ambrite-mdd

ambrite-mdd library

Use this library in your server side (node) TypeScript (or javascript) code.

Uses Stackdriver to LOG and MONTIOR.

Logging

Log-line based reporting, featuring:

• text payloads
• json payloads
• severities
• logBook name

Monitoring

Write logs to custom metrics. NOTE: don't use too many different log names.

• custom names
• custom labels
• time-series based

Installation

npm install -s ambrite-mdd

Note: includes TypeScript types.

Usage

Logging

Installation

Import the library, Logger part: import { Logger } from "ambrite-mdd"

Add new logBook (or reference an old one)

Setup a new log: const log = new Logger(projectId, logName) where projectId is the name of the GCP project and logName is the name of the log to write to.

Write a log

And finally to write a log: log.debug("some string")

or generic: log.level(payload)

LogLevel

where the level is one of:

• log.log(...) uses DEFAULT
• log.debug(...) uses DEBUG
• log.notice(...) uses NOTICE
• log.warn(...) uses WARNING
• log.error(...) uses ERROR
• log.critical(...) uses CRITICAL
• log.alert(...) uses ALERT
• log.emergency(...) uses EMERGENCY

All these log levels are supported by stackdriver.

Payload

Currently, the payload you submit to the function is directly used in the stackdriver methods.

Specifically, you can use:

Text payloads; log.debug("Hello")

JSON payloads: log.debug({foo: {bar: 5}})

Retrieving logs

• Go to the GCP console.
• Go to the Logging subsection
• Switch to advanced filter
• start typing: logName (should auto-complete)
• then type = and a drop down with all logBooks should appear, so you can just choose the log you need

Example: logName="projects/testProject/logs/testLog"

or generically: logName="projects/{projectID}/logs/{logName}"

Monitoring

Installation

Import the library, Monitoring part: import { Monitoring } from "ambrite-mdd"

Create a new metric (or reference an old one)

const monitoring = new Monitoring(projectId);
const metric = Monitoring.createMetric("testService/metric1", {
  label1: "tag1",
});

This creates a new metric on our project. Note that we could add more/other metrics with different tags.

Do NOT create too many metrics, but try to use labels to make differences.

This is all you need to configure a metric.

Writing to a metric

Now, in the code, you add a new point to the metric by:

monitoring.addPoint(123, metric);

This adds the point 123 to the metric. Note that this API allows us to simply define various metrics and ad-hoc choose the metric to write to.

Retrieving Metrics

• Go to your GCP Project
• Navigate to Monitoring
• Go to the metrics explorer
• Under the global resource, you should find your metric

Note: Might take a minute or two to feed through all the systems.