Throughput, latency, and availability in HLD - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

Understanding Throughput in a Web Service

Imagine a web service that handles user requests. Which of the following best describes throughput?

A The number of requests the service can handle per second.

B The time it takes for a single request to be processed.

C The percentage of time the service is operational.

D The amount of data transferred in a single request.

💡 Hint

Think about how many requests the system can complete over time.

Throughput measures how many requests a system can process in a given time, usually per second.

❓ Architecture

intermediate

Reducing Latency in a Distributed System

You have a distributed system with multiple servers across regions. Which architectural choice will most effectively reduce latency for users?

A Use a single powerful server to handle all requests.

B Deploy servers closer to users geographically.

C Store all data in a central database far from users.

D Increase the number of servers in a single data center.

💡 Hint

Latency depends on the time data takes to travel between user and server.

Deploying servers closer to users reduces the distance data travels, lowering latency.

❓ scaling

advanced

Scaling for High Availability

Your system must maintain high availability even during server failures. Which design choice best supports this?

A Schedule maintenance during peak hours to fix issues quickly.

B Use a single server with a powerful CPU and memory.

C Use multiple redundant servers with automatic failover.

D Store backups only once per week to save storage space.

💡 Hint

Think about how to keep the system running if one server stops working.

Redundancy and automatic failover ensure the system stays available despite failures.

❓ tradeoff

advanced

Tradeoff Between Latency and Throughput

In a system design, increasing throughput sometimes increases latency. Which scenario best explains this tradeoff?

A Adding more servers to handle requests in parallel.

B Processing each request immediately, reducing latency but lowering throughput.

C Using caching to serve requests faster without affecting throughput.

D Batch processing many requests together to improve throughput but causing delay for each request.

💡 Hint

Think about how grouping requests affects speed and quantity.

Batching increases throughput by processing many requests at once but adds waiting time, increasing latency.

❓ estimation

expert

Estimating System Availability

A system has 3 independent servers, each with 99.9% uptime. What is the approximate overall system availability if it requires at least one server to be up?

A 99.9999%

B 99.999%

C 99.0%

D 99.7%

💡 Hint

Calculate the chance all servers fail simultaneously, then subtract from 100%.

Availability = 1 - (probability all servers fail). Each server fails 0.1% = 0.001. All fail = 0.001^3 = 0.000000001. So availability ≈ 1 - 0.000000001 = 99.9999999%, closest option 99.9999% (A).