Initialization
Assuming you have completed the Node.js Tracker Setup for your project, you are now ready to initialize the Tracker.
Require the Node.js Tracker module into your code like so:
const snowplow = require('@snowplow/node-tracker');
const tracker = snowplow.newTracker( /* ... */ );
or, if using ES Modules, you can import the module like so:
import { newTracker } from '@snowplow/node-tracker';
The newTracker
call takes 2 arguments – the tracker and emitter configuration. The tracker configuration specifies the tracker namespace and app name. The emitter configuration specifies how events are queued locally and sent to the Snowplow collector.
const tracker = newTracker({
namespace: "my-tracker", // Namespace to identify the tracker instance. It will be attached to all events which the tracker fires.
appId: "my-app", // Application identifier
encodeBase64: false, // Whether to use base64 encoding for the self-describing JSON. Defaults to true.
}, {
endpoint: "https://collector.mydomain.net", // Collector endpoint
eventMethod: "post", // Method - defaults to POST
bufferSize: 1, // Only send events once n are buffered. Defaults to 1 for GET requests and 10 for POST requests.
});
There are a number of additional parameters that may optionally be configured. Please refer to the following API docs for the full list:
Using multiple Emitters
In case you want to send the events to multiple Snowplow collectors, you can provide a list of emitter configurations in the newTracker
call – one for each collector.
This may look as follows.
const tracker = newTracker({
namespace: "my-tracker", // Namespace to identify the tracker instance. It will be attached to all events which the tracker fires.
appId: "my-app", // Application identifier
encodeBase64: false, // Whether to use base64 encoding for the self-describing JSON. Defaults to true.
}, [
{
endpoint: "https://collector1.mydomain.net", // First collector endpoint
},
{
endpoint: "https://collector2.mydomain.net", // Second collector endpoint
},
]);
Create your own Emitter
Emitter is an object responsible for queuing tracked events and making requests to the Snowplow collector.
By default, the Node tracker makes use of the fetch API to make the HTTP requests.
If you want to use a different library or logic for queuing and sending events, you can provide a custom Emitter
implementation.
Emitters must conform to an Emitter
interface, which looks like:
/**
* Emitter is responsible for sending events to the collector.
* It manages the event queue and sends events in batches depending on configuration.
*/
interface Emitter {
/**
* Forces the emitter to send all events in the event store to the collector.
* @returns A Promise that resolves when all events have been sent to the collector.
*/
flush: () => Promise<void>;
/**
* Adds a payload to the event store or sends it to the collector.
* @param payload - A payload to be sent to the collector
* @returns Promise that resolves when the payload has been added to the event store or sent to the collector
*/
input: (payload: Payload) => Promise<void>;
/**
* Updates the collector URL to which events will be sent.
* @param url - New collector URL
*/
setCollectorUrl: (url: string) => void;
/**
* Sets the server anonymization flag.
*/
setAnonymousTracking: (anonymous: boolean) => void;
/**
* Updates the buffer size of the emitter.
*/
setBufferSize: (bufferSize: number) => void;
}
Once you implement your own Emitter
, you can pass it to the newTracker
call as follows.
export const expressTracker = newTracker(
{
namespace: "my-tracker",
appId: "my-app",
},
{
customEmitter: () => ({
input: (payload) => {
return Promise.resolve();
},
flush: () => {
return Promise.resolve();
},
setCollectorUrl: (url) => {},
setAnonymousTracking: (anonymous) => {},
setBufferSize: (bufferSize) => {},
}),
}
);