Mongos router behavior in MongoDB - Step-by-Step Execution

Concept Flow - Mongos router behavior

Client sends query

↓

Mongos receives query

↓

Mongos checks shard key

↓

Shard key present?

No→Broadcast query to all shards

↓

Route query to specific shard

↓

Collect results from shards

↓

Merge and return results to client

Mongos receives queries, checks if shard key is present, routes to specific shard or broadcasts to all, then merges results and returns to client.

Execution Sample

MongoDB

db.collection.find({shardKey: 5, status: 'active'})

Query with shard key routes to one shard; results returned from that shard only.

Execution Table

Step	Action	Shard Key Present?	Routing Decision	Shards Queried	Result Collected
1	Client sends query	Yes	Check shard key	N/A	N/A
2	Mongos receives query	Yes	Route to shard with shardKey=5	Shard 2	N/A
3	Shard 2 executes query	N/A	N/A	Shard 2	Documents matching query
4	Mongos collects results	N/A	N/A	Shard 2	Documents from Shard 2
5	Mongos returns results to client	N/A	N/A	N/A	Final result set

💡 Query routed to shard based on shard key; results returned from that shard only.

Variable Tracker

Variable	Start	After Step 2	After Step 4	Final
query	{shardKey:5, status:'active'}	{shardKey:5, status:'active'}	{shardKey:5, status:'active'}	{shardKey:5, status:'active'}
shardKeyPresent	undefined	true	true	true
targetShards	undefined	Shard 2	Shard 2	Shard 2
results	empty	empty	documents from Shard 2	documents from Shard 2

Key Moments - 2 Insights

Why does Mongos route the query to only one shard when the shard key is present?

Because the shard key identifies the exact shard holding the data, Mongos routes the query directly to that shard (see execution_table step 2). This avoids querying all shards and improves efficiency.

What happens if the shard key is missing in the query?

Mongos broadcasts the query to all shards since it cannot determine which shard holds the data. This is less efficient but necessary to find matching documents.

Visual Quiz - 3 Questions

Look at the execution_table at step 2, what is the routing decision Mongos makes?

A Route query to shard with shardKey=5

B Broadcast query to all shards

C Reject the query

D Cache the query for later

💡 Hint

Check the 'Routing Decision' column at step 2 in the execution_table.

At which step does Mongos collect results from shards?

A Step 3

B Step 4

C Step 1

D Step 5

💡 Hint

Look for the 'Collect results from shards' action in the execution_table.

If the query did not include the shard key, how would the 'Shards Queried' column change at step 2?

A It would list only one shard

B It would be empty

C It would list all shards

D It would list no shards

💡 Hint

Recall that without shard key, Mongos broadcasts query to all shards.

Concept Snapshot

Mongos router receives client queries.
Checks if shard key is present in query.
If yes, routes query to specific shard.
If no, broadcasts query to all shards.
Collects and merges results before returning to client.

Full Transcript

Mongos acts as a router in a MongoDB sharded cluster. When a client sends a query, Mongos first checks if the query contains the shard key. If the shard key is present, Mongos routes the query directly to the shard that holds the relevant data. This targeted routing improves efficiency by querying only one shard. If the shard key is missing, Mongos broadcasts the query to all shards to find matching documents. After shards execute the query, Mongos collects and merges the results, then returns the final result set to the client. This behavior ensures efficient and correct query routing in a sharded environment.