Write-through and write-back caching in HLD - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

Difference between write-through and write-back caching

Which statement correctly describes the main difference between write-through and write-back caching?

A Write-through caching updates the main memory immediately on every write, while write-back caching updates main memory only when the cache block is replaced.

B Write-back caching updates the main memory immediately on every write, while write-through caching updates main memory only when the cache block is replaced.

C Write-through caching never updates the main memory, only the cache, while write-back caching updates both simultaneously.

D Write-back caching writes data directly to main memory without using cache.

💡 Hint

Think about when the main memory gets updated in each caching method.

Write-through caching writes data to both cache and main memory immediately, ensuring consistency. Write-back caching delays writing to main memory until the cache block is replaced, improving performance but risking data loss on failure.

❓ Architecture

intermediate

Choosing caching strategy for a banking system

You are designing a caching layer for a banking system where data accuracy is critical. Which caching strategy is most appropriate?

A Write-through caching to ensure data is always consistent between cache and main memory.

B Write-back caching to improve performance by delaying writes to main memory.

C No caching to avoid any risk of stale data.

D Write-back caching combined with periodic full memory flushes.

💡 Hint

Consider the importance of data accuracy and consistency in banking.

Write-through caching ensures every write updates main memory immediately, which is crucial for banking systems to avoid data loss or inconsistency.

❓ scaling

advanced

Scaling write-back cache in distributed systems

In a distributed system using write-back caching, what is a major challenge when scaling to many nodes?

A Reducing cache size to fit all data in memory.

B Increasing write latency due to immediate writes to main memory.

C Ensuring cache coherence to prevent stale data across nodes.

D Avoiding network partitioning by disabling caching.

💡 Hint

Think about how delayed writes affect data consistency across multiple nodes.

Write-back caching delays updates to main memory, so in distributed systems, keeping caches coherent and consistent across nodes is challenging as stale data can be served.

❓ tradeoff

advanced

Trade-offs between write-through and write-back caching

Which option correctly describes a trade-off when choosing write-back caching over write-through caching?

A Write-back caching eliminates the need for cache coherence protocols.

B Write-back caching improves write performance but risks data loss on power failure.

C Write-back caching reduces cache complexity but increases main memory writes.

D Write-back caching guarantees data consistency but lowers system throughput.

💡 Hint

Consider performance benefits versus risks of delayed writes.

Write-back caching improves performance by delaying writes but risks losing data if the system crashes before writing back to main memory.

❓ estimation

expert

Estimating memory write load with write-through vs write-back

A system writes data 1000 times per second. With write-through caching, every write updates main memory immediately. With write-back caching, only 10% of writes cause main memory updates due to cache block replacements. How many main memory writes per second occur with write-back caching?

A 900 writes per second

B 1000 writes per second

C 10 writes per second

D 100 writes per second

💡 Hint

Calculate 10% of 1000 writes per second.

Write-back caching reduces main memory writes to only those caused by cache block replacements, which is 10% of total writes, so 100 writes per second.