Why PERSIST to remove expiry in Redis? - Purpose & Use Cases

The Big Idea

What if you could save your Redis data from disappearing with just one simple command?

The Scenario

Imagine you set a timer on your phone to remind you about an important task, but then you realize you want to keep the reminder indefinitely. Without a simple way to remove the timer, you'd have to delete and recreate the reminder manually every time.

The Problem

Manually tracking and resetting expiry times on keys in Redis is slow and error-prone. You might accidentally lose important data if the expiry is not removed correctly, or waste time constantly resetting timers.

The Solution

The PERSIST command in Redis lets you quickly remove the expiry from a key, turning it into a permanent entry without needing to delete or recreate it. This keeps your data safe and saves you time.

Before vs After

✗ Before

DEL mykey
SET mykey value

✓ After

PERSIST mykey

What It Enables

You can easily keep important data forever in Redis by removing expiry with a single command.

Real Life Example

Suppose you have a session token that you initially want to expire after 30 minutes, but the user logs in again and you want to keep their session active indefinitely. Using PERSIST lets you remove the expiry without losing the session data.

Key Takeaways

Manual expiry removal is slow and risky.

PERSIST removes expiry instantly and safely.

This keeps your important Redis data permanent when needed.

Practice

1. What does the Redis command PERSIST key do?

easy

A. Renames the key to 'persisted_key'

B. Sets a new expiration time on the key

C. Deletes the key immediately

D. Removes the expiration time from the key, making it permanent

Solution

1. Step 1: Understand the purpose of PERSIST

   The PERSIST command removes any expiration time set on a Redis key, so the key no longer expires.

2. Step 2: Compare with other commands

   Unlike EXPIRE which sets a timeout, PERSIST cancels it. It does not delete or rename the key.

3. Final Answer:

   Removes the expiration time from the key, making it permanent -> Option D

4. Quick Check:

   PERSIST removes expiry = Removes the expiration time from the key, making it permanent [OK]

Hint: PERSIST cancels expiry, keeps key forever [OK]

Common Mistakes:

• Confusing PERSIST with EXPIRE
• Thinking PERSIST deletes the key
• Assuming PERSIST renames the key

2. Which of the following is the correct syntax to remove expiry from a Redis key named session123?

easy

A. PERSIST session123

B. REMOVEEXPIRY session123

C. EXPIRE session123 0

D. TTL session123

Solution

1. Step 1: Recall the correct command for removing expiry

   The Redis command to remove expiry is PERSIST key. So for key 'session123', it is PERSIST session123.

2. Step 2: Check other options

   REMOVEEXPIRY is not a Redis command. EXPIRE with 0 does not remove expiry but sets it to expire immediately. TTL only checks expiry time.

3. Final Answer:

   PERSIST session123 -> Option A

4. Quick Check:

   Correct syntax for removing expiry = PERSIST session123 [OK]

Hint: Use PERSIST key to remove expiry [OK]

Common Mistakes:

• Using EXPIRE with 0 to remove expiry
• Trying non-existent commands like REMOVEEXPIRY
• Confusing TTL with expiry removal

3. Given the following Redis commands executed in order:

SET mykey "hello"
EXPIRE mykey 10
TTL mykey
PERSIST mykey
TTL mykey

What will be the output of the two TTL commands?

medium

A. 10 and -1

B. -1 and 10

C. 10 and 10

D. -1 and -1

Solution

1. Step 1: Analyze the first TTL after EXPIRE

   After setting key 'mykey' and applying EXPIRE 10, TTL returns the remaining seconds, which is 10.

2. Step 2: Analyze the second TTL after PERSIST

   PERSIST removes expiry, so TTL returns -1 indicating no expiration.

3. Final Answer:

   10 and -1 -> Option A

4. Quick Check:

   TTL before persist = 10, after persist = -1 [OK]

Hint: TTL shows seconds left; PERSIST makes TTL -1 [OK]

Common Mistakes:

• Expecting TTL to stay 10 after PERSIST
• Confusing TTL -1 meaning
• Thinking PERSIST sets TTL to 0

4. You run the command PERSIST mykey but it returns 0. What is the most likely reason?

medium

A. The key does not exist

B. The key has no expiry set

C. The key is already persistent

D. The command syntax is incorrect

Solution

1. Step 1: Understand PERSIST return values

   PERSIST returns 1 if expiry was removed, 0 if key had no expiry or key does not exist.

2. Step 2: Differentiate between no expiry and missing key

   If key does not exist, PERSIST returns 0, but usually you check key existence first. If key exists but no expiry, PERSIST returns 0 because nothing to remove.

3. Final Answer:

   The key has no expiry set -> Option B

4. Quick Check:

   PERSIST returns 0 if no expiry to remove [OK]

Hint: PERSIST returns 0 if no expiry on key [OK]

Common Mistakes:

• Assuming 0 means syntax error
• Confusing no expiry with key missing
• Thinking 0 means key deleted

5. You have a Redis key cache:item with an expiry of 30 seconds. You want to keep it permanently only if its current TTL is more than 10 seconds. Which sequence of commands achieves this safely?

hard

A. Run EXPIRE cache:item 0 to remove expiry

B. Run PERSIST cache:item directly without checking TTL

C. Use TTL cache:item, then if result > 10, run PERSIST cache:item

D. Rename the key to a new key without expiry

Solution

1. Step 1: Check current TTL of the key

   Use TTL cache:item to get remaining expiry time in seconds.

2. Step 2: Conditionally remove expiry

   If TTL is more than 10, run PERSIST cache:item to remove expiry and keep key permanently.

3. Step 3: Avoid wrong commands

   Running PERSIST without checking TTL removes expiry regardless. EXPIRE with 0 does not remove expiry but sets immediate expiry. Renaming does not affect expiry.

4. Final Answer:

   Use TTL, then if > 10, run PERSIST -> Option C

5. Quick Check:

   Check TTL before PERSIST to keep key only if TTL > 10 [OK]

Hint: Check TTL first, then PERSIST if TTL > 10 [OK]

Common Mistakes:

• Running PERSIST blindly without TTL check
• Using EXPIRE 0 to remove expiry
• Thinking renaming removes expiry