One of the main difficulties encountered by network administrators is scalability, i.e. the ability to meet requests in an acceptable amount of time, even in a high traffic situation.
Load balancing consists in distributing a task to a pool of machines or peripherals in order to:
- even out network traffic, i.e. distribute the overall burden to different machines
- ensure equipment availability by sending data only to those machines that are able to handle it or to those machines that have the best response time
This type of mechanism relies on a load balancer, whose job it is to distribute the work between different machines.
There are many ways to implement load balancing:
- With a layer 4 switch
- With a server that uses a Round Robin algorithm