Data center: first and foremost a production site

Most companies devote considerable resources to building new data centers or refurbishing real estate assets in order to house their IT rooms. Yet, on a day-to-day basis, it is the proper operation of the building and the equipment hosted there that ensures the expected level of service for users. Thus, over the entire life cycle of data centers (between 10 and 20 years), their operation costs more than their design.

However, the “operations” component is still far too often the poor relation in design specifications, because in this area cost control still largely takes precedence over risk management.

In practical terms, data center operations have two aspects: on the one hand, managing the regular changes to physical and logical IT infrastructures, carried out by IT teams. On the other hand, managing the building and its technical utilities infrastructure (power supply, cooling, etc.), handled by general services or real estate departments.

In the industrial sector, this dual management has been organized for many years: the day-to-day operation of factories is governed by documented procedures for installation, maintenance, and incident resolution. In data centers, this is still quite rare. Yet industrial operations for IT rooms deliver the same benefits: better overall steering of investments, harmonization of working methods and operational processes (maintenance, day-to-day management, security, protocols in the event of malfunction, etc.), and improved service quality.

Industrializing data center operating processes requires an investment of time as well as money. It requires genuine coordination between the teams responsible for IT and those responsible for managing the company’s real estate assets. Overall, around one year is generally needed to produce a comprehensive operations manual, from inventorying installations and defining procedures through to finalization, including testing the various scenarios.

This investment should be viewed in light of the benefits for the company: implementing standardized, reliable procedures helps reduce the number of incidents, particularly those caused by improper handling. Documented protocols also make it possible to establish KPIs (Key Performance Indicators) to track service-level objectives, compliance with interventions, capacity evolution, correction times following an incident, etc.

These data can be consolidated into dashboards for better overall steering of all of a company’s or group’s data centers. They also facilitate managing the relationship with the provider responsible for multi-technical maintenance (PMT) when operations are outsourced: procedures are based on precise protocols defined upstream of the contract, and service-level commitments can be monitored on the basis of objective KPIs.