Which caching pattern writes data only to the cache and flushes it to the database asynchronously in the background?

Which caching pattern writes data only to the cache and flushes it to the database asynchronously in the background?

• A. Write-Behind (Write-Back) Caching
• B. Read-Through Caching
• C. External Cache (Redis)
• D. CDN Caching

Answer: (A)

Write-behind caching is the pattern where the application writes data to the cache and the database is updated asynchronously in the background. This means the write path is fast because the cache acknowledges the write immediately, while the actual persistence to the database happens later, typically by a background worker or flush queue. This can greatly increase write throughput and reduce latency, since the database isn’t blocking the application on a synchronous commit. Understand that this introduces eventual consistency: if the cache or the flush pipeline fails before the data reaches the database, you can lose updates or serve stale data until recovery. Systems using this pattern must handle durability guarantees, failure scenarios, and proper invalidation or reloading of data to maintain correctness. Choice focusing on reads, like a mechanism that automatically loads data from the database on a miss, doesn’t specify how writes propagate. Simply placing a cache in front of a data store or using a CDN describes caching content or delivery rather than how data is persisted.

Write-behind caching is the pattern where the application writes data to the cache and the database is updated asynchronously in the background. This means the write path is fast because the cache acknowledges the write immediately, while the actual persistence to the database happens later, typically by a background worker or flush queue. This can greatly increase write throughput and reduce latency, since the database isn’t blocking the application on a synchronous commit.

Understand that this introduces eventual consistency: if the cache or the flush pipeline fails before the data reaches the database, you can lose updates or serve stale data until recovery. Systems using this pattern must handle durability guarantees, failure scenarios, and proper invalidation or reloading of data to maintain correctness.

Choice focusing on reads, like a mechanism that automatically loads data from the database on a miss, doesn’t specify how writes propagate. Simply placing a cache in front of a data store or using a CDN describes caching content or delivery rather than how data is persisted.