Why should you care?

Without mongos, clients would need to know chunk maps. The router hides that complexity — like a smart load-aware proxy for data.

See it live — copy this example

Open mongosh or MongoDB Compass, select database nosqlverse, then run the example. Change one field and run again.

// Client URI points at mongos host(s)
// mongodb://mongos1:27017,mongos2:27017/NoSQLVerse
db.hello()
db.adminCommand({ listShards: 1 })
db.readings.find({ deviceId: "sensor-42" }).explain("executionStats")

Run Example »

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

What happened?

• db.hello from mongos shows you are on a mongos.
• listShards shows backends.
• explain reveals whether the query targeted one shard or scattered.
• Always give apps multiple mongos hosts for HA.

Practice next

1. Connect mongosh to mongos, not a shard.
2. Run listShards.
3. explain a query that includes the shard key vs one that does not.
4. Compare explain shards for targeted vs scatter query.
5. Add a second mongos in a lab topology.

Remember

mongos routes sharded queries. Clients connect to mongos. Targeted queries need the shard key.

App tier → mongos pool

Kubernetes pods connect to two mongos services for NoSQLVerse production.

Outcome: Router restarts do not take the API fully down.