It is important to have a familiarity with the OSI 7 layer stack model when considering what type of traffic or devices can be load balanced. This knowledge is important because load balancing of traffic and, for that matter, devices is normally performed at Layer 2 (Data), Layer 4 (Transport) and Layer 7 (Application). It is also important to consider what information you require from your load balancers. For example, a business may need a health check on the performance of servers and applications, and/or information about server loading levels together with the ability to use load balancers to optimize the bandwidth that is available.

If you use an ADSL router at home you may discover that you are already experiencing the benefits of load balancing (multiplexing) without even knowing it. The better ADSL routers load balance according to the applications that are running and the number of users currently connected. They are able to support a number of users and can identify the different types of applications, giving priority to VoIP phone calls over video player content such as BBC iPlayer and content streaming over email download and upload. While the argument that you get what you pay for does apply, you could be surprised to find that your ADSL router costing less than $60 also includes load balancing as well.

Moving from the home or the small business to medium or larger businesses and organizations, the argument for installing network load balancers internally becomes more compelling. In this scenario the servers that host the business-critical data are the subject for evaluating the value of load balancing. When an organization has two servers or more, load balancing becomes a valid argument for discussion in order to define a company policy. The first thing to do is to define where the company mission -critical content servers are located.

Even if your organization has decided to put your servers, for example Microsoft Exchange, web site or ecommerce content in the cloud, it is important to confirm that your cloud service provider has made sure that a load balanced network architecture has been implemented to ensure best performance and 24/7 uptime. Many cloud service provides today provide this as a value-added service to their "VIP" clients-take a moment to check the charges for this service and make sure that you have an SLA that works for you.

There is a misconception that in-house load balancers can only be purchased for extremely large amounts of money and that you need to have either thousands of employees and racks of servers generating content to even think about an in house load balancer deployment. This is a misconception because you can actually install an industry-leading award-winning load balancer from as little as $1850 that includes a full year of technical support in the purchase price. The key requirement is to be able to scale the load balancers that you need according to both budget available and performance. Load balancers are capable of balancing diverse content including Microsoft Exchange, Microsoft Lync and Microsoft RDS as well as web content delivery and ecommerce portals.

Load balancing at Layer 4 is pretty common to all network load balancers; however, be sure to ask potential vendors whether the load balancer being recommended is also capable of health checking at Layer 7, the application layer. This is vitally important as the actual performance of each individual application can be monitored rather than the overall performance of the server, thereby greatly improving load balancing performance. As a good recent example, Microsoft initially advised that future users of Microsoft Exchange 2013 would be able to return to using Windows Network Load Balancer. On reflection, Microsoft now recommends that the suggestion to use hardware network load balancers for Exchange 2010 also applies to Exchange 2013 in order to benefit from the service-aware Layer 7 health checking that load balancers range offers.