Server Load Balancer Algorithm

Question

What all I know about a load balancer is that:-

When we have high traffic on our site, or we are using multiple servers, then a load balancer is established in front of all the servers. and any http request directly hits the load balancer and from there it reaches the respective server, depending the server loads.

Q1 Can Someone exactly explain the Algorithm that the load balancer uses to balance the load in multiple servers.

Q2 Can we create our own load balancer or do we have to take it from some one else like Barracuda networks or others.

Answer 1

1) there are various ways to do this (round-robin, least-connection, weighted, ...) a good overview: http://www.centos.org/docs/5/html/Virtual_Server_Administration/s2-lvs-sched-VSA.html

2) "create our own" - you probably don't want to reinvent the wheel, there are lots of existing products around, both commercial and open-source/freeware. some of them are specialized on http requests, others support all sorts of protocols.

Answer 2

Q1: The most simple algoritm is round-robin. It just goes trough every existing sever and takes the next one for the next request.

Q2: Of course you can create your own one, or you install one of the available open-source o commercial products on one of you servers.