Why should you care?

Chat fan-out, cache invalidation, and event-driven microservices need “tell me when data changes” without polling.

See it live — copy this example

Open mongosh or MongoDB Compass, select database nosqlverse, then run the example. Change one field and run again.

// mongosh / Node concept — watch new chat messages
const stream = db.messages.watch([
  { $match: { "operationType": "insert", "fullDocument.roomId": "lobby" } }
], { fullDocument: "updateLookup" })
// In Node:
// const change = await stream.next();
// change.fullDocument contains the new message
db.messages.insertOne({ roomId: "lobby", text: "hi", at: new Date() })

Run Example »

Edit the code below and click Run to see the result in Toolliyo’s live editor.

What happened?

• watch opens a cursor on changes.
• The pipeline filters to inserts for room lobby.
• Inserting a message emits an event the app can push over WebSockets.
• Resume tokens allow restart after disconnects.

Practice next

1. Ensure you are on a replica set (Atlas is).
2. Open a change stream on messages in a Node script.
3. Insert a document in another shell and see the event.
4. Watch update operations on orders.status.
5. Add a $project stage in the watch pipeline.

Remember

Change streams push data-change events. Need replica set / Atlas. Filter early; resume carefully.

Live order tracking

When status flips to shipped, a change stream notifies the mobile push service.

Outcome: Customers get “out for delivery” without the app polling Mongo.