Safe as in a safe
Data centers are focused on solving business problems by providing services in the form of information services. Modern data center, despite such relatively new requirements as a high degree of flexibility and scalability, still must, above all, ensure high reliability. All that you wanted to know about the reliability of data centers, but were afraid to ask - under the cut.
What is a data center (storage and) center? Modern data center - a comprehensive solution for managing enterprise information systems. This is a set of premises, external sites, engineering equipment, forming a single system for the placement of means of transmission, storage and processing of information with a given degree of accessibility . The data center includes a set of interrelated software and hardware components, organizational procedures, locations of these components and personnel for secure centralized processing, storage and provision of data, services and applications. From a system point of view, a data center is a fault-tolerant, integrated, centralized solution covering information and engineering systems. In the usual view, the data center is a specialized building where the server, network equipment and data storage systems are located, which are located in standard racks.
The equipment used in the data center is housed in special cabinets and racks.
')
The data center is a complex engineering structure for the centralized placement and maintenance of information systems, computer and telecommunications equipment, a high-tech platform equipped with air conditioning systems, uninterrupted power supply, fire extinguishing, SCS, and access control. The main purpose of data centers is to store, process and transfer customer-critical information, therefore the infrastructure of the data center includes many complex and redundant engineering systems of high reliability.
Reliability is the main characteristic of the data center. It includes many factors: reliability of the building, power supply, engineering systems and communications, installed equipment. High reliability of the data center means uninterrupted provision of services to customers in the case of a commercial data center and reliable operation of its own IT infrastructure of the company operating the corporate data center.
The data center provides access to enterprise information, management of its information systems, organization of data storage and processing. As a single multi-component system with consolidated centralized storage and data center data processing, first of all, it is designed to ensure uninterrupted automated operation of business processes.
The main tasks of the data center primarily include efficient consolidated data storage and processing, the provision of application services to users, and support for the operation of corporate applications. Data centers are created, first of all, to improve the efficiency of companies that actively use information technology in their activities and to improve the quality of services provided. They provide an opportunity to securely store important information throughout its entire life cycle.
Any data center consists of three technological units: informational (responsible for processing and storing information), telecommunications (ensuring the interconnection of center elements and data transmission) and engineering (this includes uninterrupted power supply, maintaining temperature and humidity in the room, fire extinguishing systems). Thus, data center elements include information, network, engineering infrastructure, as well as monitoring / control and security systems.
The main elements of the data center are data processing and storage systems, active network equipment, and engineering systems. It houses computing platforms and storage systems, data transmission system, power supply and electric lighting system, precision air conditioning system, structured cable system, cable channel system, raised floor, suspended ceiling, integrated security system, physical protection of computing systems and telecommunications equipment, early detection systems fire and gas extinguishing.
The information infrastructure includes a server complex, data storage and backup systems. The network and telecommunications infrastructure provides interaction between servers, combines logical layers and forms communication channels, including backbone connections with telecom operators and channels that provide users access to data center resources. Telecommunication infrastructure allows interconnection of data center elements, interaction between data center and users.
The structure of the data center.
The engineering infrastructure ensures the normal functioning of the main data center systems. It includes subsystems for climate control and air conditioning, power supply, fire alarm and fire suppression.
Network infrastructure (SCS) provides interaction between servers, and also includes the main channels for communication with operators and communication for users to communicate with the data center.
Automated systems for monitoring, dispatching, and managing information resources are a necessary element in increasing the reliability of the data center and reducing the negative impact of the “human factor”. The security system prevents unauthorized intrusion into areas of confidential information. It includes protections, alerts and access controls. Often, to prevent unauthorized entry into the data center (physical security), surveillance cameras are installed.
The main function of modern data centers is to increase the reliability of information processing and storage. One of the main criteria for assessing the performance of any data center is the availability of information systems. The data center should ensure business security and minimize downtime risks by meeting the following requirements:
These primary requirements are determined by the cost of downtime and the cost of ensuring the availability of applications and data. That is, business requirements — solution cost, recovery time of applications and data after a failure — determine the corresponding technical requirements for data center subsystems, including their redundancy, performance, throughput, capacity, and data integrity.
