TTL behavior and timing in DynamoDB - Time & Space Complexity

Time Complexity: TTL behavior and timing

O(n)

Understanding Time Complexity

We want to understand how the time it takes for DynamoDB to remove expired items grows as more items have TTL set.

How does the system handle many items expiring, and how does that affect timing?

Scenario Under Consideration

Analyze the time complexity of DynamoDB's TTL expiration process.

// DynamoDB TTL process (conceptual):
// 1. Scan table for items with TTL attribute set and expired.
// 2. Mark expired items for deletion.
// 3. Delete expired items asynchronously.
// 4. Repeat periodically.
    

This process checks items with TTL timestamps and deletes those expired in batches over time.

Identify Repeating Operations

Look at what repeats during TTL expiration.

• Primary operation: Scanning items to find expired ones.
• How many times: Once per periodic TTL check, scanning many items.

How Execution Grows With Input

As the number of items with TTL grows, the scan to find expired items takes longer.

Input Size (n)	Approx. Operations
10	Scan 10 items for expiration
100	Scan 100 items for expiration
1000	Scan 1000 items for expiration

Pattern observation: The time to find expired items grows roughly in proportion to the number of items checked.

Final Time Complexity

Time Complexity: O(n)

This means the time to process TTL expiration grows linearly with the number of items having TTL.

Common Mistake

[X] Wrong: "TTL expiration happens instantly for each item as soon as it expires."

[OK] Correct: TTL expiration is a background process that scans items periodically, so there can be a delay and the time depends on how many items need checking.

Interview Connect

Understanding how background expiration processes scale helps you reason about system performance and delays in real-world databases.

Self-Check

"What if DynamoDB used an index to track only expired items instead of scanning all? How would the time complexity change?"

Practice

1. What does TTL (Time To Live) do in DynamoDB?

easy

A. Encrypts data stored in the table.

B. Backs up data periodically.

C. Automatically deletes items after their expiration time.

D. Indexes data for faster queries.

Solution

1. Step 1: Understand TTL purpose

   TTL is designed to remove expired data automatically to save storage and keep data fresh.

2. Step 2: Identify TTL behavior

   It deletes items after their expiration timestamp passes, not encrypting or backing up data.

3. Final Answer:

   Automatically deletes items after their expiration time. -> Option C

4. Quick Check:

   TTL = automatic deletion [OK]

Hint: TTL means automatic deletion after expiry [OK]

Common Mistakes:

• Thinking TTL encrypts data
• Confusing TTL with backup
• Assuming TTL speeds up queries

2. Which attribute type must the TTL attribute have in DynamoDB?

easy

A. String representing a date in ISO format

B. Number representing Unix epoch time in seconds

C. Boolean indicating if item is expired

D. List of timestamps

Solution

1. Step 1: Recall TTL attribute format

   The TTL attribute must be a number storing the expiration time as Unix epoch seconds.

2. Step 2: Eliminate other types

   Strings, booleans, or lists are not valid for TTL expiration timestamps.

3. Final Answer:

   Number representing Unix epoch time in seconds -> Option B

4. Quick Check:

   TTL attribute = Unix epoch seconds [OK]

Hint: TTL attribute stores expiration as Unix epoch seconds [OK]

Common Mistakes:

• Using string dates instead of numbers
• Using boolean instead of timestamp
• Trying to store multiple timestamps

3. If an item has a TTL attribute set to 1650000000 (Unix epoch seconds), when will DynamoDB delete it?

medium

A. Within 48 hours after 1650000000 seconds

B. Exactly at 1650000000 seconds

C. Immediately after setting the TTL attribute

D. Only when manually deleted

Solution

1. Step 1: Understand TTL deletion timing

   DynamoDB deletes expired items approximately after the TTL timestamp, not exactly at that second.

2. Step 2: Recall deletion delay

   Deletion can take up to 48 hours after the expiration time to occur.

3. Final Answer:

   Within 48 hours after 1650000000 seconds -> Option A

4. Quick Check:

   TTL deletion delay = up to 48 hours [OK]

Hint: TTL deletion can be delayed up to 48 hours [OK]

Common Mistakes:

• Expecting deletion exactly at expiration time
• Thinking deletion is immediate
• Assuming manual deletion is required

4. You set TTL on an attribute but notice expired items are not deleted after 3 days. What is a likely cause?

medium

A. TTL attribute is not a number with Unix epoch seconds

B. TTL is enabled but the table is in a different AWS region

C. You forgot to enable encryption on the table

D. The table has no primary key

Solution

1. Step 1: Check TTL attribute format

   If the TTL attribute is not a number with Unix epoch seconds, DynamoDB ignores it for expiration.

2. Step 2: Verify other options

   Region difference, encryption, or primary key absence do not prevent TTL deletion.

3. Final Answer:

   TTL attribute is not a number with Unix epoch seconds -> Option A

4. Quick Check:

   TTL attribute format error blocks deletion [OK]

Hint: TTL attribute must be Unix epoch number [OK]

Common Mistakes:

• Assuming region affects TTL
• Thinking encryption affects TTL
• Ignoring TTL attribute data type

5. You want to ensure items expire exactly after 24 hours using TTL. Which approach is best considering DynamoDB's TTL behavior?

hard

A. Disable TTL and run a scheduled job to delete expired items

B. Set TTL attribute to current time + 86400 seconds and manually delete items after 24 hours

C. Use a string date format for TTL attribute for precise timing

D. Set TTL attribute to current time + 86400 seconds and accept up to 48-hour deletion delay

Solution

1. Step 1: Understand TTL timing limits

   DynamoDB TTL deletes items approximately after expiration, with up to 48 hours delay.

2. Step 2: Choose best practical approach

   Setting TTL attribute to current time + 86400 seconds (24 hours) is correct; manual deletion adds complexity and string format is invalid.

3. Step 3: Consider alternatives

   Disabling TTL and scheduling deletes is possible but less efficient than TTL's automatic deletion.

4. Final Answer:

   Set TTL attribute to current time + 86400 seconds and accept up to 48-hour deletion delay -> Option D

5. Quick Check:

   TTL delay means accept approximate expiry [OK]

Hint: TTL deletion timing is approximate, plan accordingly [OK]

Common Mistakes:

• Expecting exact deletion timing
• Using string instead of number for TTL
• Disabling TTL unnecessarily