DPC Engineering Infrastructure
Requirements of operation reliability and continuity specified for the state-of-the-art computing centers become higher from year to year. To ensure required reliability level high reliable server complexes and active DPC equipment are insufficient. Today the requirements as rigid are imposed to the DPC premise, stable equipment operation conditions and physical safety of the equipment and data.
Today DPC is a complex engineering objects equipped with independent operation support systems. The DPC engineering infrastructure is a complex of engineering structures aimed to ensure constant power supply, microclimate parameter stability, DPC equipment physical protection.
Today the DPC consists of the following premises:
- server premises;
- switch rooms;
- auxiliary area for engineering systems (accumulator premises, AGFFS bulb premise);
- system administrators on duty room.
Open Technologies complex projects on design, construction and updating of DPC engineering infrastructure often include issues of power supply updating and organization of the building complex security system.
Upon creation of DPCs it is necessary to pay special attention not only to server complexes and active equipment but also to construction of high quality engineering infrastructure. Engineering system reliability directly affects the level of DPC service availability and shall be higher than reliability of other components. The design shall consider standby schemes ensuring the same false-safety level as the standby schemes used in DPC technical architecture design.
The DPC without automated engineering infrastructure control system cannot ensure the service reliability level required for the business today. Presence of the unified automated engineering infrastructure monitoring and control system in DPCs is recommended for any level of its operation reliability and compulsory to achieve reliability of over 99.98% (according to The Uptime Institute research and ANSI-TIA-942/2005 standard).
Qualified consultants and engineers of Open Technologies execute the whole cycle of works on designing and introduction of DPC engineering infrastructure taking into account the organization scale and infrastructure, objectives and customer's tasks.
- General uninterrupted continuous power supply:
Uninterrupted Power Supply System (UPSS). Power supply in case of the main power supply outages for the time required to recover external power supply or switch it to standby (emergency) uninterruptible power supply sources. Uninterrupted Power Supply units, DC systems and accumulators are used.
Continuous Power Supply System (CPSS). Provision of DPC computing and engineering complex with power supply during long power outages. Duplicated external supply terminal, standby auto cut-in devices and diesel generator units are used.
General power supply system. Power supply to consumers of B, C groups and uninterrupted power supply system. The main components are: group networks, power outlets.
- Precision conditioning and supply ventilation
Climate maintenance in the DPC premises according to preset temperature and humidity parameters to ensure reliable equipment operation. Conditioning, ventilation and air heating. Ensuring space air overpressure to create overpressure in DPC premises equipped with gas fire fighting system.
- Installation components
Aimed for deployment and fixation of equipment in optimum way regarding density, availability and optimum environment. Equipment used - mounting racks, floor cupboards, wall frames, wall cupboards.
- False floor. Ceiling and floor baseboards
Aimed to lay cable channels, insert system, air ducts of suction-and-exhaust ventilation in DPCs and buildings. Ensures mechanic protection for cables, preservation of building design and easy access for maintenance and updating of cable system.
- Complex Physical Security System:
Access Monitoring and Control System (AMCS). The system is aimed to control transit routes in the DPC premises. Ensures selective permission of people to the DPC adjacent territory and registration of entrance.
CCTV system. The system ensures 24x7 surveillance in DPC premises and registers movements at the territory, informs operator and makes video records.
Intrusion alarm. The system is aimed to control state of perimeter/ certain access zones with movement sensors, opening sensors, stress sensors, surface break sensors. Aimed to warn security about intrusions to DPC territory.
- Fire alarm. The system is aimed to ensure smoke or fire detection in all premises under control. This system is able to initiate fire alarm processes, personnel evacuation, start of automatic fire extinguishing systems, stop air supply, disconnect power supply and start smoke exhaust system.
- Emergency Power Shut-Down System The system ensures power shut-down in case of water fire extinguishing to prevent short circuits in electric lines.
- Gas Fire Fighting and Smoke Exhaust System. Ensures automatic fire fighting in DPC premises supplying gases repressing combustion (freon, carbon dioxide, inergen) and stops air supply to premise on fire. After activation of the gas fire fighting system the smoke exhaust system removes combustion and extinguishing gas products from the premise.
