Cache eviction policies (LRU, LFU, TTL) in HLD - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

Understanding LRU Cache Behavior

Consider a cache with a capacity of 3 items using the Least Recently Used (LRU) eviction policy. The cache is initially empty. The following keys are accessed in order: 1, 2, 3, 1, 4.

After these accesses, which keys remain in the cache?

A [2, 3, 4]

B [1, 3, 4]

C [1, 2, 4]

D [2, 3, 1]

💡 Hint

Remember that LRU removes the least recently used item when the cache is full.

Initially, keys 1, 2, 3 fill the cache. Accessing key 1 makes it most recently used. When key 4 is accessed, the least recently used key (2) is evicted. Remaining keys are 1, 3, and 4.

❓ Architecture

intermediate

Choosing Cache Eviction Policy for Read-Heavy Workload

You are designing a caching layer for a read-heavy web application where some items are accessed very frequently, and others rarely. Which cache eviction policy is most suitable to maximize cache hit rate?

A Least Frequently Used (LFU)

B Random eviction

C Time To Live (TTL) based eviction

D Least Recently Used (LRU)

💡 Hint

Consider which policy favors items accessed many times over time.

LFU keeps items accessed most frequently, which suits read-heavy workloads with hot items. LRU favors recent access but may evict frequently accessed older items. TTL evicts based on time, not usage frequency.

❓ scaling

advanced

Scaling Cache with TTL Eviction in Distributed Systems

You have a distributed cache system with TTL-based eviction. What is a major challenge when scaling this system across multiple nodes?

A Randomly evicting items to balance load

B Handling cache misses due to LRU eviction

C Tracking frequency counts for LFU eviction

D Ensuring consistent TTL expiration times across nodes

💡 Hint

Think about time synchronization and expiration consistency.

In distributed caches with TTL, nodes must agree on expiration times. Clock skew or delays can cause inconsistent evictions, leading to stale or missing data. LRU and LFU challenges differ and are unrelated to TTL expiration.

❓ tradeoff

advanced

Tradeoffs Between LRU and LFU Cache Policies

Which statement best describes a key tradeoff between LRU and LFU cache eviction policies?

A LRU requires tracking access frequency, LFU only tracks recency

B LFU is simpler to implement than LRU and uses less memory

C LRU adapts quickly to changing access patterns, while LFU may retain stale popular items longer

D Both LRU and LFU evict items randomly when full

💡 Hint

Consider how each policy responds to changes in item popularity over time.

LRU evicts least recently used items, adapting quickly if access patterns change. LFU evicts least frequently used items, which can cause it to keep items that were popular in the past but are no longer accessed.

❓ estimation

expert

Estimating Cache Size for TTL Eviction

A cache stores items with a TTL of 10 minutes. On average, 1000 new items are added per minute. Assuming steady state and no evictions other than TTL expiration, approximately how many items will the cache hold?

A 10,000 items

B 1000 items

C 100 items

D 1,000,000 items

💡 Hint

Think about how many items accumulate over the TTL duration.

With 1000 items added per minute and a TTL of 10 minutes, the cache holds items added in the last 10 minutes: 1000 * 10 = 10,000 items.