The UDC was first introduced by Palo Alto, California-based HP in late 2001 as a concept, rather than a product, and Philips Semiconductor has become the first company to actually build and run one of the data center infrastructures.
Until now, in order to demonstrate the concept of its utility computing vision, HP has relied on outsourcing services deals with the likes of Bank of Ireland [IRE] and Procter & Gamble [PG], which use UDC capabilities as part of outsourcing deals.
Essentially a UDC is a flexible, virtualized data center that is pre-built to be responsive to a business’s processing requirements and utilizes metering technologies so that users of computing power can be billed, or at least measured, by how much they use.
Designed to be assembled using both new and existing technologies, the UDC is more about making the best use of resources to deliver business applications and services than it is about making use of the latest hardware.
One of the greatest arguments for moving to utility computing is that companies pay for what they use, with typical utilization rates seeing 40% of servers being used at 40% capacity.
This situation has arisen due to companies deploying new hardware for each application, along with new test and back-up environments. With a UDC, in order to add a new service an administrator just has to identify and reconfigure existing available resources and allocate them to the application.
HP’s OpenView service management software also allows for administrators to deploy applications based on rules to automatically allow for peaks and troughs in demand.
A UDC does not have to be built from scratch, and indeed is unlikely to be if a customer is aiming to reduce costs, and Philips has utilized existing servers as part of its UDC. Philips also took the opportunity to rethink some operational processes to manage its IT resources as a service, employing HP’s service management methodology.
This article was based on material originally published by ComputerWire.
