Horizontal Scaling in MongoDB?

Horizontal scaling (or scaling out) is the process of increasing database capacity by adding more servers instead of upgrading a single server. In MongoDB, sharding is the method used for horizontal scaling.

Implemention Of Horizontal Scaling

Follow the steps below to implement sharding in MongoDB successfully.

Step 1: Set Up the Environment

Before starting, ensure you have multiple servers available to act as shards, config servers, and query routers.

Step 2: Install MongoDB of All Servers

Install MongoDB on all servers that will act as shards, config servers, and query routers.

Step 3: Configure Shards

Start each MongoDB instance that will be a shard with the appropriate configuration. For example:

mongod --shardsvr --port 27018 --dbpath /data/shard1 --logpath /var/log/mongodb/shard1.log --fork

Repeat the process for all shard servers, ensuring each one is configured correctly.

Step 4: Configure Config Servers

Start the config servers with the appropriate configuration. Ensure you have at least three config servers for redundancy.

mongod --configsvr --port 27019 --dbpath /data/config --logpath /var/log/mongodb/config.log --fork

Repeat the process for all config servers.

Step 5: Start the Query Routers (Mongos)

Start the Mongos instances to act as query routers:

mongos --configdb configReplSet/localhost:27019,localhost:27020,
localhost:27021 --logpath /var/log/mongodb/mongos.log --fork

Ensure the configDB setting correctly points to the config servers.

Step 6: Connect to the Mongos

Connect to the mongos instance using the MongoDB shell:

mongo --port 27017

Step 7: Add Shards to the Cluster

Within the MongoDB shell, add each shard to the cluster:

sh.addShard("shard1/localhost:27018")
sh.addShard("shard2/localhost:27019")
sh.addShard("shard3/localhost:27020")

Step 8: Enable Sharding for a Database

Enable sharding for a specific database:

sh.enableSharding("myDatabase") 

Step 9: Shard a Collection Using a Shard Key

Shard a collection within the database by specifying a shard key. Choosing the right shard key is critical for even data distribution and query efficiency.

sh.shardCollection("myDatabase.myCollection", { shardKey: 1 })

Why Horizontal Scaling is Needed

• Handle Large Data: Efficiently manage large datasets and high-traffic workloads.
• Prevent Bottlenecks: Avoid overloading a single server by distributing data and queries.
• Improve Performance: Reads and writes are distributed across multiple servers.
• High Availability: Combined with replica sets, sharding ensures fault tolerance.