Panel Discussion: Pedal to the Metal: Accelerating Big Data with Bare-Metal Cloud

No ratings

Presented at UP 2014 by

Satish Hemachandran

Shah Rahman

Arjun Ramaratnam

Young Paik

Despite fanfare around big data and the cloud, are traditional virtual cloud servers the right choice for big data apps? As customers embrace big data, they are realizing that to make a real-time business impact, they need their big data to behave like “fast data.” Traditional virtual cloud servers must share resources among multiple tenants and also include a hypervisor layer that creates additional processing overhead, and are therefore not the right choice to address the high-performance needs of many businesses. This is especially true for big data apps handling massive transactional workloads in real-time, in areas like digital marketing, eCommerce, online gaming, among others. These applications are helping drive the rapid emergence of bare-metal cloud as an alternative to public cloud. This involves non-virtualized dedicated servers that feature the same on-demand scaling, automation, self-provisioning and hourly billing as virtualized cloud. This panel will be moderated by Satish Hemachandran, VP of product management at Internap, and will include end-user organizations that are deploying large-scale bare metal environments to meet their high-performance needs. Attendees will draw the following takeaways: Learn the role of bare-metal cloud and database technologies in ensuring real-time data delivery and analysis, while delivering cost and management benefits. Hear firsthand how organizations are utilizing bare-metal architecture and the advantages it has brought them. Understand the ideal big data/fast data use cases and workloads for both public and bare metal cloud; where price: performance of public cloud can begin to break down; and leave with a list of key criteria to consider when deciding which model is right to support a given business.