The round-robin algorithm is a straightforward and widely used technique for load balancing. It distributes incoming client requests sequentially and evenly across a pool of available servers in a cyclical manner.
- Initial Request: The first client request is sent to the first server in the pool.
- Subsequent Requests: Each subsequent request is sent to the next server in the list, cycling back to the first server once the last server has been reached.
- Continuous Cycle: This process repeats indefinitely, ensuring that all servers receive an equal number of requests over time.
- Even Distribution: Ensures all servers receive a similar load, assuming requests are of equal weight.
- Stateless: Doesn't require tracking client-server mappings.
Illustration
- Assume you have 3 servers: S1, S2, and S3.
- Incoming requests: R1, R2, R3, R4, R5, R6.
The round-robin distribution will be:
- R1 → S1
- R2 → S2
- R3 → S3
- R4 → S1 (cycle restarts)
- R5 → S2
- R6 → S3
The Least Connections algorithm is a load-balancing strategy that directs incoming requests to the server with the fewest active connections at the time. This ensures that the server handling the lightest load receives the next request, helping to distribute traffic more evenly, especially when server workloads are not uniform.
- Each server tracks the number of active connections it is currently handling.
- When a new request arrives:
- The load balancer checks the number of active connections on all servers.
- The request is forwarded to the server with the fewest active connections.
- The server processes the request, and once completed, the connection count for that server decreases.
Example
If you have three servers with the following connection states:
- Server 1: 5 active connections
- Server 2: 2 active connections
- Server 3: 3 active connections
The next incoming request will be routed to Server 2 because it has the fewest active connections.