aquipa / cloudwatch-metrics Goto Github PK

This module provides a simplified wrapper for creating and publishing CloudWatch metrics.

$ npm install cloudwatch-metrics

or

$ npm install cloudwatch-metrics --save

By default, the library will log metrics to the us-east-1 region and read AWS credentials from the AWS SDK's default environment variables.

If you want to change these values, you can call initialize:

var cloudwatchMetrics = require('cloudwatch-metrics');
cloudwatchMetrics.initialize({
	region: 'us-east-1'
});

For creating a metric, we simply need to provide the namespace and the default type of metric:

var myMetric = new cloudwatchMetrics.Metric('namespace', 'Count');

If we want to add our own default dimensions, such as environment information, we can add it in the following manner:

var myMetric = new cloudwatchMetrics.Metric('namespace', 'Count', [{
	Name: 'environment',
	Value: 'PROD'
}]);

If we want to disable a metric in certain environments (such as local development), we can make the metric in the following manner:

// isLocal is a boolean
var isLocal = someWayOfDetermingIfLocal();

var myMetric = new cloudwatchMetrics.Metric('namespace', 'Count', [{
	Name: 'environment',
	Value: 'PROD'
}], {
	enabled: isLocal
});

The full list of configuration options is:

Then, whenever we want to publish a metric, we simply do:

myMetric.put(value, metric, units, additionalDimensions);

Instead of sending individual data points for your metric, you may want to send summary metrics. Summary metrics track statistics over time, and send those statistics to CloudWatch on a configurable interval. For instance, you might want to know your total network throughput, but you don't care about individual request size percentiles. You could use summaryPut to track this data and send it to CloudWatch with fewer requests:

var metric = new cloudwatchMetrics.Metric('namespace', 'Bytes');

function onRequest(req) {
	// This will still track maximum, minimum, sum, count, and average, but won't
	// take up lots of CloudWatch requests doing so.
	metric.summaryPut(req.size, 'requestSize');
}

Note that metrics use different summaries for different dimensions, and that the order of the dimensions is significant! In other words, these track different metric sets:

var metric = new cloudwatchMetrics.Metric('namespace', 'Bytes');
// Different statistic sets!
metric.summaryPut(45, 'requestSize', [{Name: 'Region', Value: 'US'}, {Name: 'Server', Value: 'Card'}]);
metric.summaryPut(894, 'requestSize', [{Name: 'Server', Value: 'Card'}, {Name: 'Region', Value: 'US'}]);

Be aware that the put call does not actually send the metric to CloudWatch at that moment. Instead, it stores unsent metrics and sends them to CloudWatch on a predetermined interval (to help get around sending too many metrics at once - CloudWatch limits you by default to 150 put-metric data calls per second). The default interval is 5 seconds, if you want metrics sent at a different interval, then provide that option when constructing your CloudWatch Metric:

var myMetric = new cloudwatchMetrics.Metric('namespace', 'Count', [{
	Name: 'environment',
	Value: 'PROD'
}], {
	sendInterval: 3 * 1000 // It's specified in milliseconds.
});

You can also register a callback to be called when we actually send metrics to CloudWatch - this can be useful for logging put-metric-data errors:

var myMetric = new cloudwatchMetrics.Metric('namespace', 'Count', [{
	Name: 'environment',
	Value: 'PROD'
}], {
	sendCallback: (err) => {
		if (!err) return;
		// Do your error handling here.
	}
});