Extend the lifespan of an existing data center while controlling costs

The secure hosting of IT and telecom equipment is the raison d'être of a data center. As such, the first action to take to give it a new lease on life is to conduct a precise inventory of IT equipment. This assessment will identify the oldest equipment, which is energy-intensive and requires high power for cooling. It will also measure the actual utilization rate of IT equipment: most of the time, servers are only used at 10% or 20% of their capacity.

Replacing the oldest equipment, consolidating servers, and optimizing their capacity are essential steps to extend a data center's lifespan. The evolution of technologies, which allows for ever-increasing power in less space, also contributes to achieving this goal: thus, virtualization avoids systematically installing new physical servers when application or storage needs evolve.

Optimizing IT equipment helps contain, or even reduce, electrical consumption within a constant scope. However, data processing needs, storage, and strong availability and performance constraints are constantly growing. They lead to increased heat dissipation and the creation of hot spots. In just 10 years, the electrical and thermal density of data centers has thus increased from 500 to 1200, or even 2000W/m².

This progressive densification, which generates increased complexity in the operation of IT rooms, is generally poorly managed. The optimization of IT equipment must therefore necessarily be accompanied by a re-urbanization of the data center. For example, a better distribution of IT equipment in racks can reduce their number. The space thus gained can then be used to add new equipment or to optimize cooling. Another example: a new equipment layout, coupled with a reorganization of containments and a reorientation of airflows, leads to a reduction in cooling-related electrical consumption, without significant investment.

To go even further, the ideal is also, of course, to increase the available cooling capacity. Provided it is possible: everything generally depends on the data center's location. Thus, in urban areas, the constraints are such that the project often proves very complex: limited space, difficulty in obtaining ICPE (Classified Installations for Environmental Protection) authorizations, or the addition of air coolers that require acoustic protection from the neighborhood.

Conversely, data centers located away from cities, in industrial zones for example, do not experience the constraints associated with a dense environment. They thus offer more flexibility to increase cooling capacity, but also to reinforce electrical installations. In particular, it is easier to add generators to ensure business continuity in the event of a power outage.

In any case, whether in the city or on the outskirts, it is better to plan for reserve areas from the construction of a new data center. While the initial investment may be slightly higher, it will be compensated over time as it will allow subsequent phases to be carried out without interruption and within a controlled financial envelope.

The main difficulty of a data center lifespan extension project lies in the need to continue operations during the work phase. A voluntary or accidental shutdown, even for a short period, can have disastrous consequences for the company's activity (or hosted clients). It must therefore be prepared, which requires great rigor and very specific skills in the following areas: conducting an inventory, drafting the specifications step by step, defining intervention protocols and procedures (scheduled shutdowns, acceptable periods of degraded service), coordinating the various stakeholders (construction, electricians, IT teams), etc.

All these steps require expertise and an overall view that is often difficult to have internally. Support from specialists is recommended, especially since it also helps to 'guarantee' against insurance claims.

Furthermore, external intervention facilitates dialogue, understanding, and project management between the company's general services, responsible for the building, electrical, and cooling infrastructures, and the IT team, responsible for IT equipment and production.