- Emergency lighting. Ensures lighting of the DPC premise evacuation routes in case of emergency. Includes individual power supply sources used for lighting and as separate UPSs.
- Evacuation system. The system is aimed to produce audible messages and inform employees at the territory of DPC. Also includes a complex of organizational measures aimed to control evacuation of personnel.
- Structural Cable System (SCS). Used to construct communication infrastructure for technological equipment control systems, internal phone network, intrusion alarm, video communication systems, surveillance and industry television. The SCS is aimed to transmit all main low current signals via cable used by enterprises irrespective of applications and equipment.
- Monitoring and control. DPC engineering infrastructure consists of tens of devices many of which have built-in microprocessor control systems. Number of parameters to be controlled by today's DPC can amount to thousands (climate parameters of separate racks, server consumption currents, security sensor state, etc.). All parameters shall be collected and analyzed in 24x7 real-time mode. This can be ensured only by using unified control and monitoring system which regularly collects information from all data sources (sensors, equipment state) and controls the equipment to ensure the most comfortable operation conditions.
Difficulty of computing system engineering infrastructure construction includes organization of unified designing, installation and test procedures for all infrastructure subsystems. At the same time the quality of works defines critical parameters such as information service reliability and availability.
Considering the fact that the today's DPC can consist of tens, sometimes hundreds of servers, large number of telecommunication equipment and other components, we use total time of unavailability of services provided to users by the DPC or provision of service not in full (not all users or insufficient response time) to define the DPC reliability.
The total annual DPC dead time in hours divided by 8760 (24 hours х 365 days) defines possibility of DPC unavailability. Accordingly, 100% possibility of unavailability defines DPC availability level. The state-of-the-art DPC availability level usually fluctuates between 99% (87.6 dead hours per year) and 99.95% (0.4 dead hours per year).
The benefits of state-of-the-art DPC engineering infrastructure based on СН 512-78 and ANSI-TIA 942 standards can be divided into direct and indirect ones.
The direct benefits include reliable DPC operation in case of ideal constant power supply to consumers, optimum environmental parameters and constant control of physical access to equipment. This operation mode is the primary objective of design and construction of the DPC engineering infrastructure.
Indirect benefits are usually deeper and become clear only during active DPC operation:
- reduction of DPC update and extension time - correctly designed DPC engineering infrastructure allows reduction of updating time from 4-6 months to 2-6 weeks due to flexible use of redundant equipment resources;
- reduction of the number of used vendors and facilitation of equipment service maintenance - unified DPC engineering infrastructure designing and use of industry equipment significantly reduces expenses of conjugation of engineering infrastructure subsystems and equipment maintenance.
Separately, the benefits of the unified engineering infrastructure monitoring and control:
- almost immediate response to event - data from the most critical points can be received several times per second thus ensuring fast system response, minimize possible negative consequences of emergency situations;
- no human presence in decision-making in case of emergency - the majority of threats are known and the action procedure is known; the unification of engineering systems into unified logical network including decision-making mechanisms integrated into automated control systems ensures preset response of the whole DPC engineering infrastructure to emergency situation;
- lower human factor in engineering infrastructure operation;
- automated system allows setting necessary parameters of the engineering system operation, control compliance to these parameters, continuously control parameters of environment, power, technological equipment state with independent control devices. If the parameters are out of limits or failures are detected the personnel on duty receives warnings and recommendations on elimination of emergency situation;
- extended possibilities of engineering systems by extension of the control points thus allowing to guarantee compliance to necessary environmental and equipment operation parameters (for example, regulation of microclimate maintenance system control not only depending on hot air temperature supplied to conditioner but depending on data received from additional sensors installed directly in the equipment air intakes);
- unified monitoring and control interface - the personnel on duty has unified graphic monitoring and control interface for all DPC engineering systems thus facilitating information perception, allowing real-time control of all parameters, timely detection of changes in operation and execution of maintenance and repairs;
- possibility to forecast engineering infrastructure element failures - connection of additional sensors and measurement of nonstandard parameters of engineering infrastructure allows prevention of possible emergencies (for example, certain value sagging sensors tell that pressure on cover has reached maximum permissible value and further load increase will lead to breaks of constructions).