codedebug / pmx

Keymetrics++ driver

Driver for Keymetrics

PMX is a module that allows you to create advanced interactions with Keymetrics.

With it you can:

• Trigger remote actions / functions
• Analyze custom metrics / variables (with utilities like Histogram/Counter/Metric/Meters)
• Report errors (uncaught exceptions and custom errors)
• Emit events
• Analyze HTTP latency

Installation

Install PMX and add it to your package.json via:

$ npm install pmx --save

Then init the module to monitor HTTP, Errors and diverse metrics.

var pmx = require('pmx').init(); // By default everything is enabled and ignore_routes is empty

Or choose what to monitor.

var pmx = require('pmx').init({
  http          : true,
  errors        : true,
  custom_probes : true,
  ignore_routes : [/socket\.io/, /notFound/]
});

Custom monitoring

Emit Events

Emit events and get historical and statistics:

var pmx = require('pmx');

pmx.emit('user:register', {
  user : 'Alex registered',
  email : '[email protected]'
});

Trigger function from remote

var pmx = require('pmx');

pmx.action('db:clean', { comment : 'Description for this action' }, function(reply) {
  clean.db(function() {
    /**
     * reply() must be called at the end of the action
     */
     reply({success : true});
  });
});

Note: in case of exceptions in the function, your app will not be affected

Errors

Catch uncaught exceptions:

var pmx = require('pmx').init();

Attach more data from errors that happens in Express:

var pmx = require('pmx');

app.get('/' ...);
app.post(...);

app.use(pmx.expressErrorHandler());

Trigger custom errors:

var pmx = require('pmx');

pmx.notify({ success : false });

pmx.notify('This is an error');

pmx.notify(new Error('This is an error'));

HTTP latency analysis

Monitor routes, latency and codes. REST complient.

pmx.http(); // You must do this BEFORE any require('http')

Ignore some routes by passing a list of regular expressions.

pmx.http([/socket\.io/, /notFound/]);

This can also be done via pmx.init() by passing the ignore_routes option.

pmx.init({
  http          : true,
  ignore_routes : [/socket\.io/, /notFound/]
});

This module is enabled by default if you called pmx with the init() function.

Measure

Measure critical segments of you code thanks to 4 kind of probes:

• Simple metrics: Values that can be read instantly
  • Monitor variable value
• Counter: Things that increment or decrement
  • Downloads being processed, user connected
• Meter: Things that are measured as events / interval
  • Request per minute for a http server
• Histogram: Keeps a resevoir of statistically relevant values biased towards the last 5 minutes to explore their distribution
  • Monitor the mean of execution of a query into database

Metric

Values that can be read instantly.

var probe = pmx.probe();

var metric = probe.metric({
  name  : 'Realtime user',
  value : function() {
    return Object.keys(users).length;
  }
});

Counter

Things that increment or decrement.

var probe = pmx.probe();

var counter = probe.counter({
  name : 'Downloads'
});

http.createServer(function(req, res) {
  counter.inc();
  req.on('end', function() {
    counter.dec();
  });
});

Meter

Things that are measured as events / interval.

var probe = pmx.probe();

var meter = probe.meter({
  name    : 'req/min',
  seconds : 60
});

http.createServer(function(req, res) {
  meter.mark();
  res.end({success:true});
});

Options

seconds option is the measurement rate of the meter, default is 1 seconds

Histogram

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

var probe = pmx.probe();

var histogram = probe.histogram({
  name        : 'latency',
  measurement : 'mean'
});

var latency = 0;

setInterval(function() {
  latency = Math.round(Math.random() * 100);
  histogram.update(latency);
}, 100);

Options

measurement option can be:

• 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.

Modules

Simple app

process.env.MODULE_DEBUG = true;

var pmx  = require('pmx');

var conf = pmx.initModule();

License

MIT