Write concern and data durability in MongoDB - Step-by-Step Execution

Concept Flow - Write concern and data durability

Client sends write request

↓

MongoDB receives write

↓

Apply write to primary

↓

Check write concern level

↓

w:0

↓

No ack

↓

Confirm durability

↓

Send response to client

This flow shows how MongoDB processes a write request and confirms it based on the write concern level, ensuring data durability as requested.

Execution Sample

MongoDB

db.collection.insertOne({name: "Alice"}, {writeConcern: {w: "majority"}})

Insert a document with write concern set to majority, waiting for confirmation from most nodes.

Execution Table

Step	Action	Write Concern Level	Result	Client Response
1	Client sends insertOne request	majority	Request received by primary	No response yet
2	Primary applies write	majority	Write applied on primary	No response yet
3	Primary waits for replicas	majority	Replicas apply write	No response yet
4	Majority of nodes confirm write	majority	Write durable on majority	Acknowledgement sent
5	Client receives acknowledgement	majority	Write confirmed durable	Success response

💡 Write concern majority met, client receives confirmation, execution ends.

Variable Tracker

Variable	Start	After Step 2	After Step 3	After Step 4	Final
writeAppliedOnPrimary	false	true	true	true	true
writeAppliedOnReplicas	false	false	true	true	true
writeDurable	false	false	false	true	true
clientAcknowledged	false	false	false	true	true

Key Moments - 2 Insights

Why does the client not get a response immediately after the primary applies the write?

Because with write concern 'majority', MongoDB waits for most replicas to confirm the write before sending acknowledgement, as shown in execution_table step 3 and 4.

What happens if writeConcern is set to w:0?

The client gets no acknowledgement at all; the write is sent but the client does not wait for any confirmation, as indicated in the concept_flow branch for w:0.

Visual Quiz - 3 Questions

Look at the execution_table, at which step does the write become durable on the majority of nodes?

A Step 4

B Step 3

C Step 2

D Step 5

💡 Hint

Check the 'Result' column for when 'Write durable on majority' occurs.

According to variable_tracker, when does clientAcknowledged change to true?

A After Step 5

B After Step 3

C After Step 4

D After Step 2

💡 Hint

Look at the 'clientAcknowledged' row and see when it becomes true.

If writeConcern was set to w:1, how would the client response timing change compared to w:majority?

A Client waits for all replicas

B Client gets acknowledgement after primary applies write

C Client never gets acknowledgement

D Client waits for no nodes

💡 Hint

Refer to concept_flow branches for different write concern levels.

Concept Snapshot

Write concern controls how MongoDB confirms writes.
'w:0' means no acknowledgement.
'w:1' waits for primary only.
'w:majority' waits for most nodes.
Higher write concern means better durability but slower response.

Full Transcript

This visual execution shows how MongoDB handles write concern and data durability. The client sends a write request with a specified write concern level. The primary node applies the write and, depending on the write concern, waits for replicas to confirm the write. For 'majority', the client only receives acknowledgement after most nodes confirm the write, ensuring durability. Variables track the write application and acknowledgement status step-by-step. Key moments clarify why acknowledgements are delayed and what happens with different write concern levels. Quizzes test understanding of when writes become durable and when clients get responses.

Practice

1. What does the writeConcern option in MongoDB control?

easy

A. How MongoDB confirms that data is saved to the database

B. The format of data stored in MongoDB

C. The speed of data retrieval from MongoDB

D. The size limit of documents in MongoDB

Solution

1. Step 1: Understand the role of write concern

   Write concern tells MongoDB how to confirm that a write operation is saved and durable.

2. Step 2: Differentiate from other options

   It does not control data format, retrieval speed, or document size.

