Frequencies(int buckets, double min, double max, Frequency... frequencies)
      Create a Frequencies that captures the values in the specified range into the given number of buckets, where the buckets are centered around the minimum, maximum, and intermediate values.
      buckets - the number of buckets; must be at least 1
      min - the minimum value to be captured
      max - the maximum value to be captured
      frequencies - the list of Frequency metrics, which at most should be one per bucket centered on the bucket's value, though not every bucket need to correspond to a metric if the value is not needed
      IllegalArgumentException - if any of the Frequency objects do not have a center value within the specified range
  • Method Details

    • forBooleanValues

      public static Frequencies forBooleanValues(MetricName falseMetricName, MetricName trueMetricName)
      Create a Frequencies instance with metrics for the frequency of a boolean sensor that records 0.0 for false and 1.0 for true.
      falseMetricName - the name of the metric capturing the frequency of failures; may be null if not needed
      trueMetricName - the name of the metric capturing the frequency of successes; may be null if not needed
      the Frequencies instance; never null
      IllegalArgumentException - if both falseMetricName and trueMetricName are null
    • stats

      Specified by:
      stats in interface CompoundStat
    • frequency

      public double frequency(MetricConfig config, long now, double centerValue)
      Return the computed frequency describing the number of occurrences of the values in the bucket for the given center point, relative to the total number of occurrences in the samples.
      config - the metric configuration
      now - the current time in milliseconds
      centerValue - the value corresponding to the center point of the bucket
      the frequency of the values in the bucket relative to the total number of samples
    • combine

      public double combine(List<SampledStat.Sample> samples, MetricConfig config, long now)
      Specified by:
      combine in class SampledStat
    • newSample

      protected org.apache.kafka.common.metrics.stats.Frequencies.HistogramSample newSample(long timeMs)
      newSample in class SampledStat
    • update

      protected void update(SampledStat.Sample sample, MetricConfig config, double value, long timeMs)
      Specified by:
      update in class SampledStat