How to Choose a Load Balancer: 5 Tips for Finding the Best Load Balancing Solution

Posted on

If you were in IT in the mid-90s, you may have heard of load balancing. These were low-level hardware devices designed to distribute traffic across the network and focused largely on traffic between clients and servers. 

Today’s load balancers, however, are a whole new world. Instead of balancing just between clients and servers — and amongst servers themselves — load balancers can be equipped to handle almost anything, including: 

  • Virtual machines 
  • Applications 
  • Cloud services 

Just as the environment of digital infrastructures has dramatically changed over the last several decades, so too has the nature of load balancing. It’s not the mid-1990s anymore, so let’s walk through load balancing capabilities in 2022.  

Download the Load Balancing Buyer’s Guide

What is a Load Balancer?

Though the scope and abilities of load balancing has changed in tandem with the demands of modern digital environments and IT, the concept is the same: load balancers balance the load of traffic and work requests between resources on physical or virtual servers in a network.  

Using algorithms and intelligent network monitoring running constant health checks on servers, load balancing ensures no server in the network is over-allocated beyond its limits with work requests. Though not all servers are created equal, in a sense, load balancing is a democratizer of workloads, making sure all servers on the network are getting generally equal, balanced work distributions.  

“A load balancer can be deployed as software or hardware to a device that distributes connections from clients between a set of servers. A load balancer acts as a ‘reverse-proxy’ to represent the application servers to the client through a virtual IP address (VIP). This technology is known as server load balancing (SLB). SLB is designed for pools of application servers within a single site or local area network (LAN),” as explained in our What is Load Balancing? Blog. 

While server load balancing is a primary function and technological underpinning on modern networks, there is a lot more to it. Server load balancers can provide additional functionality based on their role in the network, including allowing and/or denying certain security, authenticating IP addresses and scaling loads up or down based on availability or location (GSLB).  

One of the biggest areas of use is applications, providing not only availability but also scalability that allows the application to scale beyond the capacity of a single server. The load balancer works to steer the traffic to a pool of available servers through various load balancing algorithms. If more resources are needed, additional servers can be added and expand to include scaling up and down as demands change. 

Load balancers health check the application on the server to determine its availability. If the health check fails, the load balancer takes that instance of the application out of its pool of available servers. When the application comes back online, the health check validates its availability, and the server is put back into the availability pool. 

Today’s load balancers are extremely advanced and can run software that provides a large sum of functions. Load balancers not only sit between the client and application server but also manage this connection, performing content switching and providing content-based security like web application firewalls (WAF) and two factor authentication (2FA). 

What do Load Balancers do?

Availability, scalability and security are all well and good, but what else can load balancers do? Well, as a reverse-proxy acting as a multi-functional access point directing and controlling traffic between clients and servers, load balancers can: 

  • Automatically detect server failures and redirect client traffic 
  • Allow for server maintenance without any impact 
  • Provide automated disaster recovery to backup sites 
  • Add and remove application servers without disruption 
  • Monitor and block malicious content

Five Key Load Balancing Types

With services and abilities as varied as that, you might be wondering what types of load balancers are out there. Here are five key types of load balancers: 

Application Load Balancers

Application load balancers — or Layer 7 load balancers — were one of the first load balancers to develop out of the original network and hardware-based balancers that hit the market in the 90s. Unlike their predecessors, these load balancers provide content awareness and content switching, looking into content payloads via URL, HTTP headers and more.  

Elastic Load Balancers

More sophisticated and flexible (and useful in the Cloud) elastic load balancer (ELB) solutions offer operators scalable capacities based on traffic requirements in real time. As elastic as their name indicates, ELBs are at your disposal, capable of scaling traffic to applications as demands evolve over time, automatically, and on-demand, leveraging request routing algorithms to distribute incoming application traffic across multiple instances.  

