ts_prometheus

A prometheus client for Deno that supports counter, gauge, histrogram and summary metric types.

Usage

By default all metrics are registered in the global Registry accessible via Registry.default. The Registry class has the method metrics() that returns the text-based exposition for all metrics collected. But it is possible specified one or more custom registry:

const myRegistry = new Registry();
const myCounter = Counter.with({
  name: "my_counter",
  help: "a counter with custom registry",
  registry: [myRegistry],
});

Examples

Metric Types

Counter

const counter = Counter.with({
  name: "http_requests_total",
  help: "The total number of HTTP requests.",
  labels: ["method", "status"],
});

const totalGetCreate = counter.labels({
  method: "GET",
  status: "201",
});

totalGetCreate.inc();
totalGetCreate.inc(42);

# HELP http_requests_total The total number of HTTP requests.
# TYPE http_requests_total counter
http_requests_total{method="GET",status="201"} 43

Gauge

const gauge = Gauge.with({
  name: "cpu_time_usage",
  help: "The CPU time usage.",
  labels: ["mode"],
});

const cpuIdle = gauge.labels({
  mode: "idle",
});

cpuIdle.set(0);
cpuIdle.inc();
cpuIdle.inc(42);
cpuIdle.dec();
cpuIdle.dec(3.14);

# HELP cpu_time_usage The CPU time usage.
# TYPE cpu_time_usage gauge
cpu_time_usage{mode="idle"} 38.86

Histogram

const histogram = Histogram.with({
  name: "http_requests_duration",
  help: "A histogram of the requests duration.",
  buckets: [.05, .1, .2, .5, 1, 3],
});

histogram.observe(.42);
histogram.observe(.58);

# HELP http_requests_duration A histogram of the requests duration.
# TYPE http_requests_duration histogram
http_requests_duration_bucket{le="0.05"} 0
http_requests_duration_bucket{le="0.1"} 0
http_requests_duration_bucket{le="0.2"} 0
http_requests_duration_bucket{le="0.5"} 1
http_requests_duration_bucket{le="1"} 2
http_requests_duration_bucket{le="3"} 2
http_requests_duration_bucket{le="+Inf"} 2
http_requests_duration_sum 1
http_requests_duration_count 2

Summary

By default percentiles when not set are [ .01, .05, .9, .95, .99 ].

let summary = Summary.with({
  name: "http_response_size",
  help: "A summary of the response size.",
  percentiles: [.25, .5, .75, 1],
});

let values = [1, 1, 2, 3, 5, 8, 13, 21, 34, 55];
values.forEach((v) => summary.observe(v));

# HELP http_response_size A summary of the response size.
# TYPE http_response_size summary
http_response_size{percentile="0.25"} 2
http_response_size{percentile="0.5"} 5
http_response_size{percentile="0.75"} 21
http_response_size{percentile="1"} 55
http_response_size_sum 143
http_response_size_count 10

A sliding time window can be set using maxAge for defining the age of observation in milliseconds, or ageBuckets for limiting the max number of observations.

Summary.with({
  name: "http_response_size",
  help: "A summary of the response size.",
  percentiles: [ .25, .5, .75, 1 ]
  maxAge: 1000 // milliseconds
  ageBuckets: 5 // number of observations
});