The reliability of a data center usually includes such components as a properly selected room, the availability of a well-functioning uninterrupted system of guaranteed power supply, a competent ventilation and air-conditioning system, water supply and drainage, as well as the proper construction of telecommunication systems. Ensuring the integrated security of the data processing center - access control, burglar and fire alarms, fire suppression - remains critical for the user. Aspects of disaster recovery and the availability of a backup site are becoming increasingly important.
In the design of modern data centers, in addition to exceptional reliability, fault tolerance and performance, systems are subject to a number of additional requirements, such as manageability, maintainability, low total cost of ownership, scalability, adaptability to change and security. Therefore, a data center is usually designed as an integrated solution, closely integrated with both the building management services and the enterprise's IT infrastructure. Modularity and standardization allows the whole complex to develop, adapt to changing conditions and loads.
The data center provides guaranteed uninterrupted power to the equipment being placed, even if the external power supply fails, to meet the requirements for climatic parameters (temperature, humidity) and security (physical and informational). As a rule, for high availability, data centers are connected to several (channels of various telecom operators or nodes of the corporate network. The quality and width of channels affect the level of services provided.
The main characteristic of the data center is its level of reliability. According to the TIA-942 standard, there are four levels of reliability:
Tier I (N) - without reservation. Availability of 99.71%. This is the most unreliable data center in which in the event of malfunctions and equipment failure or at the start of repair work, all data exchange centers are stopped. Moreover, in such a data center there are no raised floors and additional power sources, and the engineering infrastructure is not reserved.
Tier II (N + 1) - reservation of critical nodes. Availability of 99.741%. In this case, there is a certain level of redundancy, in such a data center there are raised floors and additional power sources, but in the event of the start of repairs, the data center will suspend its work.
Tier III (N + 1) - reservation of critical nodes, ways of obtaining electricity and routes of coolant delivery. At the same time, it is possible to decommission any node for maintenance, while maintaining the full functionality of the facility as a whole. Availability 99.982%. In such data centers there is the possibility of repair work without stopping the data center. Engineering systems are once reserved and there are several channels for power distribution and cooling, but only one of them is constantly running.
Tier IV (2 (N + 1)) is the most fault-tolerant level, where one accident is allowed (and not the planned decommissioning of a node) at one time. Availability is 99.995%, which means idle for only 26 minutes per year. In data centers of this type it is possible to conduct all types of work without stopping its work. Engineering systems are reserved twice, the main and additional systems are duplicated.
Four levels of fault tolerance data center.
According to experts, a fully redundant data center usually costs 2.5 times as much as a simple data center, so it is extremely important to determine at the level of pre-project preparation which category the site should fit. Both underestimation and overestimation of the importance of the parameter for permissible downtime equally affect the company's budget. After all, financial losses are possible either due to downtime and disruptions in critical systems, or due to irrational investments.
The main characteristic of the data center is fault tolerance, but no less important are the cost of operation, the indicator of energy consumption and energy efficiency (PUE). In determining the requirements for the data center, it is necessary to take into account the development trends of information systems and telecommunications, to provide options for expanding the capacity of the data center in accordance with the enterprise development plans.
The requirements for space and data center space are determined along with the overall performance and capacity of IT systems as well as such factors as calculated heat removal and the necessary level of redundancy for fault tolerance. An important parameter is the power consumption of the main equipment. For example, when designing a new data center, it is recommended to focus not on 5–6 kW per rack, but on connecting 10–15 kW. The power of uninterruptible power supplies is selected depending on the power consumption of the main equipment.
Driven to the data center should be approximately 250% of the power consumed by IT systems. Along with IT, it is consumed by air conditioning and power supply systems, as well as lighting, ventilation, fire suppression, access control, and video surveillance systems.
The dimensions and total weight of the racks determine the requirements for the minimum load capacity of floors and raised floors, and the composition and performance of information systems for cable systems and external cable entries for communication channels. Based on the capacity and types of equipment to be placed, as well as the required levels of service, requirements for auxiliary facilities are drawn up.
The requirements for the data center are also determined by the characteristics of the application systems and the need for the allocation of IT resources, legislative acts of state regulators of economic activity. In general, they depend on business needs and business problems to be solved, however, there are general requirements defined by standards.
Data centers can be divided into several types, which depend on some parameters. In terms of use, two types of data centers are usually distinguished: corporate and commercial (CSC). The purpose of creating a corporate data center is the automation of business processes, storage and processing of information of a company or organization. The main task of a commercial data center - making a profit from the rental of space, the provision of services and services.
