Why load balancers distribute traffic in HLD - The Real Reasons

The Big Idea

What if one simple tool could stop your website from crashing during its busiest moments?

The Scenario

Imagine a busy restaurant where all customers line up at a single counter to place orders. The staff at that counter tries to handle every request, but the line grows longer and customers get frustrated waiting.

The Problem

Handling all requests through one point causes slow service and mistakes. The staff gets overwhelmed, orders get mixed up, and some customers leave unhappy. This manual approach can't keep up when many people arrive at once.

The Solution

A load balancer acts like a smart host who directs each customer to the shortest line or the next available staff member. It spreads the work evenly, so no one gets overwhelmed and everyone is served faster and more reliably.

Before vs After

✗ Before

All requests -> Single server -> Slow response

✓ After

Requests -> Load balancer -> Multiple servers -> Fast response

What It Enables

It enables systems to handle many users smoothly by sharing the work across multiple servers, preventing overload and downtime.

Real Life Example

Popular websites like online stores use load balancers to make sure thousands of shoppers can browse and buy without delays or crashes, even during big sales.

Key Takeaways

Manual single-point handling causes delays and errors.

Load balancers distribute traffic evenly to multiple servers.

This improves speed, reliability, and user experience.

Practice

1. Why do load balancers distribute incoming traffic across multiple servers?

easy

A. To prevent any single server from becoming overloaded and improve system reliability

B. To increase the number of servers needed for the system

C. To slow down the response time for users

D. To reduce the total number of user requests

Solution

1. Step 1: Understand server overload

   When too many requests go to one server, it slows down or crashes.

2. Step 2: Role of load balancers

   Load balancers spread requests evenly to keep servers healthy and fast.

3. Final Answer:

   To prevent any single server from becoming overloaded and improve system reliability -> Option A

4. Quick Check:

   Load balancers prevent overload = C [OK]

Hint: Load balancers keep servers from getting too busy [OK]

Common Mistakes:

• Thinking load balancers reduce total requests
• Believing load balancers slow down responses
• Assuming load balancers increase server count unnecessarily

2. Which of the following is a common method used by load balancers to distribute traffic?

easy

A. Bubble sort

B. Binary search

C. Depth-first search

D. Round-robin

Solution

1. Step 1: Identify load balancing methods

   Common methods include round-robin and least connections.

2. Step 2: Match options to known methods

   Round-robin cycles through servers evenly; others are unrelated algorithms.

3. Final Answer:

   Round-robin -> Option D

4. Quick Check:

   Round-robin is a load balancing method = D [OK]

Hint: Round-robin cycles requests evenly [OK]

Common Mistakes:

• Confusing sorting/search algorithms with load balancing
• Choosing unrelated computer science terms
• Ignoring common load balancing strategies

3. Consider a load balancer using the least connections strategy. If Server A has 3 active connections and Server B has 1 active connection, where will the next request be sent?

medium

A. Server B

B. Server A

C. Both servers equally

D. To a new server not listed

Solution

1. Step 1: Understand least connections strategy

   It sends new requests to the server with fewer active connections.

2. Step 2: Compare active connections

   Server A has 3, Server B has 1, so Server B has fewer connections.

3. Final Answer:

   Server B -> Option A

4. Quick Check:

   Least connections picks server with fewer connections = B [OK]

Hint: Send to server with least active connections [OK]

Common Mistakes:

• Choosing server with more connections
• Assuming equal distribution regardless of load
• Thinking new servers are automatically chosen

4. A load balancer is configured to use round-robin but always sends requests to the same server. What is the most likely cause?

medium

A. The servers are all down

B. The load balancer is not tracking the last server used correctly

C. The network is too slow

D. The user requests are too few

Solution

1. Step 1: Understand round-robin behavior

   Round-robin cycles through servers in order to balance load.

2. Step 2: Identify cause of repeated server selection

   If it always picks the same server, it likely lost track of which server was last used.

3. Final Answer:

   The load balancer is not tracking the last server used correctly -> Option B

4. Quick Check:

   Round-robin needs correct tracking = A [OK]

Hint: Round-robin must remember last server used [OK]

Common Mistakes:

• Blaming server downtime without checking load balancer
• Assuming network speed affects server choice
• Thinking low request volume causes imbalance

5. You design a system with a load balancer distributing traffic to 5 servers. Suddenly, one server becomes much slower but still receives equal requests. What change to the load balancing strategy can improve overall system performance?

hard

A. Use round-robin but add more user requests

B. Increase the number of servers without changing strategy

C. Switch from round-robin to least connections to avoid slow server overload

D. Disable the load balancer and send requests to one server

Solution

1. Step 1: Identify problem with round-robin

   Round-robin sends equal requests regardless of server speed, causing slow server overload.

2. Step 2: Choose better strategy

   Least connections sends requests to less busy servers, reducing load on slow server.

3. Final Answer:

   Switch from round-robin to least connections to avoid slow server overload -> Option C

4. Quick Check:

   Least connections adapts to server speed = A [OK]

Hint: Use least connections to avoid slow servers [OK]

Common Mistakes:

• Adding servers without fixing load distribution
• Increasing requests worsens overload
• Disabling load balancer causes single point failure