When creating data centers, management issues are often given secondary attention. Nevertheless, analysts note a trend characteristic today for this market where the number of data center automation projects is growing.

The incentives for the introduction of such systems are changes in the infrastructure of data centers associated with virtualization, high density of equipment, the transition to cloud computing and a number of other factors. The introduction of automated data center management tools has specific economic and technological motives. Quality management tools not only increase the availability of data center infrastructure but also allow you to achieve significant economic benefits by reducing energy costs and reducing the time to deploy services that experts can spend on more important tasks.

The cost of the data center includes not only the investment in the acquisition of servers but also the cost of their administration. Analysts predict that over the next two to three years, the cost of administering the acquired servers will be seven times higher than the funds actually purchased. Therefore, the administration segment is the highest priority in terms of optimizing the cost of data centers.

Control mode
According to the experts from Managed Network, if we use the maturity model to assess the state of the data center, then the best condition is the service-oriented data center, the worst is the center on manual control. In the second case, there is no documentation at the facility. As a rule, in such a data center, IT equipment is replaced whenever possible, and often it works as long as it can work at all. The downtime of such a center can also be absolutely anything and, in fact, does not affect the company’s operations.

If the business is not indifferent to the duration of downtime, apply a more advanced, reactive system. In the event of an accident, such a data center quickly finds a replacement for the equipment and corrects the situation. As a rule, some documentation is kept for such sites, the troubleshooting procedure is regulated. However, in general, the operation of such a data center rests on the experience of IT professionals who work there.

“Often when creating a data processing center, we are faced with the fact that the customer does not trust the automated control system (ADCS) and is trying to bring a manual mode into it,” and in order to minimize the risks of introducing the ADCS, first of all, it is necessary to develop regulations for its operation, and secondly, to launch and test operation of the entire engineering complex with a duration of at least one month.”- stated by most of the IT Outsourcing companies.

The data center management model, which experts in the field of operating data centers today are aiming at, is proactive, suggesting work in real time. This is a model that allows you to anticipate problems, predict the impact of changes and not only find the causes of errors reported by troubleshooting, but also establish workarounds for the restoration of services. This is a single system that can, for example, predict how a certain area of a data center will react to a power supply or UPS shutdown.

In such a service-oriented model, the facility is fully documented, the procedures for the replacement of equipment are clearly defined, and various types and forms of reports are created for the operational services. Initially, the system has a number of fixed report formats for the parameters of engineering systems, for accidents and alarms, time of occurrence, time of acknowledgment and actions of personnel to eliminate accidents. The built-in designer allows you to quickly set up the required type of reports and analyze emergency situations for any period you like. Thus, the management system provides a basis for litigation - for example, between the owner of the data center and the electricity supplier.

Colorful landscape
One of the most common approaches to managing a data center is to separate control loops — by application systems, IT Outsourcing companies, and engineering infrastructure. These circuits are even different teams of operators. “The owners of data centers invest in infrastructure, organize mandatory monitoring of servers, but do not pay enough attention to engineering infrastructure.

At the moment, very few companies in the UK, especially in Managed Network Services, are engaged in the optimization and management of the data center. Many invest in infrastructure, buying complex solutions - engineering, server, but rarely anyone creates a unified management platform and forms a unified management team.