Core Concepts & L4 vs L7

What a Load Balancer Does

A load balancer sits between clients and backend servers, distributing incoming requests across a pool of identical servers. It solves two problems simultaneously: the single-server performance ceiling and the single point of failure.

Without a load balancer, scaling is vertical — expensive, finite, and still a SPOF. With one, scaling is horizontal: add commodity servers, and the load balancer distributes the work.

Common Distribution Algorithms

• Round Robin — Requests distributed sequentially. Best when servers are identical and requests have similar cost.
• Least Connections — Route to the server with the fewest active connections. Better when request cost varies significantly.
• IP Hash — Route based on a hash of the client IP, ensuring the same client always reaches the same server. Used for sticky session requirements.

L4 vs L7 Load Balancers

Layer 4 (Transport Layer) — Operates at TCP/UDP level. Forwards packets based on IP and port without inspecting content.

• Strengths: Extremely fast; minimal CPU overhead; handles millions of connections per second.
• Use when: Maximum throughput is critical; traffic is non-HTTP (game servers, databases, streaming); no content-based routing needed.

Layer 7 (Application Layer) — Terminates the TCP connection and inspects the HTTP request before routing. Routes based on URL path, headers, or cookies.

• Strengths: Intelligent routing (e.g., /api/* to API servers, /static/* to file servers); SSL termination; WAF integration; session persistence.
• Trade-off: Higher CPU and memory cost than L4.

Decision Factor	Choose L4	Choose L7
Primary goal	Maximum throughput	Content-based routing
Traffic type	Non-HTTP (databases, game servers)	HTTP/HTTPS
Routing logic	Round-robin or least-connections	URL, header, or cookie-based
Feature needs	No SSL or WAF required	SSL termination, WAF, session persistence
Architecture	Homogeneous backend servers	Multiple microservice pools