Caching Strategies

"There are only two hard things in Computer Science: cache invalidation and naming things."

If you cache a user's profile and they change their name, but your cache still shows the old name, your app is "broken" in the user's eyes. This is called Stale Data. We use different strategies to prevent this.

1. TTL (Time to Live)

This is the simplest and most common strategy. You give every piece of data an "expiry date."

• How it works: When you save data to the cache (like Redis), you set a timer (e.g., 3600 seconds). Once the timer hits zero, the cache automatically deletes that data.
• Best For: Data that changes occasionally but isn't "mission critical" (like weather updates or trending posts).

2. Write Strategies

How does the data get into the cache in the first place?

🏹 Cache Aside (Lazy Loading)

The "Check First" Method

The application code is responsible for managing the cache.

1. Check cache.
2. If found (Hit), return.
3. If not found (Miss), get from DB and then save to cache.

• Pros: Only requested data is cached.
• Cons: The first user always experiences a "slow" request (Cache Miss).

✍️ Write Through

The "Always Sync" Method

Data is written to the cache and the database at the same time.

• Pros: Cache is never stale; data is always ready.
• Cons: Writing data takes slightly longer because you are saving it in two places.

3. Cache Eviction (The Cleanup)

What happens when your cache (RAM) is full? You can't just stop saving data. You have to kick something out.

• LRU (Least Recently Used): The most popular strategy. It deletes the data that hasn't been touched in the longest time.
• LFU (Least Frequently Used): Deletes data that is requested the least number of times.
• FIFO (First In, First Out): Deletes the oldest data regardless of how often it's used.

4. Cache Invalidation (The "Kill" Switch)

This is the process of manually removing data from the cache because you know the source data has changed.

Example at CodeHarborHub: A student updates their "Bio."

1. Your code updates the Database.
2. Your code immediately runs cache.delete('user_101_bio').
3. The next time the user visits their profile, the app sees a Cache Miss and fetches the new bio from the DB.

Summary Checklist

• I understand that TTL is a timer that automatically clears old data.
• I know that Cache Aside is the most common way to load data.
• I understand that LRU helps manage limited memory by deleting unused data.
• I recognize that manual Invalidation is necessary when data is updated.

The 80/20 Rule

In most applications, 80% of your traffic comes from 20% of your data. Focus your caching efforts on that 20% (like top-selling products or homepage banners) to get the biggest performance boost!