LoadBalancing - Types,Techniques!!!

What is Load – Balancing?

It is a process/technique of distributing the requests evenly across a computer network so that no single device is overloaded. It becomes an essential entity for networks where it's difficult to predict the number of requests that will be issued to a server. If one server starts to get swamped, requests are forwarded to another server so that the requests are not failed.

Why do we need load balancing?

Load balancing is an important attribute in a high complicated applications or when the user requests are high. It became mandate in the current trend applications because of the following benefits:

·         Better performance - By sharing the web traffic

·         High Scalability - By distributing the client requests across multiple servers

·         High availability – Re-routes the traffic to another server if any of the server is down

Where Load-Balancing can be applied?

Load Balancing can be applied in any layer. It can be applied in

·         Web Servers

·         Application Servers

·         Database Servers

How does Load Balancing work?

Load Balancing intercepts each incoming HTTP request and redirects them to a cluster of servers. This logic is achieved on various means based on the type of sprayer used.

Equipment Used:

Load balancing can be done by

·         Load dispatcher

·         Network dispatcher

·         Load balancer

Load Balancing Algorithms:

1.    Random Allocation

In this algorithm, HTTP requests are assigned to a server picked randomly from a group of servers.  One of the servers may be assigned many requests to process, while the other servers can be idle.

Pros:

Simple to implement

Cons:

May lead to overloading of one server while others are under-utilized

2.    Round Robin Allocation

IP sprayer assigns the requests to a list of the servers based on a circular queue

Only the first request is allocated to a server randomly whereas the IP sprayer follows the circular order for subsequent requests

pros:

Requests are equally divided among the available servers

cons:

Not suitable if the server specifications are not identical to each other in the server group

3.    Weighted Round-Robin Allocation

A weight is assigned to each server in the group. If one server is capable of handling twice as much load as the   other, the powerful server gets a weight of 2. Two requests are assigned to a powerful server wherein the weaker one gets one request.  

 

Pros:

Capacity of the servers in the group are taken care of

Cons:

For each individual request, the advanced load balancing requirements such as processing times are not considered

Method of load balancing

4.    Round Robin DNS Load Balancing

In-built round-robin feature of a DNS server is used. It’s an early adopted load balancing technique to cycle through the IP addresses corresponding to a group of servers in a cluster

pros:

simple

inexpensive

easy to implement

cons:

No knowledge about server availability

2. Hardware Load balancing

Here TCP/IP packets are routed to various servers in a cluster by providing a robust topology with high availability

Pros:

Uses circuit level network gateway to route traffic

Cons:

Higher costs compared to software versions

3.    Software Load Balancing

It comes as an integrated component of expensive web server and application server software packages

pros:

Cheaper than hardware load balancers

More configurable based on requirements

Incorporate intelligent routing based on multiple input parameters

Cons:

Need to provide additional hardware to isolate the load balancer.