MongoDB Sharding Cluster Setup - Complete Beginner Tutorial

Understanding MongoDB Sharded Cluster Components

A MongoDB sharded cluster is composed of three main components: shard servers, config servers, and mongos query routers.

Each component plays a critical role in distributing data and managing cluster metadata.

• Shard Servers: Store the actual data and handle read/write operations for their portion of the dataset.
• Config Servers: Maintain metadata and configuration settings for the cluster.
• Mongos Routers: Interface between client applications and the sharded cluster, routing queries to appropriate shards.

Setting Up Shard Servers

Shard servers hold subsets of the data. Each shard can be a single mongod instance or a replica set for high availability.

For production environments, it is recommended to deploy each shard as a replica set.

• Install MongoDB on each shard server.
• Configure each shard as a replica set with a unique replica set name.
• Start the mongod process with the --shardsvr option to designate it as a shard server.

Configuring Config Servers

Config servers store metadata about the cluster’s structure and the location of data chunks.

A minimum of three config servers is required for production to ensure redundancy and fault tolerance.

• Deploy config servers as a replica set for high availability.
• Start mongod instances with the --configsvr option.
• Ensure config servers are accessible to all shard servers and mongos routers.

Deploying Mongos Query Routers

Mongos acts as the interface between client applications and the sharded cluster.

It routes queries to the appropriate shards based on cluster metadata.

• Install mongos on one or more servers accessible to your application.
• Configure mongos with the addresses of the config servers.
• Start mongos processes to enable query routing.

Adding Shards to the Cluster

Once shard servers, config servers, and mongos routers are running, you can add shards to the cluster.

Use the mongos shell to connect and run commands to add shards.

• Connect to mongos using the mongo shell.
• Run the command: sh.addShard("replicaSetName/host1:port,host2:port,...") to add each shard.
• Verify shards are added with sh.status().

Example: Basic Sharded Cluster Setup

Here is a simplified example of commands to start a shard server, config server, and mongos router.

• Start a shard server: mongod --shardsvr --replSet shard1 --dbpath /data/shard1 --port 27018
• Start a config server: mongod --configsvr --replSet configReplSet --dbpath /data/config --port 27019
• Start mongos: mongos --configdb configReplSet/localhost:27019 --port 27017