3. Final Answer:

   How MongoDB confirms that data is saved to the database -> Option A

4. Quick Check:

   Write concern = data save confirmation [OK]

Hint: Write concern = confirmation of data saved [OK]

Common Mistakes:

• Confusing write concern with data format
• Thinking it controls read speed
• Assuming it limits document size

2. Which of the following is the correct way to specify a write concern that waits for acknowledgment from the primary in MongoDB?

easy

A. { writeConcern: { w: -1 } }

B. { writeConcern: { w: 0 } }

C. { writeConcern: { w: 'majority' } }

D. { writeConcern: { w: 1 } }

Solution

1. Step 1: Recall write concern values

   w:1 means wait for acknowledgment from the primary node only.

2. Step 2: Check other options

   w:0 means no acknowledgment; w:'majority' waits for majority nodes; w:-1 is invalid.

3. Final Answer:

   { writeConcern: { w: 1 } } -> Option D

4. Quick Check:

   w:1 = primary ack wait [OK]

Hint: w:1 waits for primary acknowledgment [OK]

Common Mistakes:

• Using w:0 which means no acknowledgment
• Confusing 'majority' with primary only
• Using invalid negative values

3. Consider this MongoDB insert command with write concern:

db.collection.insertOne({name: 'Alice'}, {writeConcern: {w: 'majority', wtimeout: 5000}})

What happens if the majority of nodes do not acknowledge the write within 5 seconds?

medium

A. The write operation succeeds immediately without waiting

B. The write operation retries indefinitely until acknowledged

C. The write operation fails with a timeout error

D. The write operation ignores the write concern and writes anyway

Solution

1. Step 1: Understand w:'majority' and wtimeout

   The write waits for majority nodes to confirm within 5000 ms (5 seconds).

2. Step 2: Outcome if timeout exceeded

   If majority nodes don't respond in time, MongoDB returns a timeout error and the write is considered failed.

3. Final Answer:

   The write operation fails with a timeout error -> Option C

4. Quick Check:

   w:'majority' + wtimeout = fail if no timely ack [OK]

Hint: wtimeout limits wait; timeout causes failure [OK]

Common Mistakes:

• Assuming write succeeds immediately
• Thinking MongoDB retries forever
• Ignoring write concern on timeout

4. You wrote this code to insert a document with write concern:

db.collection.insertOne({item: 'book'}, {writeConcern: {w: 2, wtimeout: '3000'}})

But it throws an error. What is the problem?

medium

A. wtimeout must be a number, not a string

B. w:2 is invalid; only 0,1, or 'majority' allowed

C. insertOne does not accept writeConcern option

D. The document field 'item' is not allowed

Solution

1. Step 1: Check writeConcern syntax

   wtimeout expects a number (milliseconds), not a string.

2. Step 2: Validate other parts

   w:2 is valid if the replica set has at least 2 nodes; insertOne accepts writeConcern; 'item' field is valid.

3. Final Answer:

   wtimeout must be a number, not a string -> Option A

4. Quick Check:

   wtimeout type must be number [OK]

Hint: wtimeout needs number, not quotes [OK]

Common Mistakes:

• Using string for wtimeout
• Thinking w:2 is invalid always
• Assuming insertOne disallows writeConcern

5. You want to ensure that a write is saved to disk on the primary and replicated to at least two secondary nodes before confirming success. Which write concern setting achieves this?

hard

A. { writeConcern: { w: 1, j: true } }

B. { writeConcern: { w: 3, j: true } }

C. { writeConcern: { w: 'majority', j: false } }

D. { writeConcern: { w: 2, j: false } }

Solution

1. Step 1: Understand w and j options

   w:3 means wait for acknowledgment from primary plus two secondaries (total 3 nodes). j:true means wait for write to be committed to journal (disk) on primary.

2. Step 2: Check other options

   B waits only for primary with journaling; C waits for majority but no journaling; D waits for 2 nodes but no journaling.

3. Final Answer:

   { writeConcern: { w: 3, j: true } } -> Option B

4. Quick Check:

   w:3 + j:true = 3 nodes + disk write [OK]

Hint: w:number = nodes ack; j:true = disk commit [OK]

Common Mistakes:

• Ignoring journaling for durability
• Using w:'majority' without knowing node count
• Assuming w:2 means two secondaries only