Why caching reduces latency in HLD - Challenge Your Understanding

Challenge - 5 Problems

🧠 Conceptual

intermediate

How does caching reduce latency in a web application?

Imagine a web application that fetches user profile data from a database. Which of the following best explains why adding a cache layer reduces the time it takes to get the data?

A Cache duplicates the database, so the system uses more storage but does not affect speed.

B Cache stores frequently accessed data closer to the user, so requests avoid slow database queries.

C Cache compresses data, making the database queries faster by reducing data size.

D Cache delays requests to the database, batching them to reduce load but increasing latency.

💡 Hint

Think about where the data is stored and how fast it can be accessed compared to the original source.

Caching stores copies of frequently requested data in a faster storage layer, often memory, which is closer to the application or user. This avoids the slower process of querying the database every time, thus reducing latency.

❓ Architecture

intermediate

Where to place cache to reduce latency effectively?

In a typical client-server system, where is the best place to add a cache to reduce latency for repeated data requests?

A On the server's hard disk, to store data permanently for faster access.

B Only inside the database server, so queries run faster internally.

C Between the client and server, so the client can get data without contacting the server every time.

D On the network routers, to speed up data packets.

💡 Hint

Consider where repeated requests happen and where storing data temporarily can save time.

Placing cache between client and server (like in-memory cache or CDN) allows repeated requests to be served quickly without hitting the server or database, reducing latency significantly.

❓ scaling

advanced

Impact of cache size on latency reduction

A system uses caching to reduce latency. What happens if the cache size is too small compared to the data requested?

A Cache hits increase, so latency decreases further.

B Cache automatically expands to fit all data, so latency stays low.

C Cache size does not affect latency as long as caching is enabled.

D Cache misses increase, causing more requests to go to the slower database, increasing latency.

💡 Hint

Think about what happens when the cache cannot hold all the data needed.

If the cache is too small, it cannot store all frequently requested data, leading to more cache misses. This forces the system to fetch data from the slower database more often, increasing latency.

❓ tradeoff

advanced

Tradeoff between cache freshness and latency

Why might a system choose to serve slightly stale data from cache instead of always fetching fresh data from the database?

A Serving stale data reduces latency because it avoids slower database queries, improving user experience.

B Serving stale data requires more network bandwidth, increasing latency.

C Serving fresh data always reduces latency because it avoids cache overhead.

D Serving stale data increases latency because the system must check timestamps before responding.

💡 Hint

Consider the speed difference between cache and database and the impact of data freshness.

Serving slightly stale data from cache avoids the slower process of querying the database every time. This reduces latency and improves responsiveness, though at the cost of data freshness.

❓ estimation

expert

Estimating latency reduction from caching

A database query takes 200 ms on average. A cache lookup takes 5 ms. If 80% of requests hit the cache, what is the average latency per request?

A 45 ms

B 50 ms

C 40 ms

D 60 ms

💡 Hint

Calculate weighted average latency based on cache hit and miss rates.

Average latency = (cache hit rate * cache latency) + (cache miss rate * (cache latency + database latency)) = (0.8 * 5) + (0.2 * (5 + 200)) = 4 + 41 = 45 ms.