Master MongoDB Sharding: Step-by-Step Setup and Verification

Data Sharding

Why Sharding? Sharding is MongoDB's method for splitting large collections across multiple servers (a cluster) when a single server's CPU, memory, or I/O cannot meet demand, thereby reducing server pressure.

Sharding Architecture

Experimental Topology

Configuration Process

Clean Existing Data

[root@node1 ~]# service mongod stop

Stopping mongod:                               [  OK  ]

[root@node1 ~]# rm -rf /mongodb/data/*

Execute the above on each node; skip if this is the first run.

Config Server Setup

Install required packages:

[root@node2 mongodb]# ls

mongodb-org-server-2.6.10-1.x86_64.rpm  mongodb-org-tools-2.6.10-1.x86_64.rpm

mongodb-org-shell-2.6.10-1.x86_64.rpm

[root@node2 mongodb]# yum install *.rpm -y

Modify Configuration File [root@node2 ~]# vim /etc//mongod.conf Set the following options:

logpath=/var/log/mongodb/mongod.log
logappend=true
fork=true
dbpath=/mongodb/data
configsvr=true   # enable config server
pidfilepath=/var/run/mongodb/mongod.pid
#bind_ip=127.0.0.1
httpinterface=true
rest=true

Create Data Directory

[root@node2 ~]# mkdir /mongodb/data -pv

mkdir: created directory `/mongodb'

mkdir: created directory `/mongodb/data'

[root@node2 ~]# chown -R mongod.mongod /mongodb

Start Service

[root@node2 ~]# service mongod start

Starting mongod:                               [  OK  ]

[root@node2 ~]# ss -tnl | grep 27019

LISTEN      0      128               *:27019                *

Note: the listening port has changed from the default 27017.

Mongos Configuration

Install only the mongos package:

[root@node1 ~]# yum install mongodb-org-mongos-2.6.10-1.x86_64.rpm -y

Start mongos:

[root@node1 ~]# mongos --configdb=172.16.10.124 --fork --logpath=/var/log/mongodb/mongos.log

Output indicates successful fork and background execution.

Shard Configuration

Modify each shard's /etc/mongod.conf similarly to the config server (set configsvr=true for config server only).

[root@node3 ~]# vim /etc/mongod.conf

logpath=/var/log/mongodb/mongod.log
logappend=true
fork=true
dbpath=/mongodb/data
pidfilepath=/var/run/mongodb/mongod.pid
#bind_ip=127.0.0.1
httpinterface=true
rest=true

[root@node3 ~]# service mongod start

Starting mongod:                               [  OK  ]

Repeat on both shard nodes.

Add Shard Nodes to Mongos

[root@node1 ~]# mongo --host 172.16.10.123

mongos> sh.addShard("172.16.10.125")

{ "shardAdded" : "shard0000", "ok" : 1 }

mongos> sh.addShard("172.16.10.126")

{ "shardAdded" : "shard0001", "ok" : 1 }

mongos> sh.status()

Shows the two shards and their hosts.

Enable Sharding on a Database

mongos> sh.enableSharding("testdb")

{ "ok" : 1 }

Shard a Collection

mongos> sh.shardCollection("testdb.students", {"age": 1})

{ "collectionsharded" : "testdb.students", "ok" : 1 }

Insert Test Data and Verify Distribution

mongos> use testdb

mongos> for (i=1;i<=100000;i++) db.students.insert({name:"student"+i, age:(i%120), addr:"china"})

WriteResult({ "nInserted" : 1 })

mongos> db.students.find().count()

100000

mongos> sh.status()

The status output shows chunks distributed across shard0000 and shard0001, confirming successful sharding.

Conclusion

MongoDB replication sets and data sharding provide superior performance and simpler operations compared to MySQL for large‑scale data handling. Although MongoDB is still evolving and has open issues, its capabilities make it a strong choice for future projects.