Whether you're administering a highly virtualized enterprise data center or a distributed computing cluster, the application requirements can quickly outpace the available technology (CPUs, memory, network I/O, storage and security services). This can result in performance issues, constant upgrade cycles, oversubscription of some resources, underutilization of others, and numerous maintenance and management issues. Aprius systems remove network interface and services bottlenecks allowing virtualized and distributed data center environments to deliver the same or better application performance as non-virtualized environments.

The Data Center Today

Aprius systems enable server, storage and network administrators to take a fresh approach to rack-level connectivity. This new category of data center systems consolidates and virtualizes network interfaces and services on a cost-effective, high-performance PCI Express-based fabric, while preserving all of the traffic bandwidth, priority and personality of each protocol with extremely low latency.

For example, bottlenecks can arise when application workloads on a server "burst", but do not have sufficient network bandwidth, or fast enough access to storage, or sufficient computing resources to satisfy their need. One way to address these bottlenecks is to statically over provision each server with all the resources (cards) it could possibly need to meet peak load. An alternative is to provide these network interfaces and services 'on-demand' to the servers, giving each server access to the resources it needs, when it needs it.

Just-in-time provisioning is far more cost-effective than just-in-case provisioning.

This allows the data center manager to dramatically simplify, provision and manage all resources at the rack or cluster level rather than at the individual server or blade chassis level, eliminating the need to manage various, discrete hardware configurations within each server. A single set of data and storage network resources can be provisioned and shared by the entire rack of servers, simplifying data center topologies and significantly reducing capital and operating expenses.

The Aprius Vision

To read industry coverage of network consolidation, I/O Virtualization and related trends, please click here

For more information please click here