Load balancing algorithms (round robin, least connections) in HLD - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

How does the round robin load balancing algorithm distribute requests?

Imagine you have 4 servers and a load balancer using the round robin algorithm. How does it assign incoming requests?

A It sends each new request to the next server in order, cycling back to the first after the last.

B It sends all requests to the server with the fewest active connections.

C It sends requests randomly to any server without order.

D It sends requests only to the server with the highest CPU usage.

💡 Hint

Think about a simple way to share requests evenly without checking server load.

Round robin cycles through servers one by one, sending each new request to the next server in the list. This balances load evenly if all servers have similar capacity.

❓ Architecture

intermediate

Which load balancing algorithm is best when servers have uneven workloads?

You have 3 servers with different current loads. Which algorithm helps distribute new requests to the least busy server?

A Round robin, because it cycles evenly through servers.

B Random, because it picks any server without checking load.

C IP hash, because it sends requests based on client IP.

D Least connections, because it sends requests to the server with the fewest active connections.

💡 Hint

Consider which algorithm checks server load before assigning requests.

Least connections algorithm sends requests to the server with the fewest active connections, helping balance uneven workloads.

❓ scaling

advanced

What is a key limitation of round robin load balancing in a system with servers of different capacities?

Consider a system with 2 servers: one powerful and one weaker. What problem can round robin cause?

A It causes requests to be dropped randomly.

B It may overload the weaker server by sending equal requests to both servers regardless of capacity.

C It requires servers to share session state.

D It always sends requests to the most powerful server only.

💡 Hint

Think about how equal distribution affects servers with different power.

Round robin does not consider server capacity, so it can overload weaker servers by sending them the same number of requests as stronger ones.

❓ tradeoff

advanced

What is a tradeoff when using least connections load balancing compared to round robin?

Least connections balances load better but has a cost. What is a downside compared to round robin?

A It requires tracking active connections on each server, increasing complexity and overhead.

B It always sends requests to the same server, causing imbalance.

C It cannot handle more than 2 servers.

D It ignores server health status.

💡 Hint

Think about what extra information least connections needs to work.

Least connections needs to monitor active connections on each server, which adds complexity and overhead compared to simple round robin.

❓ estimation

expert

Estimate the maximum number of concurrent connections a load balancer can handle using least connections algorithm with 5 servers, each supporting 1000 connections.

Assuming the load balancer perfectly distributes connections using least connections algorithm, what is the maximum total concurrent connections supported?

A 1000 connections, limited by the weakest server.

B 2000 connections, assuming half the servers are used at a time.

C 5000 connections, as it sums the capacity of all servers.

D 10000 connections, doubling the total server capacity.

💡 Hint

Think about how least connections distributes load across all servers.

Least connections balances load so all servers are used up to their capacity, so total max connections is sum of all servers' capacities.