The new virtual IT data center must support multiple application and data workloads efficiently, while meeting performance requirements. IT administrators are looking to deliver the resources necessary – without over-provisioning equipment or experiencing performance bottlenecks.
The same trends described above that have been driving server consolidation are placing new demands on storage networks and storage subsystems. Current server technology supports multi-core processors, each capable of running 10 or more virtual machines simultaneously. This can result in over 1,000 virtual systems running in one rack, driving a massive increase in I/O density, performance and capacity requirements.
Along with the increase in density, virtual computing workloads are significantly more demanding than applications running in traditional environments. The combination of high density, driving high data rates, with more demanding workloads places a significant burden on the storage systems used to support these workloads. In order to reap the benefits of virtualization, it is imperative that I/O and storage systems are able to support these new, demanding workloads.
This paper covers the impact of virtual servers on storage and critical considerations for designing the storage environments. It will compare block versus storage, VMFS vs NFS and discuss strategies for storage efficiencies. The paper highlights the use of 3PAR for virtual environments and outlines features that enable an efficient environment.