What is Load balancing

Load balancing technology is a key solution for delivering faster service with higher efficiency in a computer network. Most commonly, the load balancing refers to a technique that distributes processing work and communicating activities evenly across main network servers in an effort to not overload one server and avoid uncoordinated processing work of the servers, so that no single server is overwhelmed.

In overall, load-balancing technique is significant to the servers, for which it is difficult to predict the number of requests that might be issued to a server. Busy websites and university websites, in which multi user registration is required, are in this group.

Most of busy and reliable websites typically employ two or more parallel Web servers in a load balancing scheme. If one server starts to get swamped or its resource get overused, in the way that continuing the activity might cause it serious damage or down the server, requests are forwarded to another server with more capacity.

Load balancing performance:

In general, load balancing placed between a host and a client. When a host encounters trouble or fail, the load balancer service detects the problem, and immediately route requests, which referred to the failed host to the valid host.

It must be considered that the processes will be done automatically. The entire connections, which are referred to the failed host, will be disconnected and routed to the valid host.

It worth to be noted that the user will not be aware about any of the mentioned process. So that service availability in this situation in comparing with using applying only one server will be maximized.

Host detecting process will canceled, and the routing process will restart and finally finding new host will take less than 10 seconds, so that the user will never get service disconnecting due to one server downing problem.

Load balancing algorithms:

Load balancing algorithm is determined according to different measure, in all of them host server prefers has been determined according to client requests. These algorithms aim is intelligent distribution of processing work or server availability maximization.

• Round-robin:

  This algorithm evenly distributes process across servers without considering number of active connections and their replying time. This algorithm is useful when the servers process powers are the same, otherwise some of them might get overload whereas more powerful servers are using part of their sources and abilities.

• Weighted round robin:

  This algorithm considers each server processing power. The administrator manually assign each server performance weight, then a scheduled table will be created according to the assigned weight, and then the request will be send to each server scheduled.

• Least connection:

  When the least connection is assigned to the load balancer, new requests to the available server with least active connection will be sent.

• Load-based:

  The load based algorithm send request to the servers according to the factor, which is least processing load of each server.

Conclusion:

In overall, load-balancing process provides minimization of server processing noises and server failing or unavailability. It worth to be noted, that the load balancing service might be set up in both software and hardware models.