Front-end Monitoring How-To

You can use ts-mon.js to instrument your front-end code, view metrics in your dashboards, and set up alerts based on those metrics. This document shows you how to set up ts_mon monitoring on the front-end.

Installation and Setup

The front-end library lives at crdx/chopsui/tsmon-client.js. Import it on your page to get started. The easiest way is to symlink in the chopsui directory and serve it.


Next, in your JS, instantiate the client and register a metric. You can add additional metric fields, for example a status field for XHR response status.

const tsMon = new window.chops.tsmon.TSMonClient();

// An object of metric metadata.
const metricMetaData = {};
// A Map of metric fields.
const metricFields = new Map([
  ['status', TSMonClient.intField('status')],
const metric = tsMon.cumulativeDistribution(
  'Latency for loading rutabagas.',
  metricMetaData, metricFields,

You'll also need to set up the server-side proxy handler. For Python, subclass the handler like this:

from gae_ts_mon.handlers import TSMonJSHandler

# For now, metrics need to be declared twice, once client side and once server side.
RUTABAGA_LATENCY_METRIC = ts_mon.CumulativeDistributionMetric(
  'Latency for loading rutabagas.',

class YourAppTSMonJSHandler(TSMonJSHandler):

  def __init__(self, request=None, response=None):
    super(YourAppTSMonJSHandler, self).__init__(request, response)

  def xsrf_is_valid(self, _body):
    # Do your own XSRF checking here.
    return True

Note: the Go server-side proxy handler is not yet available.

Sending Measurements

Then in JavaScript when something comes up that you want to measure:

const elapsedMs = new Date().getTime() - startTime;

const metricFields = new Map([
  ['status', <compute the value>],
metric.add(elapsedMs, metricFields);

Metrics are flushed automatically by the JS handler every 60 seconds.

The types of metrics you can measure are the same as the back-end ts_mon library. See crdx/chopsui/tsmon-client.js for more details.