Data virtualisation - an important data centre building block

A recent study by Emerson Network Power has revealed that the average power density per rack will increase significantly in the coming years, creating new challenges in thermal management.

  • 8 years ago Posted in
 The study, which interviewed 800 global data centre specialists about the data centre of the future, examined expectations of how power requirements would evolve between now and 2025. The results highlighted improvements in energy efficiency as a big issue for almost two thirds of respondents, with many expecting that data centres will leverage emerging technologies to consume significantly less energy over the next ten years while providing the same computing power.
 
"Copy data virtualisation has the power to make data management dramatically more power-efficient," explains Ash Ashutosh, Founder & CEO of Actifio. "Few would argue that the efficiencies enabled by virtualisation technology will play an essential role in the data centre of the future, though ironically most seem focused on the virtualisation of compute and networking as the places to begin. It’s called a “DATA Centre” for a reason, and virtualising the data is absolutely key to decoupling the growth of primary data from the growth of power-hungry storage."
 
According to the Emerson Network Power study, there is a lack of "transparency over the different systems". In many data centres, "silo-like structures" still prevail because individual systems communicate with different protocols. The big challenge is to "develop solutions that collect data across all systems, and translate them, so that one management solution can work with it." Data virtualisation makes it possible already today to control access to application data centrally, so automatically there is better transparency for data management.

Silo structures have much to do with the traditional approach of data management, where copies of the same data are held for different purposes: analysis, backup, development, testing, disaster recovery etc. These data silos can be replaced by a central Copy Data Management platform. Data will then be recorded directly from within applications, centrally stored, managed and globally de-duplicated. A master ‘golden’ copy is created and saved. Changed blocks are always stored incrementally, so the master copy is always up to date. Whenever required, this master copy can generate countless virtual copies for any purpose. Fewer memory resources are used, since there is only storage space required for one physical master copy. Virtual data copies are available at any time - even faster than in conventional memory variants.
Infinidat has achieved significant milestones in an aggressive expansion of its channel...
NetApp extends its collaboration to accelerate Ducati Corse’s digital transformation and deliver...
Infinidat says that Richard Bradbury has been appointed SVP, EMEA & APJ. Leveraging his extensive...
Just months after announcing the availability of its backup as a service offering (BaaS), Cohesity...
Tintri®, a DDN® subsidiary and the leading provider of auto adaptive, workload intelligent...
Veritas InfoScale native deployment in Kubernetes environments, including Red Hat OpenShift, will...
Portworx by Pure Storage delivers scalability, availability, and better security to data rich...
The combination of Storage Made Easy’s Enterprise File Fabric and Object Storage from Cloudian...