measured

measured

This is an alternative port of Coda Hale's metrics library.

I created this despite the existing metrics port for node.js to fully understand the underlaying algorithms, and to provide a solid, tested and documented module.

Install

npm install measured

Usage

Step 1: Add measurements to your code. For example, lets track the requests/sec of a http server:

var http  = require('http');
var stats = require('measured').createCollection();

http.createServer(function(req, res) {
  stats.meter('requestsPerSecond').mark();
  res.end('Thanks');
}).listen(3000);

Step 2: Show the collected measurements (more advanced examples follow later):

setInterval(function() {
  console.log(stats.toJSON());
}, 1000);

This will output something like this every second:

{ requestsPerSecond:
   { mean: 1710.2180279856818,
     count: 10511,
     'currentRate': 1941.4893498239829,
     '1MinuteRate': 168.08263156623656,
     '5MinuteRate': 34.74630977619571,
     '15MinuteRate': 11.646507524106095 } }

Step 3: Aggregate the data into your backend of choice. I recommend graphite.

Metrics

The following metrics are available (both standalone and on the Collection API):

Gauge

Values that can be read instantly. Example:

var Measured = require('measured')
var gauge = new Measured.Gauge(function() {
  return process.memoryUsage().rss;
});

There is currently no callback support for Gauges because otherwise it would be very difficult to report the metrics inside a collection within a regular interval.

Options:

• Gauges take a function as parameter which needs to return their current value.

Methods:

None.

toJSON Output:

Gauges directly return their currently value.

Counter

Things that increment or decrement. Example:

var Measured = require('measured')
var activeUploads = new Measured.Counter();
http.createServer(function(req, res) {
  activeUploads.inc();
  req.on('end', function() {
    activeUploads.dec();
  });
});

Options:

• count An initial count for the counter. Defaults to 0.

Methods:

• inc(n) Increment the counter by n. Defaults to 1.
• dec(n) Decrement the counter by n. Defaults to 1.
• reset(count) Resets the counter back to count Defaults to 0.

toJSON Output:

Counters directly return their currently value.

Meter

Things that are measured as events / interval. Example:

var Measured = require('measured')
var meter = new Measured.Meter();
http.createServer(function(req, res) {
  meter.mark();
});

Options:

• rateUnit The rate unit. Defaults to 1000 (1 sec).
• tickInterval The interval in which the averages are updated. Defaults to 5000 (5 sec).

Methods:

• mark(n) Register n events as having just occured. Defaults to `1.
• reset() Resets all values. Meters initialized with custom options will be reset to the default settings (patch welcome).
• unref() Unrefs the backing timer. The meter will not keep the event loop alive. Idempotent.
• ref() Refs the backing timer again. Idempotent.

toJSON Output:

• mean: The average rate since the meter was started.
• count: The total of all values added to the meter.
• currentRate: The rate of the meter since the last toJSON() call.
• 1MinuteRate: The rate of the meter biased towards the last 1 minute.
• 5MinuteRate: The rate of the meter biased towards the last 5 minutes.
• 15MinuteRate: The rate of the meter biased towards the last 15 minutes.

Histogram

Keeps a resevoir of statistically relevant values biased towards the last 5 minutes to explore their distribution. Example:

var Measured = require('measured')
var histogram = new Measured.Histogram();
http.createServer(function(req, res) {
  if (req.headers['content-length']) {
    histogram.update(parseInt(req.headers['content-length'], 10));
  }
});

Options:

• sample The sample resevoir to use. Defaults to an ExponentiallyDecayingSample.

Methods:

• update(value, timestamp) Pushes value into the sample. timestamp defaults to Date.now().
• reset() Resets all values. Histograms initialized with custom options will be reset to the default settings (patch welcome).

toJSON Output:

• min: The lowest observed value.
• max: The highest observed value.
• sum: The sum of all observed values.
• variance: The variance of all observed values.
• mean: The average of all observed values.
• stddev: The stddev of all observed values.
• count: The number of observed values.
• median: 50% of all values in the resevoir are at or below this value.
• p75: See median, 75% percentile.
• p95: See median, 95% percentile.
• p99: See median, 99% percentile.
• p999: See median, 99.9% percentile.

Timers

Timers are a combination of Meters and Histograms. They measure the rate as well as distribution of scalar events. Since they are frequently used for tracking how long certain things take, they expose an API for that:

var Measured = require('measured')
var timer = new Measured.Timer();
http.createServer(function(req, res) {
  var stopwatch = timer.start();
  req.on('end', function() {
    stopwatch.end();
  });
});

But you can also use them as generic histograms that also track the rate of events:

var Measured = require('measured')
var timer = new Measured.Timer();
http.createServer(function(req, res) {
  if (req.headers['content-length']) {
    timer.update(parseInt(req.headers['content-length'], 10));
  }
});

Options:

• meter The internal meter to use. Defaults to a new Meter.
• histogram The internal histogram to use. Defaults to a new Histogram.

Methods:

• start() Returns a Stopwatch.
• update(value) Updates the internal histogram with value and marks one event on the internal meter.
• reset() Resets all values. Timers initialized with custom options will be reset to the default settings (patch welcome).
• unref() Unrefs the backing timer. The internal meter will not keep the event loop alive. Idempotent.
• ref() Refs the backing timer again. Idempotent.

toJSON Output:

• meter: See Meter toJSON output docs above.
• histogram: See Histogram toJSON output docs above.

Todo

• Implement flatten() so reporters can use it
• Implement async gauges
• Document using this with graphite / zabbix

License

This module is licensed under the MIT license.