According to its purpose in the information infrastructure of the data center can be primary or backup. The latter is necessary to replace the main in case of unforeseen accidents or preventive maintenance.
Data centers can also be divided into several types by size. Large data centers have their own building, which is created or selected in accordance with the requirements for the data center requirements. Average data centers, as a rule, rent a room of the required size, as well as channels with a certain bandwidth. Small data centers and server do not necessarily located in poorly adapted for this room, and the list of services presented at a minimum. In recent years, high-tech container data centers have become popular - specially designed and quickly deployable modular solutions. Finally, data centers are classified according to their reliability levels — the Tier I-IV mentioned above.
There are several types of commercial use of data centers. The first is virtual hosting, which implies the rental of a portion of disk space, processor time and RAM (one physical server is shared between multiple clients). The second - dedicated hosting - a similar scheme, but with guaranteed resources. The third is the placement of the server and other client equipment in the center (co-location).
There are two main documents that are most often mentioned when discussing data center standards: these are the TIA 942 standard and the level classification from the Uptime Institute. Both of these documents regulate the levels of reliability (Tier), which often leads to confusion: for example, the Tier III by TIA 942 and Tier III by the Uptime Institute are two big differences.
The ANSI / TIA / EIA-942 standard regulates the requirements for the location of data processing centers, external infrastructure, telecommunications space inside the data center, cable system and data center cable channels, and its infrastructure, depending on the required level of data center reliability. Data Centers I and II class can occupy part of a room, and objects of class III and IV are located only in separate buildings. In total, EIA / TIA-942 describes up to 200 different data center parameters.
The TIA 942 standard (Telecommunications Industry Association - Telecommunications Infrastructure Standard for Data Centers) was developed by the United States Telecommunications Industry Association, and primarily concerns the organization of structured cabling systems in the data center and fault tolerance and other engineering subsystems. It is a recommendation. The document developed by the Uptime Institute is not a standard, but rather a methodology for rationing the resiliency of the data center. It formulates the basic principles of design and approaches. Certified as a project (Tier Certification of Design Documents), and working platform (Tier Certification of Constructed Facility) and its operation (Operational Sustainability Certification). Certification is carried out only by the Uptime Institute.
Although in essence these levels are similar, the principles of assessment vary considerably. Uptime Institute verifies the result - how the data center works, rather than strict compliance with the documents. In Russia, there is no single standard, however, data centers focus on the requirements developed for communication facilities, on TIA-942, and also on GOSTs. There is also a relatively new standard BICSI 002 2010 Data Center Design and Implementation Best Practices, which appeared in 2010 and was updated in 2011. It was created to complement existing standards. European standard EN 50173-5 is also close to completion, on the basis of which in the future it is planned to form the corresponding Russian GOST.
Data processing centers are the core business of SAFEDATA , the operator of its own distributed data center network and IT outsourcing service provider. The SAFEDATA group, which today includes the Data Storage Center, the Center for Virtualization Technology, the CDN operator NGENIX and the operator of the traffic exchange platform MSC-IX, owns a network of data centers built in accordance with the requirements of the international standards TIA-942 (Tier III).
According to iKS-Consulting, the structure of Rostelecom in 2015 will occupy 12.5% ​​of the Russian data center market. Now the state-owned company has launched 3081 racks, including the MMTS-9 and SAFEDATA data centers. At the end of 2014, the income of its data center is estimated at 1.67 billion rubles.
SAFEDATA data centers - modern and energy-efficient engineering solutions. They are designed to function in virtually any environment. The data center SAFEDATA Moscow-I with a total area of ​​more than 2000 sq. M is located in Ostapovsky passage.
The data center Moscow-I began its work in September 2007. Since the start of the data center has not been a single emergency stop.
The infrastructure of the data center, designed for 500 racks, built in accordance with the standard TIA-942.
This 4 MW data center directly feeds from Mosenergo TPP-8. The data center provides uninterrupted power supply category 1, has independent power inlets, redundant diesel generator sets. For uninterrupted operation of the data center, four FG Wilson DGIs with a capacity of 1,250 kVA each, and four UPSs with a capacity of 800 kVA are involved. DGU automatically starts when the voltage from the transformer substation is lost and lasts up to 8 hours without refueling.
