Why should you care?

Guessing indexes wastes RAM. Real optimization uses executionStats: nReturned vs totalDocsExamined ratio should be close to 1 for selective queries.

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.products.createIndex({ category: 1, price: 1 })
const plan = db.products.find({ category: "audio", price: { $lte: 3000 } })
  .sort({ price: 1 })
  .explain("executionStats")
printjson({
  stage: plan.queryPlanner.winningPlan,
  examined: plan.executionStats.totalDocsExamined,
  returned: plan.executionStats.nReturned
})

Run Example »

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

What happened?

• The compound index supports filter + price sort.
• printjson summarizes examined vs returned.
• If examined >> returned, tighten the filter or fix the index order.

Practice next

1. Pick your slowest find from logs.
2. Run explain("executionStats").
3. Create or adjust one index.
4. Force a hint({ category: 1, price: 1 }) and compare plans.
5. Remove an unused index after checking $indexStats.

Remember

Use explain executionStats. Minimize docs examined. Index for real filters and sorts.

Black Friday latency war room

Engineers sort Atlas profiler by duration and fix the worst COLLSCAN first.

Outcome: Checkout browse p99 returns under budget.