Global Server Load Balancing

On the other hand, global server load balancing (GSLB) is a slightly different technology than the average Layer 4-7 load balancer — and most load balancers today offer it as a component. GSLB is based on domain name server (DNS) and acts as a DNS proxy providing responses in — you guessed it — real time. Explained in our What is Load Balancing blog: “It is easiest to think of GSLB as a dynamic DNS technology that manages and monitors multiple sites through configurations and health checks.” 

Hardware vs Software vs Virtual Load Balancing

Load balancers originated as hardware solutions, and as such hardware-based load balancers are the longest tenured in the load balancing marketplace. As simple turn-key solutions, these load balancers skirt the requirements of software-based solutions such as hypervisors and commercial off-the-shelf hardware. 

But as network technologies evolved and moved towards more software-defined, virtualization, and Cloud technologies, load balancers changed in tow. “Software-based load balancing solutions offer flexibility and the ability to integrate into the virtualization orchestration solutions. Some environments such as the Cloud, which often uses DevOps and/or CI/CD processes, require software solutions. Therefore, the flexibility and integration of a software load balancer is more suited for these environments,” as stated on our What is Load Balancing page.  

Network Server Load Balancers 

Network server load balancers are the original load balancers (also known as Layer 4 load balancers), the ones that first entered the market in the mid-90s to pool server resources to support the explosive surge of traffic on the internet. These load balancers are designed to manage connections based on the 5-tuple packet headers – source IP, destination IP, source port, destination port, and IP protocol.  

So, with all these options at your disposal, how do you pick the load balancer that is right for you? 

Five Steps for Selecting the Perfect Load Balancer

When it comes to selecting the best load balancing solution for your business, there are five key steps to identifying the right load balancer for your environment.  

Step One: Identify the Business Goals of the Application 

The Load Balancing Buyer’s Guide says that the perfect load balancer for your business ultimately depends on your business requirements and how the application you are servicing behaves. From the Buyer’s Guide: “Load balancing solutions should enable organizations to cost-effectively scale their operations,” while ensuring high availability and outstanding application experience (AX).  

According to Forrester Research, speed and agility is critical in this regard, with 91% of enterprises saying so. Because your load balancer is the proxy for your application, you can’t have a fast application without a fast load balancer. 

The proliferation of remote work has also changed the business demands of load balancers. With a need to support 24/7 access to business-critical applications and data, you need a load balancer that is “secure, quick to deploy, easy to manage and at a value point that fits future IT budgets.” Without a load balancer of this nature, you’re looking at a lot of unnecessary downtime.  

Step Two: Understand the Features and Capabilities You Require from a Load Balancing Solution

Step Two is all about understanding the real nitty-gritty details of what you’ll require from a load balancer. Key to this understanding is knowing your own network and application needs inside-out. How is your infrastructure supported? Do you need a hardware-based load balancer? Software-based? Or how about virtual? Is your application available in the Cloud marketplace? 

Load balancers are constantly performing health checks on servers. Since more advanced Open Systems Interconnection means more advanced health checks, you’ll need to know just how advanced your load balancer should be. Your options can run between Layer 3 ICMP health checks all the way to Layer 7 for any TCP/UDP port. 

Load balancers also need to be resilient enough to manage application connections even when they themselves fail. As application demand increases, you’ll need a load balancer that offers your application scalability by providing more load balancing resources, such as:  

  • Active/Standby high availability between pairs of load balancers 
  • Stateful connection failover to prevent the loss of existing active application connections  
  • Clustering beyond 1+1 availability to deliver high performance scalability  
  • Disaster recovery (DR) active/standby ability for multiple sites 
  • Active/Active site availability through dynamic global server load balancing (GSLB) 

Not to be overshadowed too is session persistence. Sometimes connections will need to be maintained to a specific server because of local content associated with the session, or different applications needing access to the same local session-specific information at once. Session persistence solves this, offering different parameters for: 

  • Client IP address 
  • TLS (SSL) Session ID 
  • Web/HTTP cookie and/or Session ID 
  • Port following for multiple applications (i.e., web store and checkout sites)”  

 To make the most informed decision on load balancers you’ll also need to consider different scheduling methods and encryption features, and know your management interface and security requirements among other considerations — all of which can be found in the Load Balancing Buyer’s Guide.  

