Differences Between Legacy Application Delivery and Application Delivery Fabric

Posted on

The way applications are developed and delivered is rapidly changing. The adoption of Agile methodology, DevOps practices, and multi-cloud hybrid platform delivery means that the load balancing infrastructure practices of the past are often no longer fit for purpose.

Application Delivery in the Past

What IT Operations have typically done to deliver load balancing is to specify a shared sizeable multi-tenant load balancer for each site hosting applications, or they have deployed dedicated load balancers for each application on each site. In many cases, a combination of shared and dedicated load balancers has been used, often in pairs to provide resilience. This approach is inefficient as the estimated maximum traffic requirements for each load balancer has to be worked out and paid for in advance. This path leads to organisations paying for unused capacity for much of the time.

Application Delivery Fabric

A better way to approach load balancing in the new agile and multi-cloud world is to think of the complete application delivery infrastructure as a single fabric. Irrespective of the delivery location for an organisation’s applications – data centres, private cloud, or from multiple public cloud providers, tracking of the load balancing traffic should include all locations. Choosing a load balancing solution that allows for aggregated traffic monitoring and billing over all sites means that you only have to pay for the capacity used across the whole organisation.

MELA Licensing

MELA (Metered Enterprise Licensing Agreement) from KEMP is designed to deliver this licensing model. With MELA the peak monthly usage of each load balancer instance is measured, and the sum of these peaks is used to determine the overall usage of the whole application delivery fabric. Unlike the past multi-tenancy load balancer model there are no hard limits on the number of load balancer instances that can be deployed, and each load balancer does not have any capacity limits. The MELA model means that the most suitable load balancing deployment can be used, and be changed as required, while still only paying for the overall traffic used. Additionally, unexpected spikes in throughput will not lead to unforeseen additional costs as outliers in traffic use each month (based on a 95-percentile model) are discarded when calculating overall traffic used.

Monitoring Multiple Vendors

In addition to delivering a flexible licensing model for load balancing the application delivery fabric, we also provide solutions that integrate the monitoring and management of third-party load balancing solutions via a single management console. KEMP 360 Vision prevents issues before they happen by correlating data from the entire application delivery fabric and converting the results into actionable insights for operations staff. KEMP 360 Central allows for multi-platform deployment, management, and monitoring of LoadMaster, F5, Nginx, and other load balancers across infrastructure on VMware, Hyper-V, or Xen, as well as on Azure and AWS public cloud. It also has a built-in automation engine to allow load balancing operations to be incorporated into DevOps and other workflows.


Selecting KEMP solutions to deliver your application delivery fabric and monitoring infrastructure will give your organisation a modern, cost-effective, cross-platform, smart system that provides a better service to your users and makes your IT staff more productive. Contact us today for more details or to discuss your needs.

Posted on

Maurice McMullin

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