--- title: "What is a Load Balancer?" description: "What is a Load Balancer?" author: "ICSM Computer" published: 2026-01-15 updated: 2026-01-15 canonical: https://www.mindstick.com/interview/34443/what-is-a-load-balancer category: "website & mobile applications" tags: ["api(s)", "cloud", "application"] reading_time: 5 minutes --- # What is a Load Balancer? A [**Load Balancer**](https://github.com/SocketCluster/loadbalancer) is a system (hardware or software) that **distributes incoming network traffic across multiple servers** to ensure **high availability, scalability, performance, and reliability** of an application Instead of all users hitting one server, the load balancer decides **which server should handle each request**. ## Why Load Balancers Are Needed ## Without a load balancer: - One server can get overloaded - Application may become slow or crash - Single point of failure exists ## With a load balancer: - Traffic is evenly distributed - Failed servers are automatically avoided - Application remains fast and available ## Basic Architecture (Image / Diagram) ![What is a Load Balancer?](https://www.mindstick.com/interviewquestion/18843e8d-9e57-47b7-a8f6-fbd7a78eff65/images/f575583d-3a52-4ff1-969c-58fa0deeefed.png) ## Working Process of a Load Balancer (Step by Step) ### 1. Client Sends Request A user opens a website or API: ```plaintext https://example.com ``` The request first goes to the **Load Balancer**, not directly to a server. ### 2. Health Check The load balancer continuously checks: - Is Server A alive? - Is Server B responding? - Is Server C overloaded? Unhealthy servers are **removed from rotation**. ### 3. Load Balancing Algorithm Selection The load balancer chooses a server using an algorithm such as: - Round Robin - Least Connections - IP Hash ### 4. Request Forwarding The request is forwarded to the selected server. ```plaintext Client → Load Balancer → Server B ``` ### 5. Server Responds The chosen server processes the request and sends the response back **through the load balancer** to the client. ## Request Flow Diagram (Working Process) ![What is a Load Balancer?](https://www.mindstick.com/interviewquestion/18843e8d-9e57-47b7-a8f6-fbd7a78eff65/images/93805bf4-b072-4699-90df-f19454c726fe.png) ## Common Load Balancing Algorithms ### 1. Round Robin Requests are distributed sequentially. ```plaintext Request 1 → Server A Request 2 → Server B Request 3 → Server C ``` Best for: servers with equal capacity. ### 2. Least Connections Request goes to the server with **fewest active connections**. Best for: long-running requests. ### 3. IP Hash Same client IP always goes to the same server. Best for: session-based applications. ## Types of Load Balancers ### 1. Hardware Load Balancer - Physical device - Very fast but expensive - Used in enterprise data centers ### 2. Software Load Balancer - Runs as software - Flexible and cost-effective - Examples: - NGINX - HAProxy - Traefik ### 3. Cloud Load Balancer - Managed by cloud providers - Auto-scaling and highly available - Examples: - AWS ALB / NLB - Azure Load Balancer - Google Cloud Load Balancer ## Layer-wise Load Balancing ### Layer 4 (Transport Layer) - Based on IP + Port - Faster - No content inspection Example: ```plaintext TCP traffic on port 443 ``` ## Real-World Example ### E-commerce Website Sale Day Without Load Balancer: ```plaintext Users → Single Server → Crash ``` With Load Balancer: ```plaintext Users → Load Balancer → 10 Servers → Smooth experience ``` ## Key Benefits Summary - High availability - Fault tolerance - Scalability - Better performance - Zero-downtime maintenance ## Answers ### Answer by ICSM Computer A [**Load Balancer**](https://github.com/SocketCluster/loadbalancer) is a system (hardware or software) that **distributes incoming network traffic across multiple servers** to ensure **high availability, scalability, performance, and reliability** of an application Instead of all users hitting one server, the load balancer decides **which server should handle each request**. ## Why Load Balancers Are Needed ## Without a load balancer: - One server can get overloaded - Application may become slow or crash - Single point of failure exists ## With a load balancer: - Traffic is evenly distributed - Failed servers are automatically avoided - Application remains fast and available ## Basic Architecture (Image / Diagram) ![What is a Load Balancer?](https://www.mindstick.com/interviewquestion/18843e8d-9e57-47b7-a8f6-fbd7a78eff65/images/f575583d-3a52-4ff1-969c-58fa0deeefed.png) ## Working Process of a Load Balancer (Step by Step) ### 1. Client Sends Request A user opens a website or API: ```plaintext https://example.com ``` The request first goes to the **Load Balancer**, not directly to a server. ### 2. Health Check The load balancer continuously checks: - Is Server A alive? - Is Server B responding? - Is Server C overloaded? Unhealthy servers are **removed from rotation**. ### 3. Load Balancing Algorithm Selection The load balancer chooses a server using an algorithm such as: - Round Robin - Least Connections - IP Hash ### 4. Request Forwarding The request is forwarded to the selected server. ```plaintext Client → Load Balancer → Server B ``` ### 5. Server Responds The chosen server processes the request and sends the response back **through the load balancer** to the client. ## Request Flow Diagram (Working Process) ![What is a Load Balancer?](https://www.mindstick.com/interviewquestion/18843e8d-9e57-47b7-a8f6-fbd7a78eff65/images/93805bf4-b072-4699-90df-f19454c726fe.png) ## Common Load Balancing Algorithms ### 1. Round Robin Requests are distributed sequentially. ```plaintext Request 1 → Server A Request 2 → Server B Request 3 → Server C ``` Best for: servers with equal capacity. ### 2. Least Connections Request goes to the server with **fewest active connections**. Best for: long-running requests. ### 3. IP Hash Same client IP always goes to the same server. Best for: session-based applications. ## Types of Load Balancers ### 1. Hardware Load Balancer - Physical device - Very fast but expensive - Used in enterprise data centers ### 2. Software Load Balancer - Runs as software - Flexible and cost-effective - Examples: - NGINX - HAProxy - Traefik ### 3. Cloud Load Balancer - Managed by cloud providers - Auto-scaling and highly available - Examples: - AWS ALB / NLB - Azure Load Balancer - Google Cloud Load Balancer ## Layer-wise Load Balancing ### Layer 4 (Transport Layer) - Based on IP + Port - Faster - No content inspection Example: ```plaintext TCP traffic on port 443 ``` ## Real-World Example ### E-commerce Website Sale Day Without Load Balancer: ```plaintext Users → Single Server → Crash ``` With Load Balancer: ```plaintext Users → Load Balancer → 10 Servers → Smooth experience ``` ## Key Benefits Summary - High availability - Fault tolerance - Scalability - Better performance - Zero-downtime maintenance --- Original Source: https://www.mindstick.com/interview/34443/what-is-a-load-balancer Copyright © MindStick Software Pvt. Ltd. This Markdown version is provided for developers, AI systems, and offline reading.