# Cache-Aside Pattern (Lazy Loading)

Use Cache-Aside for read-heavy workloads: on cache miss, fetch from the database and populate the cache; on write, invalidate or update the cache explicitly.

The application treats the cache as an auxiliary data store, managing it explicitly rather than having the cache sit transparently between the application and the database.

## How It Works

When the application needs data, it follows this sequence:

1. **Check the cache first**: Query Redis for the requested key.
2. **On cache hit**: Return the cached data immediately.
3. **On cache miss**: Query the primary database, store the result in Redis with an appropriate TTL, then return the data.

The key insight is that the application is responsible for populating the cache. Redis never automatically fetches data from the database.

## Redis Commands Used

A typical cache-aside flow uses these commands:

    GET user:123

If this returns nil (cache miss), the application queries the database and then:

    SET user:123 "{...user data...}" EX 3600

The `EX 3600` sets a one-hour expiration, ensuring stale data eventually expires.

## Advantages

**Resilience to cache failure**: If Redis becomes unavailable, the application degrades gracefully by falling back to the database. Latency increases, but the system remains functional.

**Memory efficiency**: Only data that is actually requested gets cached. This prevents "cold storage" where unused data occupies valuable cache memory.

**Simplicity**: The pattern is straightforward to implement and reason about.

## The Staleness Problem

The primary disadvantage is the potential for stale data. Consider this scenario:

1. Application reads `user:123` from cache (cache hit)
2. Another process updates the user in the database
3. The cached copy remains unchanged until TTL expires

The cache now contains outdated information.

## Mitigating Staleness

The standard mitigation is **cache invalidation on write**. When the application successfully updates data in the database, it immediately deletes the corresponding cache key:

    DEL user:123

The next read will miss the cache and fetch fresh data from the database. This is simpler and safer than trying to update the cache, which risks race conditions.

## When to Use Cache-Aside

This pattern is ideal when:

- The workload is read-heavy
- Brief staleness is acceptable
- You want graceful degradation if the cache fails
- You need fine-grained control over what gets cached

## When to Avoid

Consider other patterns when:

- You need strong consistency between cache and database
- Write-then-immediately-read patterns are common (the read may hit stale cache)
- The cost of a cache miss is extremely high