Load Balancing is a concept often used in the network to distributing processing and communications activity evenly across a computer network so that no single device is overwhelmed. Load balancing is not confined to a single geographic site. It may be distributed across the Internet.
There are three main types of load balancing:
- Server load balancing: distributing requests from clients across a group of servers;
- Load Balancing in a multi-processor system: distributing tasks evenly among all processors;
- Network traffic load balance: distributing and sharing links of communications to achieve higher bandwidth and reduce probability of failure.
Server Load balancing is widely used for networks where it's difficult to predict the number of requests that will be issued to a server. Web sites service systems typically employ multiple servers in a load balancing scheme. If one server is too busy, requests are forwarded to another server with more capacity.
Multi-processor in a single platform becomes a common design in the high-end server, in which load balancing schemes are part of the system features. Very often an array of computers are configured into multiprocessing cluster to serve as a load balanced system.
In the case of network traffic load balancing, link aggregation techniques are used to bond multiple links together logically which appears like a single, high bandwidth, fault tolerance link.
Load Balancing
Related Terms: Link Aggregation, Multi-processing, Server