Why should you care?

IoT devices can emit a reading every second. One document per reading explodes storage and indexes; bucketing keeps time-series manageable (or use time-series collections).

See it live — copy this example

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

db.readings.insertOne({
  deviceId: "sensor-42",
  hour: ISODate("2026-07-19T10:00:00Z"),
  samples: [
    { t: ISODate("2026-07-19T10:00:01Z"), temp: 28.1 },
    { t: ISODate("2026-07-19T10:00:02Z"), temp: 28.2 }
  ],
  count: 2
})
db.readings.updateOne(
  { deviceId: "sensor-42", hour: ISODate("2026-07-19T10:00:00Z") },
  {
    $push: { samples: { t: ISODate("2026-07-19T10:00:03Z"), temp: 28.3 } },
    $inc: { count: 1 }
  }
)

Run Example »

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

What happened?

• One bucket document covers an hour for a device.
• New samples $push into samples and $inc count.
• Queries for an hour fetch one doc instead of 3600.

Practice next

1. Insert the bucket document.
2. Push two more samples with updateOne.
3. Find by deviceId and hour range.
4. Add minTemp/maxTemp maintained on each push.
5. Query count: { $gte: 100 } for busy hours.

Remember

Bucket many tiny events into one document. Key buckets by entity + time window. Great for IoT and metrics.

Factory temperature monitors

Each machine writes buckets per minute to MongoDB for dashboards.

Outcome: Hourly charts load from dozens of docs, not millions.