Why should you care?

Ad-hoc $group on a 2 TB orders collection during board week can stall checkout. Reporting architecture protects operational traffic.

See it live — copy this example

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

// Materialize a daily rollup collection
db.orders.aggregate([
  { $match: { placedAt: { $gte: ISODate("2026-07-01"), $lt: ISODate("2026-07-02") } } },
  { $group: { _id: "$tenantId", revenue: { $sum: "$total" }, orders: { $sum: 1 } } },
  { $merge: { into: "dailyTenantRevenue", on: "_id", whenMatched: "replace", whenNotMatched: "insert" } }
], { allowDiskUse: true })
db.dailyTenantRevenue.find().sort({ revenue: -1 }).limit(20)

Run Example »

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

What happened?

• $merge writes rollup results into dailyTenantRevenue.
• Dashboards read the small rollup collection.
• allowDiskUse helps large sorts/groups.
• Schedule this job hourly/nightly.

Practice next

1. Create a rollup with $merge on practice data.
2. Point a fake dashboard query at the rollup.
3. Run the job with readPreference secondary in drivers.
4. Merge into a collection partitioned by date field.
5. Add a generatedAt timestamp in $project before merge.

Remember

Materialize reports with $merge. Isolate analytics from OLTP. Dashboards read rollups, not raw facts.

CXO morning pack

Nightly jobs build tenant revenue rollups; the morning PDF reads only rollups.

Outcome: Board metrics refresh without touching checkout databases.