The data center has a VOLS to MMTS-9, MMTS-10, and customers are given the opportunity to enter their own FOCL.
The operation of the power supply system is monitored round-the-clock by SAFEDATA staff
In the Moscow-I data center there are installed precision cabinet air conditioners operating according to the (N + 1) scheme with a cooling capacity of 100 kW each. The precision air conditioning system is based on Hiref cabinet air conditioners and Geoclima refrigerators. From the air conditioners, the cooled air is fed under the raised floor and further, through the perforated raised floor tiles, into the “cold” corridors.
Guntner external condensing units are located on the roof of the building, connected to the air conditioners by two independent tracks.
On the open technological sites of the data center, two chillers with free cooling technology of 700 kW each have been installed, as well as two chillers with Turbocool technology. The total cooling capacity of the two circuits of the independent cooling system of the data center has been increased to 2,600 kW. A 40% ethylene glycol solution is used as a refrigerant. The data center is equipped with automatic gas fire extinguishing systems (AGPT) based on the gas mixture Freon 227.
Data Center Moscow-II , located at pl. Academician Kurchatov, 1, p.19, built in the framework of cooperation with the Russian Scientific Center "Kurchatov Institute".
Data Center Moscow-II with a total area of ​​3500 sq. M. built in accordance with the standard TIA-942 (Tier III) and put into operation in 2014.
It is designed for 900 racks, and the allocated electrical power is 10 MW.
Like Moscow-I, this data center provides its own fiber optic links to MMTS-9 and MMTS-10, and large customers are given the opportunity to enter their own fiber-optic links.
The data center has uninterrupted power supply category 1, independent power inputs duplicated by diesel-rotary uninterruptible power supplies. It is equipped with eight diesel rotor uninterruptible power supplies Hitec Power Protection with a capacity of 1,680 kVA each, reserved according to the N + 1 scheme.
Special backup capacity ensures the autonomous operation time of the entire diesel-rotary UPS system for at least 20 hours (prompt delivery of diesel fuel ensures uninterrupted operation of the data center, even with a long-term lack of power to the city power network).
Maintaining the required climatic parameters by using precision air conditioners Hiref. In the Moscow-II data center, several dozens of precision cabinet air conditioners operating under the (N + 1) scheme with a cooling capacity of 100 kW each are installed. The entire air conditioning system is reserved according to the N + 1 scheme for each engine room. From the air conditioners, the cooled air is supplied under the raised floor and further, through the perforated tiles of the raised floor to the “cold” corridors.
The data center is equipped with automatic gas fire extinguishing systems (AGPT) based on Freon 125 gas mixture. AGPT includes light and sound alarms, warning of gas start-up, which is integrated with the cooling system to turn off the air-conditioning equipment in case of an accident, and a system of physical access control.
The security system and access restrictions of the Moscow-I and Moscow-II data centers are organized on several levels and include a protected area, organizational procedures for controlling access to the data center, physical access control based on "proximity" cards, video surveillance system inside, outside the data center and in each module with thirty-day archiving of records. At the request of the customer can be installed additional security features.
The main feature of the modern data center is reliability. It does not stop even in case of accidents in the power systems of the city. The failed equipment and communication channels replace the backup systems. Emergency situations will not prevent SAFEDATA data centers - in case of any disaster there is an answer.
Datacenter: what is it?
What is a data center (storage and) center? Modern data center - a comprehensive solution for managing enterprise information systems. This is a set of premises, external sites, engineering equipment, forming a single system for the placement of means of transmission, storage and processing of information with a given degree of accessibility . The data center includes a set of interrelated software and hardware components, organizational procedures, locations of these components and personnel for secure centralized processing, storage and provision of data, services and applications. From a system point of view, a data center is a fault-tolerant, integrated, centralized solution covering information and engineering systems. In the usual view, the data center is a specialized building where the server, network equipment and data storage systems are located, which are located in standard racks.
The equipment used in the data center is housed in special cabinets and racks.
')
Data Center Assignment
The data center is a complex engineering structure for the centralized placement and maintenance of information systems, computer and telecommunications equipment, a high-tech platform equipped with air conditioning systems, uninterrupted power supply, fire extinguishing, SCS, and access control. The main purpose of data centers is to store, process and transfer customer-critical information, therefore the infrastructure of the data center includes many complex and redundant engineering systems of high reliability.
