Caching Strategies

Question 1

In a write-back caching strategy, when is data written to the database?

Accepted Answer

Asynchronously after the write — Write-back caching writes to cache first and asynchronously persists to the database later, improving write performance.

Question 2

The primary benefit of caching is to…

Accepted Answer

Serve hot data faster — A cache stores copies of hot data in fast memory (or nearer the client) to reduce latency and load on the origin.

Question 3

A cache HIT occurs when…

Accepted Answer

Data is found in the cache — On a hit, the cache serves the data directly; a MISS forces a slower fetch from the underlying data store.

Question 4

In the CACHE-ASIDE (lazy loading) pattern, the application…

Accepted Answer

Reads cache first, queries DB on miss — Cache-aside loads data on demand: the app manages reads, fetching from the DB on a miss and storing it for next time.

Question 5

WRITE-THROUGH caching writes data…

Accepted Answer

Cache and DB, synchronously — Write-through keeps the cache and DB consistent by updating both on each write — reads are fast, writes incur extra latency.

Question 6

WRITE-BACK (write-behind) caching improves write performance by…

Accepted Answer

Write to cache first, flush DB async — It acknowledges writes from the cache and persists to the DB in the background — fast, but risks data loss if the cache fails before flushing.

Question 7

TTL (Time To Live) on a cache entry controls…

Accepted Answer

Controls when the entry expires — TTL bounds staleness: after it expires, the entry is evicted/refetched, trading freshness against origin load.

Question 8

Which item does an LRU cache evict when it is full?

Accepted Answer

The item not accessed for the longest time — LRU evicts the entry not accessed for the longest time, assuming recently-used data is likely to be used again.

Question 9

Cache INVALIDATION is hard because…

Accepted Answer

Cache and DB values can diverge — Famously 'one of the two hard things in CS': deciding when/what to invalidate to avoid serving stale data is subtle.

Question 10

A 'cache stampede' (thundering herd) happens when…

Accepted Answer

Many concurrent requests miss at once — Simultaneous misses for a hot key overload the backend; mitigations include locking, request coalescing, and staggered/early TTL refresh.

Write-through or write-back?

Test your knowledge

In a write-back caching strategy, when is data written to the database?

Interview Questions

References & Further Reading

Related topics

Love this visualization?