T-Systems cuts size of datacentre estate by 85% through technology refresh

T-Systems, the IT services arm of Deutsche Telekom, claims to have completed the world’s largest datacentre rationalisation project, which has seen the number of facilities it operates drop from 89 to 13.

The downsizing in its datacentre footprint has taken six years to complete and has not, the company said in a statement, come at the expense of the amount of compute power and storage capacity it can offer its clients.

“We are particularly proud of the fact we were able to implement the transformation without any loss of quality in customer operations,” said Jörn Kellermann, senior vice-president of global IT operations at T-Systems International.

The firm claims the work is saving the firm a “three-digit million Euro sum” a year, as a direct result of deploying datacentre technologies that are far more efficient.

In fact, the company claims, despite the reduction in datacentre floor space, its storage and compute capacity has increased by 25% overall, before going on to reveal the work required 5,200 separate migrate projects involving 23,600 servers.

It is also claimed the project has contributed to a 56% reduction in the amount of carbon emissions generated by the firm’s server farm estate.

“Our datacentres are highly scalable and automated,” said Kellermann. “The infrastructure is consistently geared to the cloud.”

While 60% of the servers involved in the migration process are known to have been located in Germany, the firm has also – in recent years – embarked on a revamp of its UK datacentre setup.

This work has seen it commit to a five-year decommissioning of one of its legacy datacentres in Feltham, Middlesex, and migrating customers to Virtus’s colocation facility in Hayes. It also known to have colocation capacity set aside at Equinix’s datacentres in Slough.

Speaking to Computer Weekly back in spring 2016 about this specific project, the company said the emergence of higher-performing datacentre technologies, such as blade servers, meant it could get far more compute capacity within a much smaller footprint, paving the way for its Feltham facility to be taken out of use.

The company’s datacentres are used to provide its customers with access to private and public cloud environments, based on Microsoft, Amazon and OpenStack technologies.