Dev GuideAPI Reference
Dev GuideAPI ReferenceUser GuideGitHubNuGetDev CommunityDoc feedbackLog In

Event batching

This topic describes how the Optimizely JavaScript (Node) SDK uses the event processor to batch impressions and conversion events into a single payload before sending it to Optimizely.

Starting with version 3.3.0, the SDK lets you batch events and includes options to set a maximum batch size and flush interval timeout. The benefit of event batching is less network traffic for the same number of Impression and conversion events tracked.

The event batching functionality works by processing impression events from Activate and conversion events from Track placing them into a queue. The queue is drained when it either reaches its maximum size limit or when the flush interval is triggered.

By default, event batching is enabled in JavaScript SDK versions 3.3.0 and newer.

📘

Note

Event batching works with both out-of-the-box and custom event dispatchers.

The event batching process does not remove any personally identifiable information (PII) from events. You must still ensure that you aren't sending any unnecessary PII to Optimizely.

Configure event batching

We provide two options to configure event batching: eventBatchSize and eventFlushInterval. You can pass in both options during client creation. Events are held in a queue until either:

  • The number of events reaches the defined eventBatchSize.
  • The oldest event has been in the queue for longer than the defined eventFlushInterval, which is specified in milliseconds. The queue is then flushed and all queued events are sent to Optimizely in a single network request.
  • A new datafile revision is received. This occurs only when live datafile updates are enabled.
const optimizelySdk = require('@optimizely/optimizely-sdk')

optimizelySdk.createInstance({
  // other options
  eventBatchSize: 100,
  eventFlushInterval: 3000,
})

The table below defines these options and lists general recommendations for browser and server default settings. On the browser we recommend using a small eventBatchSize (10) and a short eventFlushInterval (1000). This ensures that events are sent in a relatively fast manner, since some events could be lost if a user immediately bounces, while gaining network efficiency in cases where many decisions or Track calls happen back-to-back.

For the server, these numbers are more flexible and should be set based on your organization's traffic and needs. Please contact support for recommendations for your specific implementation.

NameDescriptionRecommended Value
eventBatchSize The maximum number of events to hold in the queue. Once this number is reached, all queued events are flushed and sent to Optimizely.
Default: 10

Note: By setting this value to 1, events aren't batched and event requests behave exactly as in pre-3.3.0 JavaScript SDKs.
Based on your organization's requirements.
eventFlushInterval The maximum duration in milliseconds that an event can exist in the queue before being flushed.
Default: 30000 ms
Based on your organization's requirements.

📘

Note

The maximum payload size is 3.5 MB. If the resulting batch payload exceeds this limit, requests will be rejected with a 400 response code, Bad Request Error.

The most common cause of a large payload size is a high batch size. If your payloads are exceeding the size limit, try configuring a smaller batch size.

Close Optimizely on application exit

If you enable event batching, it i s important that you call the Close method (optimizely.close()) prior to exiting. This ensures that queued events are flushed as soon as possible to avoid any data loss.

Important

Because the Optimizely client maintains a buffer of queued events, you must call close() on the Optimizely instance before shutting down your application or whenever dereferencing the instance.

MethodDescription
close() Stops all timers and flushes the event queue. This method will also stop any timers that are happening for the datafile manager.

Note: We recommend that you connect this method to a kill signal for the running process.

You can hook into the SIGTERM event to ensure that close() is invoked, which guarantees that all events in the queue are sent.

// respond to SIGTERM and close optimizely
process.on('SIGTERM', async () => {
  console.log('SIGTERM signal received.');
  await optimizely.close()
  process.exit(0)
});