The Transformational Promise of Per-Application ADCs

May 23, 2019 11:09:45 AM

Today’s cloud-centric enterprises require agile infrastructure that can scale up and down as capacity requirements evolve. Nowhere is this shift in infrastructure requirements more apparent than in the world of application delivery controllers (ADCs) and load balancers. Today’s enterprises are shifting away from monolithic ADC appliances in favor of lightweight, per-application software ADCs and load balancers.

 Scale On-Demand, Rather Than Build Three Years from Now

 Per-application ADCs offer simplicity and cost savings on the engineering side. During the days of monolithic appliances, network engineers would install ADCs with the throughput and processing power to serve large numbers of applications from a single box. Given the three-to-five-year lifespan of a typical ADC appliance, engineers had to deploy enough capacity on day one to support traffic growth over those three to five years. This meant that enterprises typically paid for capacity they didn’t need for at least a couple of years.

 Per-application ADCs are deployed and decommissioned on demand. As application consumption grows, engineers can upgrade the ADC software serving those applications without making changes to the rest of the network. This saves on cost and complexity. 

 Change Management Simplified

 Per-application ADCs reduce change management complexity in comparison to monolithic appliances. When one large appliance is serving dozens of applications, the blast radius of a mistake is large. If an engineer commits an error while reconfiguring an ADC for one application, that error can have a ripple effect on other applications served by the same box. In a best-case scenario, this situation leads to slow implementation of change tickets, which protects against mistakes but prolongs the time-to-market for new or upgraded applications. In a worst-case scenario, this situation leads to multiple application failures, costing a business millions of dollars.

Per-application ADCs typically serve only a single application or one component of an application. When engineers make a change to the ADC software, the risk to other applications that go untouched by that ADC instance is minimal. Furthermore, the features in use on any single instance of an ADC will also decrease, since only a single application is in play. This also reduces management complexity.

Cloud Migration Simplified

Moving an application from a private data center to the public cloud also becomes less complex with per-application ADCs. Emulating the application services in the public cloud, offered by a monolithic on-premises appliance, is a difficult exercise. There is no appliance in the public cloud environment where you can cut and paste your configuration files. Engineers will have to start from scratch in a software-only environment.

 An enterprise that uses per-application ADC instances in the private data center will have less heavy lifting to do. They already operate in a software environment. Depending on the vendor, spinning up per-application ADC instances in the cloud can be relatively trivial. If your ADC vendor supports multi-cloud management and orchestration, this migration of ADC services from the private to public domain can be somewhat trivial. 

 More Granular Network Operations

Finally, per-application ADCs can offer better insights into application health and performance. A monolithic ADC typically serves multiple applications, so any performance data collected from the box will require initial filtering of data unrelated to whatever performance problem network operations is trying to gather information about. This places more of a burden on performance management tools.

A per-application ADC usually serves a single application, so most of the performance data collected from it will be pertinent to that application, reducing overhead for analytics tools.

Kemp Technologies’ per-application ADC, the LoadMaster, offers all of the advantages covered in this blog. LoadMaster is also the foundation of the Kemp 360 AX Fabric, an end-to-end solution of multi-cloud application delivery infrastructure, including orchestration and analytics. Read EMA’s impact brief about the Kemp 360 AX Fabric, or learn more about Kemp solutions at kemp.ax.

Shamus McGillicuddy

Written by Shamus McGillicuddy

Shamus is the research director for EMA's network management practice. He has more than twelve years of experience in the IT industry as an analyst and journalist. Prior to joining EMA, Shamus was the news director for TechTarget's networking publications. He led the news team's coverage of all networking topics, from the infrastructure layer to the management layer. He has published hundreds of articles about network technology, and he was a founding editor of TechTarget's website SearchSDN.com, a leading resource for technical information and news on the software-defined networking industry.

    Lists by Topic

    see all

    Posts by Topic

    see all

    Recent Posts