Building The Next Generation Application Data-Defined Infrastructure

By Sushil Kumar , CMO, Robin Systems


One of the promises of Big Data is its unique ability to hone offers and products tailored to an individual’s needs. It’s been a few years and we’ve made baby steps toward that vision. For some time, developers, IT, marketers, and visionaries have tried to surmount the hurdles that are in the way of offering the right product, at the right time, and in the right place.

While the business value and transformative nature of the data-driven enterprise is unquestioned, the reality is that most companies struggle to get there. The big part of this challenge is infrastructure complexity. Modern data applicationsutilize a distributed computing paradigm to provide scalability and availability, but this comes at the cost of huge operational complexity. These applications are not only hard to deploy and manage, the inability of the current infrastructure to provide multi-tenancy leads to cluster sprawl, data duplication, and unsustainable cost inflation. Existing hypervisor-based virtualizationis of limited help because of unacceptable performance overhead and undesirable VM and software sprawl. This problem applies to cloud deployments as well, where dedicated resources and high-performance storage translates to higher cost.

Clearly a new application-centric paradigm is needed to streamline application lifecycles, enhance infrastructure efficiency and rationalizecosts.

Major retailers are the perfect proving groundfor such innovations. They are tasked with sorting through great troves of information while keeping the costs low. They are constantly looking for new technologies that can help them do more with less. It is no wonder, therefore, that these companies are so keen to use containers to change the way modern applications are built and operated. Consequently, we are stepping into theapplication-defined datacenter era that uses container-based server and storage virtualization to create an agile and high-performance applicationplatform. This highly dynamic and elastic platform enableshardware consolidation without any performance atrophy, has a dramatically simpler application lifecycle, and streamlines data management. For application users and developers it results in incredible gain in productivity and ease of use while it allows IT manager to unlock hidden capacity, maximize hardware utilization, and reduce costs.

A major retailer that adopted this new approach experienced a revolutionary benefit in managing its data-centric applications.This particular retailer has more than 260 million customers visiting 11,000 stores, and operates 12 active websites across the globe. Their online strategy is driven in large part through innovative and extensive use of Big Data technologies that produce data-driven customer insights. It is therefore no surprise that a leading retailer like this operates a very large Big Data environment processing many petabytes of data across thousands of bare-metal servers.

UsingRobin’s container-based server and storage platform software, this retailer saw 50 percent reduction in hardware footprint, improved application performance, and dramatic operational efficiency. By deploying applications within isolated containers, multiple clusters were consolidated on shared hardware, thus shrinking the compute footprint from 16 servers/320 cores to 10 servers/160 cores.

Container-powered virtual clusters were deployed self-serve within minutes – as opposed to weeks and months before. And by decoupling compute and storage, a virtualized data layer was created that enabled data sharing across clusters, thus eliminating unnecessary data duplication and avoiding storage wastage.More details can be found here.

These benefits are nothing sort of dramatic and clearly illustrate the transformative potential of a container-powered application-defined datacenter paradigm. It is for this reason that the use of containers is on rise for data applications. According to this survey, conducted across 200 IT decision makers,  containers are emerging as the preferred platform for running performance sensitive applications such as databases. This unstoppable trend is only going to gain momentum in the times to come.