NAME

Metrics::Any::Adapter::Prometheus - a metrics reporting adapter for Prometheus

SYNOPSIS

   use Metrics::Any::Adapter 'Prometheus';

DESCRIPTION

This Metrics::Any adapter type reports metrics to Prometheus by using Net::Prometheus. Each metric added to the adapter will be registered with the global Net::Prometheus::Registry instance.

It becomes the calling program's responsibility to arrange for these to be HTTP accessible by using the Net::Prometheus API.

Distribution metrics are exported as Histograms by default. They may alternatively be exported as Summaries in order to generate smaller amounts of export data, by setting the use_histograms import argument to false:

   use Metrics::Any::Adapter 'Prometheus', use_histograms => 0;

Timer metrics are implemented as distribution metrics with the units set to seconds.

This adapter type supports batch mdoe reporting. Callbacks are invoked at the beginning of the Net::Prometheus render method.

METHODS

make_distribution

   $adapter->make_distribution( $name, %args )

In addition to the standard arguments, the following are recognised:

buckets => ARRAY[ NUM ]

If present, overrides the default Histogram bucket sizes.

bucket_min => NUM

bucket_max => NUM

buckets_per_decade => ARRAY[ NUM ]

Since version 0.04.

A more flexible alternative to specifying literal bucket sizes. The values given in buckets_per_decade are repeated, multiplied by various powers of 10 to generate values between bucket_min (or a default of 0.001 if not supplied) and bucket_max (or a default of 1000 if not supplied).

For more information, see Net::Prometheus::Histogram.

AUTHOR

Paul Evans <leonerd@leonerd.org.uk>