FindMyHost

Web Hosting Features

Setting up load balancing with cloud VPS

Setting up load balancing with cloud VPS
November 12
08:00 2013

A critical goal for every website is reliability, generally measured by the percentage of time a site is up and operational during a set period of time. Also known as uptime, reliability is an important way that businesses or organizations keep customers and users of the website happy. Ideally, every organization strives for 100 percent uptime. Cloud computing using a virtual private server (VPS) helps makes that goal a reality.

The key to 100 percent uptime is redundancy. If a single piece of equipment fails, a backup is activated, and the customer doesn’t experience any downtime. Strong systems have multiple redundancies. It’s easy to see how VPS fits into this process. A VPS is a cost-effective and efficient way for businesses to outsource their server needs, by only paying for a portion of a larger server. Using VPS, a company can establish multiple backup servers for much lower costs than buying dedicated servers. VPS also allows companies to easily distribute their virtual servers geographically, ensuring that a single catastrophic event (such as a hurricane or earthquake) can’t bring down their entire systems.

Splitting up the work
Redundancy is only part of the answer for 100 percent uptime, though. In general, organizations don’t want to establish backup systems that sit idle while other systems perform the majority of the work. The active system is strained, while the backup isn’t getting much use. Everyone who accesses the system is hitting the same active server, which gets inundated during peak usage times, and may slow down, deny page requests, or even fail.

It’s better to divide the workload among all the servers. Effectively distributing the processing between VPS systems is called load balancing. A load balancer is established at a single network address, and all traffic is routed to that point. The load balancer distributes incoming requests among the available VPS systems, and the site functions at high speed even during times of peak activity.

Methods of load balancing
Load balancers work in a number of different ways, based on the needs of the user, the capacities of the VPS systems used, and other factors. In general, when a load balancer receives a request, it chooses an appropriate server based on a specific algorithm. While there are many methods of load balancing, the common algorithms include:

  • Round robin. This is the most commonly used load balancing method. Round robin simply organizes the available VPS systems in a standard rotation, and passes each request to the next VPS in line. This distributes all requests evenly across the available servers, and it works well if all the systems are roughly equivalent in memory and processing power. A weighted round robin system forwards new requests based on both the list of servers and the individual capacities of those servers.
  • Least connections. This method balances the server load by passing new requests to the system that has the least number of active connections. Thus, the server doing the least amount of “work” is getting the new traffic. Again, this works well if the VPS systems are similar. If VPS systems have different capacities, however, a weighted least connection method can be used. This distributes the load based on both the number of active connections and the server’s capacity. For example, consider two VPS systems, one with a capacity of 100 connections, and another with a capacity of 200. If the first server has 80 active connections, and the second has 81, a load balancer using least connections will pass the next request to the smaller capacity server. A load balancer using weighted least connections will calculate the capacity of both servers (in this example, the first is at 80 percent and the second is at 40 percent) and pass the next request to the second server, which has the greater capacity.
  • Fastest response time. With the fastest method, the load balancer assigns the connection based on the fastest response time of all the servers in group. This is often used when the servers are on different types of networks.
  • Perceptive. Load balancers that use this method route traffic to the VPS in the group based on both the current number of open connections between the load balancer and the server, and the response times of the servers. Basically, when there is little difference in response times, the perceptive load balancer will distribute new requests using the least connection method. When the difference in server response times increases, the perceptive load balancer recognizes this and begins to route traffic to the servers with faster response times.

Persistence
Another important concept in load balancing is persistence, often referred to as stickiness. For many websites, it doesn’t matter if individual requests are sent to different servers. If a user is simply clicking through a website, it doesn’t matter if each page is sent from different servers.

However, if the website is conducting a transaction, such as adding items to a shopping cart and then checking out, the system needs to follow activities throughout the transaction. Each step could potentially be executed on a different server. The load balancer must be able to track the steps and identify the servers handling each portion of the transaction, typically by identifying such things as the user’s IP address, cookies issued by the application, and secure socket layer (SSL) sessions.

Always connected
Today, everyone expects website to be available and operating well at all times, even during periods of high traffic. A load balancing application coupled with multiple VPS allows organizations to meet that expectation by improving uptime and website performance. Load balancing is a critical piece of a strong website infrastructure.

To learn more about creating a load balance with a cloud VPS, contact Atlantic.net today. For almost 20 years, Atlantic.net has strategically evolved into a market-leading Cloud Hosting provider. Specializing in cloud servers, colocation, dedicated servers and virtualization hosting services, we are dedicated to implementing tailored hosting solutions that enable you to enjoy the benefits of cost savings. To learn more about Atalntic.net, visit us online at www.atlantic.net.

About Author

FindMyHost Editor

FindMyHost Editor

The 'Web Hosting Blog' at FindMyHost.com was established to provide the web hosting industry with the very latest in news, technology, interviews, event information and more. About FindMyHost.com: Launched in January 2001 to protect Web Host Consumers and Web Developers from making the wrong choice when choosing a Web host. FindMyHost.com showcases a selection of web hosting companies who have undergone our Approved Host program testing and provides reviews from customers.

Related Articles

Special Offers: