Data Centre Estate: Retain, Replace or Relocate ?

Introduction

A public body can own servers that are still serviceable while its building, cooling, connectivity or support arrangements no longer meet its needs. An NHS trust may face an ageing server room; a council may be preparing to vacate offices; a government department may need to keep a legacy service running while a replacement application is years away. The infrastructure decision is rarely confined to where the equipment sits.

Data centre colocation offers one response: the organisation places its equipment in a specialist facility and buys the space, power, cooling and physical environment required to run it. But the same customer may instead renew its own infrastructure, purchase supplier-owned managed hosting, move selected workloads to public cloud or replace applications with software as a service (SaaS). These are different purchases, with different responsibilities and commitments.

This first briefing examines the range of choices, the structure and direction of the market, and the reasons demand varies between organisations. The question underlying the series is straightforward: how should buyers preserve essential services while leaving themselves room to change course?

What is actually moving — the equipment, the workload or the responsibility?

An organisation-owned data centre combines three things: physical premises, computing equipment and the people or suppliers operating the technology. It is possible to change one without changing the others. Direct colocation typically transfers the facilities requirement to a specialist operator but leaves ownership of servers, storage and network equipment with the customer. A managed colocation arrangement adds contracted operational support. Supplier-owned managed hosting transfers equipment provision and specified day-to-day activities to the hosting…