Question 1

What is the main purpose of a cache?

Accepted Answer

A cache sits between your app and the slow data source, serving repeated reads from fast memory instead of hitting the database every time.

Question 2

A cache hit occurs when:

Accepted Answer

A cache hit means the data was already in the cache, so the request is served instantly without touching the database.

Question 3

A cache miss occurs when:

Accepted Answer

A cache miss forces the app to query the database and then populate the cache for future requests.

Question 4

Cache invalidation means:

Accepted Answer

Cache invalidation keeps the cache consistent with the source of truth by evicting or refreshing entries when the underlying data changes.

Question 5

In Cache-Aside strategy, who is responsible for populating the cache?

Accepted Answer

In Cache-Aside (lazy loading), the application checks the cache, and on a miss it queries the database and writes the result back into the cache itself.

Question 6

What is a major advantage of the Write-Through strategy?

Accepted Answer

Write-Through updates both the cache and the database synchronously on every write, so they are never out of sync.

Question 7

What is the biggest risk of the Write-Behind (Write-Back) strategy?

Accepted Answer

Write-Behind buffers writes in the cache and flushes asynchronously. If the cache crashes before flushing, those buffered writes are lost permanently.

Question 8

TTL stands for:

Accepted Answer

TTL (Time To Live) is the duration after which a cache entry automatically expires and is removed, preventing stale data from persisting indefinitely.

Question 9

Which TTL would be most appropriate for a live sports scoreboard updating every second?

Accepted Answer

Fast-changing data like live scores needs a very short TTL (1–5 seconds) so users always see near-real-time data.

Question 10

Which TTL would be most appropriate for a URL redirect that almost never changes?

Accepted Answer

Stable data like URL redirects can safely be cached for hours, dramatically reducing database load with minimal staleness risk.

Question 11

What is a cache stampede (also called thundering herd)?

Accepted Answer

A cache stampede happens when a popular entry expires and hundreds of concurrent requests all miss and pile up on the database simultaneously.

Question 12

In a URL shortener, the best cache key pattern for looking up short.ly/xyz would be:

Accepted Answer

Cache keys should be namespaced and meaningful. "redirect:xyz" makes the purpose clear and avoids collisions with other key types in the same Redis instance.

Question 13

Redis Sorted Sets are particularly useful for what use case?

Accepted Answer

Redis Sorted Sets (ZADD/ZREVRANGE) store members with scores and keep them sorted automatically, making leaderboard queries O(log n).

Question 14

Which caching strategy writes to the cache only and syncs to the database asynchronously later?

Accepted Answer

Write-Behind (Write-Back) acknowledges the write after updating only the cache, then flushes to the database in a background batch for higher write throughput.

Question 15

For a search autocomplete that should never cache single-character inputs, what is the correct condition?

Accepted Answer

Skipping cache for prefixes shorter than 2 characters avoids caching extremely broad (and likely never-reused) result sets that would pollute the cache.

Question 16

A high cache hit rate means what?

Accepted Answer

A high hit rate (e.g. 90%+) means the cache is effectively absorbing traffic that would otherwise reach the database — the goal of any caching layer.

Question 17

What does redis.setEx("key", 300, value) do?

Accepted Answer

setEx sets a key-value pair with an expiry in seconds. After 300 seconds the key is automatically removed by Redis with no extra code needed.

Question 18

In Write-Through, what happens when a write request comes in?

Accepted Answer

Write-Through keeps cache and DB in sync by writing both atomically. Reads are always fresh, but writes are slower because both stores must confirm.

Question 19

Which strategy is best for a read-heavy app where stale data for a few seconds is acceptable?

Accepted Answer

Cache-Aside with a short TTL is the most common pattern — simple to implement, low write overhead, and tolerable staleness is bounded by the TTL.

Question 20

"Debouncing" in the context of search autocomplete means:

Accepted Answer

Debouncing (e.g. 150ms delay) prevents flooding the server with a request on every keystroke — the API is only called once the user pauses.

Question 21

Which of the following is true about cache eviction?

Accepted Answer

Eviction is automatic memory management (e.g. LRU policy). Cache invalidation is explicit removal when data changes. They are related but distinct concepts.

Question 22

After a cache miss, the application should do what?

Accepted Answer

The standard cache-miss flow is read-through recovery — fetch from DB, write to cache, return to caller — so the next identical request hits the cache.

Question 23

A good cache hit rate to aim for in production is:

Accepted Answer

85–95% is a healthy target. 100% is impossible (cold starts, new data). Below 70% suggests the cache key design or TTL needs tuning.

Question 24

In the autocomplete scenario, why do we lowercase the prefix before using it as a cache key?

Accepted Answer

Normalizing case before keying ensures case-insensitive lookups share cache entries, improving hit rate and preventing redundant DB queries for the same intent.