Case study: Clifford Chance

Clifford Chance hopes its core system will perform better on the new virtualised platform

Law firm Clifford Chance hopes to achieve performance improvements via the use of virtualisation and a consolidation exercise that will reduce the number of servers used by the business by 90 per cent.

The decision to start the migration to a consolidated environment was prompted by the need to provide a scalable platform for the firm’s global practice management system (GPMS).

GPMS is the firm’s core application for the management of billing cycles worldwide. It performed well initially but was struggling to meet growing demand for its services and started to hold back the business, according to Clifford Chance.

After these limitations became apparent, the business realised there were benefits to be had from moving to a standardised server environment, such as improved performance and reduced cost and complexity.

The law firm then began a consolidation exercise worldwide, which is expected to cut costs by reducing the number of global datacentres from 10 to four. Use of virtualisation will enable the business to reduce the number of physical servers it supports by around 90 per cent.

Clifford Chance performed a proof of concept in May, using a Sun Sparc Enterprise M5000 server. This was followed by tests of the GPMS system running on Sun’s M9000 servers in its main datacentre in London and in its disaster recovery site in Germany.

The tests led to the decision to use three Sun servers for its application tier at both sites and one for its development and test environment. Fujitsu acquired and shipped the hardware to the datacentres in England and Germany.

Deployment of the new platform is expected to complete by early 2009. When the new setup is up and running, Clifford Chance expects to more than double the GPMS performance.

A boost in employee productivity and an increase in customer wins, as well as improved energy efficiency and a reduced carbon footprint are among the benefits anticipated post-implementation. Full return on investment on hardware costs is expected within two years.