A Distribution Summary is used to track the distribution of events. It is similar to a [Timer], but more general, in that the size does not have to be a period of time. For example, a distribution summary could be used to measure the payload sizes of requests hitting a server or the number of records returned from a query.
It is recommended to always use base units when recording the data. So, if measuring the payload size use bytes, not kilobytes or some other unit. This allows the presentation layer for graphing to use either SI or IEC prefixes in a natural manner, and you do not need to consider the meaning of something like "milli-milliseconds".
Distribution summaries report summarized statistics about the measurements for a time window
totalOfSquares. If you were to simply query for
the name of your timer via
you would get a nonsense value that is the sum of the reported statistics.
When querying the results of a distribution summary, either select one of the statistics above via a filter, or use one of the operators below to generate a useful response.
Average Measurement (:dist-avg)¶
To compute the average latency across an arbitrary group, use the :dist-avg function:
Maximum Measurement (:dist-max)¶
To compute the maximum latency across a group, use :dist-max:
Standard Deviation of Measurement (:dist-stddev)¶
To compute the standard deviation of measurements across all instances for a time interval:
nnf.cluster,foo,:eq, name,http.req.payload.size,:eq, :and, :dist-stddev
Note that it is possible to plot the individual statics by filtering on the
If you choose to do so, note that the
totalOfSquares are counters
thus reported as rates per second, while the
max is reported as a gauge.