TTL indexes for auto-expiry in MongoDB - Step-by-Step Execution

Concept Flow - TTL indexes for auto-expiry

Create TTL Index on Date Field

↓

MongoDB Monitors Documents

↓

Check Document Date + TTL Value

↓

Not Expired

↓

Keep Document

↓

Repeat Continuously

MongoDB uses a TTL index on a date field to automatically delete documents after a set time. It continuously checks documents and removes expired ones.

Execution Sample

MongoDB

db.sessions.createIndex({"lastAccess": 1}, {expireAfterSeconds: 3600})

Creates a TTL index on the 'lastAccess' field to auto-delete documents 1 hour after the timestamp.

Execution Table

Step	Document _id	lastAccess	Current Time	Expired?	Action
1	doc1	2024-06-01T10:00:00Z	2024-06-01T10:30:00Z	No	Keep document
2	doc2	2024-06-01T09:00:00Z	2024-06-01T10:30:00Z	Yes	Delete document
3	doc3	2024-06-01T10:15:00Z	2024-06-01T10:30:00Z	No	Keep document
4	doc1	2024-06-01T10:00:00Z	2024-06-01T11:01:00Z	Yes	Delete document
5	doc3	2024-06-01T10:15:00Z	2024-06-01T11:16:00Z	Yes	Delete document
6	-	-	2024-06-01T11:16:00Z	-	No more documents to check

💡 All documents expired or kept based on TTL; process repeats continuously.

Variable Tracker

Document _id	Initial lastAccess	After Step 1	After Step 4	Final
doc1	2024-06-01T10:00:00Z	Exists	Deleted	Deleted
doc2	2024-06-01T09:00:00Z	Deleted	Deleted	Deleted
doc3	2024-06-01T10:15:00Z	Exists	Exists	Deleted

Key Moments - 2 Insights

Why does MongoDB delete documents only after the TTL time has passed, not exactly at the moment?

MongoDB checks for expired documents periodically, not continuously every second. This is why deletion happens shortly after the TTL time, as shown in steps 2 and 4 in the execution table.

Can TTL indexes be used on any field type?

No, TTL indexes only work on fields that store date values. This is why the 'lastAccess' field must be a date, as shown in the execution table.

Visual Quiz - 3 Questions

Look at the execution table, at step 2, why is 'doc2' deleted?

A Because it was manually deleted

B Because it has the newest lastAccess date

C Because its lastAccess plus TTL is before the current time

D Because TTL indexes delete all documents immediately

💡 Hint

Check the 'Expired?' column at step 2 in the execution table.

At which step does 'doc1' get deleted according to the execution table?

A Step 4

B Step 1

C Step 2

D Step 6

💡 Hint

Look for 'doc1' with 'Delete document' action in the execution table.

If the TTL was set to 1800 seconds (30 minutes) instead of 3600, how would the action for 'doc3' at step 3 change?

A It would be deleted at step 3

B It would still be kept at step 3

C It would be deleted at step 1

D It would never be deleted

💡 Hint

Compare the TTL value with the time difference between lastAccess and current time at step 3.

Concept Snapshot

TTL Indexes in MongoDB:
- Created on a date field with expireAfterSeconds option.
- MongoDB periodically checks documents.
- Documents older than (date + TTL) are deleted automatically.
- Only works on date fields.
- Helps auto-remove expired data without manual cleanup.

Full Transcript

This visual execution shows how MongoDB TTL indexes work to auto-delete documents after a set time. First, a TTL index is created on a date field, like 'lastAccess'. MongoDB then periodically checks each document's date plus the TTL value against the current time. If the document is expired, it is deleted; otherwise, it is kept. The execution table traces documents with different lastAccess times and shows when they are deleted or kept. Key points include that TTL checks happen periodically, not instantly, and TTL indexes only work on date fields. This helps keep collections clean by removing old data automatically.