How to dis­trib­ute server workloads with a load balancer

Using Load Balancing allows server access to be evenly dis­trib­uted across various hardware resources through a technical instance called a Load Balancer. This ensures stable access speeds, which is crucial for gen­er­at­ing leads and customers.

What is load balancing?

Load balancing is es­pe­cial­ly popular in server tech­nol­o­gy and describes a method where requests are dis­trib­uted among different servers in the back­ground without users noticing. The load balancer used for this can be im­ple­ment­ed as hardware or software. It assigns several servers to a domain without causing address conflicts and is accessed under the public domain.

The sub­or­di­nate web servers are named with the domain additions www1, www2, www3, etc. This allows load balancing to make a website available under the same URL, even though multiple servers are behind it. This prevents server overload as external requests are dis­trib­uted across various physical machines within the cluster. Users generally do not notice this because the dis­tri­b­u­tion of requests happens based on complex al­go­rithms in the back­ground.

Load balancing can also play a role beyond web servers, such as with computers that operate with multiple proces­sors. In such cases, the load balancer ensures that demands are evenly dis­trib­uted across the different proces­sors to generate more computing power. Load balancers are also in demand when using container software like Ku­ber­netes. Here, they ensure that workloads are ef­fi­cient­ly dis­trib­uted across various pods.

How does load balancing work?

Requests to a web server, e.g. in the form of page views, initially run on the load balancer. This then takes over the load dis­tri­b­u­tion by for­ward­ing the access attempts to different servers. The load balancer itself can be im­ple­ment­ed as hardware or software, but the principle remains the same: a request reaches the load balancer and, depending on the method used, the device or software forwards the data to the relevant server.

The technical basis is the DNS procedure: Users access a website only through a URL. This is converted into an IP address using the DNS, which then points to the load balancer. Ideally, this process goes unnoticed by users.

Close

What are the benefits of balanced load balancing?

The three main ad­van­tages of a well-deployed load balancer are as follows:

• Optimized access times: Dis­trib­ut­ing traffic across multiple servers can result in shorter access times, even when there are many requests at the same time.
• Higher fault tolerance: A load balancer increases fault tolerance as the traffic from a slow server is au­to­mat­i­cal­ly redi­rect­ed to other servers in the cluster. If a server is un­avail­able, the hosted website remains ac­ces­si­ble.
• Sim­pli­fied system main­te­nance: Load balancing greatly supports main­tain­ing a server system. Con­fig­u­ra­tions and updates can be performed while the servers are running without no­tice­able per­for­mance loss. The load balancer detects main­te­nance states and redirects requests ac­cord­ing­ly.

What are the different types of load balancing?

How incoming requests are dis­trib­uted depends on the choice of algorithm. Popular al­go­rithms for load balancing are: Round Robin, Weighted Round Robin, Least Con­nec­tions, and Weighted Least Con­nec­tions.

Round Robin

Round Robin describes a procedure whereby incoming server requests are processed in a queue by the load balancer and dis­trib­uted through­out a series of servers. Each new request is assigned to the next server in the sequence. As a result, access requests can be evenly spread across the load balancing cluster. No matter how urgent the request or severity of the server load, Round Robin treats all processes the same. Load balancers that operate on the round-robin principle is es­pe­cial­ly suitable for en­vi­ron­ments where all instances have about the same resources available.

Weighted Round Robin

The weak­ness­es of the classic round-robin algorithm in het­ero­ge­neous server clusters can be balanced with a weighted round-robin dis­tri­b­u­tion. Incoming requests are dis­trib­uted con­sid­er­ing the static weighting of each server. This weighting is defined in advance by the ad­min­is­tra­tor.

The most powerful server, for example, can be assigned the value “100”, while less efficient servers are given the value “50”. In such a setup, the server weighted “100” would receive two requests per round from the load balancer, whereas the server weighted “50” would receive only one request. Weighted Round Robin should primarily be used in load balancing when the servers in the cluster have different resources available.

Least Con­nec­tions

Both round-robin al­go­rithms do not take into account, during the serial dis­tri­b­u­tion of server requests by the load balancer, how many con­nec­tions the sub­or­di­nate servers must maintain over a certain period. This can result in several con­nec­tions piling up on one server in the cluster. This leads to the server becoming over­loaded, even if it handles fewer con­nec­tions than others. The least-con­nec­tions algorithm protects against this. It dis­trib­utes requests based on the existing con­nec­tions of each server—the one with the fewest active con­nec­tions gets the next request from the load balancer. This load-balancing method is rec­om­mend­ed for ho­mo­ge­neous server clusters where all computers have com­pa­ra­ble resources available.

Weighted Least Con­nec­tions

If a server cluster has different ca­pac­i­ties, instead of the least-con­nec­tions algorithm, load balancing based on the weighted dis­tri­b­u­tion of existing con­nec­tions should be used. This considers both the number of active con­nec­tions a server has and the weighting defined by the ad­min­is­tra­tor. This ensures a balanced load dis­tri­b­u­tion within the server cluster. New requests are au­to­mat­i­cal­ly assigned by the load balancer to those servers whose ratio of active con­nec­tions to their re­spec­tive server weighting suggests the least current load.

What problems can arise when using load balancing?

Es­pe­cial­ly in the E-Commerce sector, load balancing often faces chal­lenges. Here’s an example: Website visitors add items to their shopping cart that they wish to purchase. These items remain saved for the duration of a session, re­gard­less of which page the users navigate to within the online mar­ket­place. A typical load balancer would dis­trib­ute requests across different servers. This means the contents of the cart would be lost.

To solve this problem, two ap­proach­es are con­ceiv­able. First, the load balancer can respond to the IP address of the users. Then, for example, requests from the same IP address are always directed to the same server. Another method would be to read a session ID from the request itself to determine which server the request must be sent to.

Why are load balancers so important?

If you earn your money over the internet, your business can’t afford a server outage. And if you use only one server and it crashes due to overload, your website is no longer ac­ces­si­ble to potential customers. This leads to several problems: For one, you can’t generate revenue during the overload. Services can’t be booked and purchases can’t be made. Ad­di­tion­al­ly, trust from (potential) customers decreases. It’s doubly bad for users who ex­pe­ri­ence a server overload during the order process. There is often great un­cer­tain­ty, and those affected don’t know if the order actually arrived and was recorded in the system.

Even if you don’t offer services directly over the internet, your website should be ac­ces­si­ble at all times. A website is one of the primary channels for accessing in­for­ma­tion. If potential customers look for details about your company online and can’t reach your site, they are more likely to consider your com­pe­ti­tion. You can minimize such risks with a load balancer.

How to implement load balancing in your business

Load balancing can be im­ple­ment­ed using both hardware and software solutions on a virtual server. Pro­fes­sion­al complete packages are offered by numerous providers either as In­fra­struc­ture-as-a-Service (IaaS) or as a network component for your own IT in­fra­struc­ture.

Since the ac­qui­si­tion of pro­pri­etary load balancers usually involves high costs, smaller companies often turn to open-source solutions like NGINX. It offers a cost-effective way to ensure high avail­abil­i­ty for your company website or other web projects through efficient load dis­tri­b­u­tion within the server network. In the web hosting sector, load balancing is also often offered as an add-on feature for cloud servers.