Horizontal scaling mental model in MongoDB - Mini Project: Build & Apply

Horizontal scaling mental model

📖 Scenario: Your e-commerce application's orders collection has grown to 500 million documents. A single MongoDB server can no longer handle the read/write load. You need to shard the collection across multiple servers to distribute the data and traffic.

🎯 Goal: Enable sharding on a database, choose an appropriate shard key, shard a collection, and verify the data distribution across shards.

📋 What You'll Learn

Enable sharding on the database

Shard a collection with an appropriate compound key

Check the shard distribution

Configure the balancer window

💡 Why This Matters

🌍 Real World

Sharding is how MongoDB scales to billions of documents across companies like eBay, Forbes, and Uber that need to handle massive data volumes.

💼 Career

Database administrators and backend engineers must understand sharding to design MongoDB deployments that scale with growing data and traffic.

1

Enable sharding on the database

Use sh.enableSharding() to enable sharding on the "ecommerce" database.

MongoDB

// Enable sharding on the ecommerce database
// Your code here

Need a hint?

Use sh.enableSharding("databaseName") to enable sharding on a specific database.

2

Create index and shard the collection

First create a compound index on { "region": 1, "orderId": 1 } for the ecommerce.orders collection using db.orders.createIndex(). Then use sh.shardCollection() to shard "ecommerce.orders" on that same compound key.

MongoDB

sh.enableSharding("ecommerce")

// Create the compound index, then shard the collection
// Your code here

Need a hint?

The shard key must have a supporting index. Create the index first with db.orders.createIndex(), then shard with sh.shardCollection().

3

Check shard distribution

Use db.orders.getShardDistribution() to check how data is distributed across the shards.

MongoDB

sh.enableSharding("ecommerce")
db.orders.createIndex({ "region": 1, "orderId": 1 })
sh.shardCollection("ecommerce.orders", { "region": 1, "orderId": 1 })

// Check how data is distributed across shards
// Your code here

Need a hint?

Use db.orders.getShardDistribution() to see the percentage of data on each shard.

4

Configure balancer window

Set the balancer to only run during off-peak hours. Use db.settings.updateOne() on the config database to set the balancer's activeWindow with start as "02:00" and stop as "06:00". Filter by { _id: "balancer" } and use $set.

MongoDB

// Set the balancer active window to 2 AM - 6 AM
// Your code here

Need a hint?

Update the balancer settings in db.settings with an activeWindow object containing start and stop times.