There's no question that virtualized public clouds have revolutionized the technology landscape, giving cloud-savvy companies the leverage to scale at the speed of business while minimizing up-front operating expenses, reducing complexity and only paying for the capacity they need.
However, for certain types of applications, virtualized public clouds can present significant performance limitations due to the multi-tenant environment and processing overhead of the hypervisor layer, especially as traffic levels scale. This is particularly significant for the emerging generation of fast, big data applications, such as those that enable personalization and targeting through customer insights, social analytics and transactional analysis.
These applications require public cloud infrastructure that can guarantee consistently high speed and ultra-low latency to enable massive amounts of uninterrupted data ingestion and real-time analysis, as well as cost-efficiency at scale.
Amid these new pressures on traditional public cloud services, Internap surveyed nearly 250 internet infrastructure decision-makers to gain insight into cloud adoption, requirements and challenges. The results of the Cloud Landscape Report survey reveal the notable performance and price challenges users are encountering with virtual public cloud, particularly for real-time, data-intensive and scale-out applications.
The survey also highlights the widespread misperception that a public cloud must include virtualization, a notion that is being challenged by these new requirements and leading to interest in alternative forms of IaaS, such as a bare-metal public cloud.
Read the details of these survey findings in the “2014 Internap Cloud Landscape Report.”