In the meantime, let’s move on to Step Three, shall we? 

Step Three: Licensing Options

An important discussion is now looming between you and your CFO, as you two might have differing views on what is an acceptable price for a load balancer. With all the insights gleaned from Step One and Step Two in your back pocket, you now need to consider four different payment models.  

Perpetual Licensing 

A perpetual license — usually based on the maximum network throughput the product can support — means you own the solution outright, whether it is a piece of hardware or an instance of load balancing software.  

However, perhaps you only require a subscription with a scheduled payment plan you can cancel at any time. In that instance, you may find greater interest in a more flexible and innovative licensing method, like pooled or metered licensing, such as pay-as-you-go in the Cloud or 1-year or 3-year support subscriptions.  

Pooled Licensing 

Pooled licensing allows your company to purchase a certain amount of load balancing capacity and then distribute it throughout your environment at your discretion. This is a beneficial model where there is a need for multiple load balancers and the needs can vary based on application and time.  

For example, a company may acquire a 100Gbps pool and they can then distribute that into individual instances so long as the total capacity stays within 100Gbps. Capacity can be distributed and recovered so long as the pool license is valid. This is usually paid annually and includes support services. 

Metered Licensing 

Metered licensing is the most innovative model, offering a license capacity similar to pooled licensing, but the individual instances of the load balancer that are deployed only go against the capacity for the actual amount of application traffic being load balanced by the system. 

A business can acquire a 10Gbps metered license that allows them to load balance up to 10Gbps across any number of load balancing instances. Metered licensing includes support services and is usually billed monthly based on the past month’s consumption. The major use case of this form of licensing is the ability to exceed capacity. A license limit will never be applied and prevent customers from handling even the greatest traffic spikes.  

Step Four: Price vs. Value

Hold it, you’re not out of the pricing woods yet. Though all load balancers equalize workloads, not all load balancing solutions are created equally. Considering all the different prices and functions offered by different load balancing vendors, it’s important to prioritize exactly what is most valuable to your business needs, and not, as Progress Kemp advises, get caught up in the “fancy features you probably do not need.” 

Below is a chart that can help you compare load balancer prices between different vendors: 

You may also watch either of the informational videos for a side-by-side comparison of LoadMaster compared with other load balancing solutions:  

Step Five: Do More Research!

Hey, what do you expect — it can never hurt to do more research! After all, picking a technology solution is important and should not be undertaken lightly. That’s why when it comes to validating your choice, you need to compare your choices with those from third-party and customer review sites like Gartner Peer Insights that offer a clear picture of the load balancing market. Look with a goal of understanding that the cost, complexity, and support of the technology does not become a barrier to the effectiveness of the solution.  

Conclusion: Why LoadMaster

As the leading load balancer available on the market today, LoadMaster offers affordable load balancers available as virtual, software, and hardware-based. You won’t find a more flexible or capable option from any other vendor. After all, LoadMaster supports the most popular hypervisors including: 

  • VMWare 
  • Hyper-V 
  • XEN 
  • KVM 

LoadMaster is also certified by leading vendors including: 

  • Microsoft 
  • Cisco 
  • Dell 
  • VMWare 
  • SAP 
  • Citrix 

Not to be outworked, LoadMaster provides load balancing solutions for: 

  • Corporate email 
  • Unified Communications 
  • Collaborative work tools 
  • ERP/CRM and other workflow applications 
  • Web content 
  • eCommerce systems 
  • Self-provisioning applications 

Want to learn even more about load balancers? Read Progress Kemp’s Load Balancing Buyer’s Guide

Posted on

Maurice McMullin

Maurice McMullin was a Principal Product Marketing Manager at Progress Kemp.