Reliability is the main characteristic of the data center. It includes many factors: reliability of the building, power supply, engineering systems and communications, installed equipment. High reliability of the data center means uninterrupted provision of services to customers in the case of a commercial data center and reliable operation of its own IT infrastructure of the company operating the corporate data center.
The data center provides access to enterprise information, management of its information systems, organization of data storage and processing. As a single multi-component system with consolidated centralized storage and data center data processing, first of all, it is designed to ensure uninterrupted automated operation of business processes.
The main tasks of the data center primarily include efficient consolidated data storage and processing, the provision of application services to users, and support for the operation of corporate applications. Data centers are created, first of all, to improve the efficiency of companies that actively use information technology in their activities and to improve the quality of services provided. They provide an opportunity to securely store important information throughout its entire life cycle.
Composition and structure of data center
Any data center consists of three technological units: informational (responsible for processing and storing information), telecommunications (ensuring the interconnection of center elements and data transmission) and engineering (this includes uninterrupted power supply, maintaining temperature and humidity in the room, fire extinguishing systems). Thus, data center elements include information, network, engineering infrastructure, as well as monitoring / control and security systems.
The main elements of the data center are data processing and storage systems, active network equipment, and engineering systems. It houses computing platforms and storage systems, data transmission system, power supply and electric lighting system, precision air conditioning system, structured cable system, cable channel system, raised floor, suspended ceiling, integrated security system, physical protection of computing systems and telecommunications equipment, early detection systems fire and gas extinguishing.
The information infrastructure includes a server complex, data storage and backup systems. The network and telecommunications infrastructure provides interaction between servers, combines logical layers and forms communication channels, including backbone connections with telecom operators and channels that provide users access to data center resources. Telecommunication infrastructure allows interconnection of data center elements, interaction between data center and users.
The structure of the data center.
The engineering infrastructure ensures the normal functioning of the main data center systems. It includes subsystems for climate control and air conditioning, power supply, fire alarm and fire suppression.
Network infrastructure (SCS) provides interaction between servers, and also includes the main channels for communication with operators and communication for users to communicate with the data center.
Automated systems for monitoring, dispatching, and managing information resources are a necessary element in increasing the reliability of the data center and reducing the negative impact of the “human factor”. The security system prevents unauthorized intrusion into areas of confidential information. It includes protections, alerts and access controls. Often, to prevent unauthorized entry into the data center (physical security), surveillance cameras are installed.
Reliability of the data center
The main function of modern data centers is to increase the reliability of information processing and storage. One of the main criteria for assessing the performance of any data center is the availability of information systems. The data center should ensure business security and minimize downtime risks by meeting the following requirements:
- Fault tolerance. This property of the technical system to maintain its performance after the failure of one or more composite components.
- High availability. This property of the system determines its reliability, the ability to perform the desired function under the specified conditions at a given time or within a specified time interval, subject to a certain set of conditions.
- Business Continuity. It includes processes and methods aimed at ensuring the non-stop execution of critical business functions.
- Disaster. This is the ability to recover from a disaster ie resistance to accidents and natural disasters.
These primary requirements are determined by the cost of downtime and the cost of ensuring the availability of applications and data. That is, business requirements — solution cost, recovery time of applications and data after a failure — determine the corresponding technical requirements for data center subsystems, including their redundancy, performance, throughput, capacity, and data integrity.
The reliability of a data center usually includes such components as a properly selected room, the availability of a well-functioning uninterrupted system of guaranteed power supply, a competent ventilation and air-conditioning system, water supply and drainage, as well as the proper construction of telecommunication systems. Ensuring the integrated security of the data processing center - access control, burglar and fire alarms, fire suppression - remains critical for the user. Aspects of disaster recovery and the availability of a backup site are becoming increasingly important.
In the design of modern data centers, in addition to exceptional reliability, fault tolerance and performance, systems are subject to a number of additional requirements, such as manageability, maintainability, low total cost of ownership, scalability, adaptability to change and security. Therefore, a data center is usually designed as an integrated solution, closely integrated with both the building management services and the enterprise's IT infrastructure. Modularity and standardization allows the whole complex to develop, adapt to changing conditions and loads.
The data center provides guaranteed uninterrupted power to the equipment being placed, even if the external power supply fails, to meet the requirements for climatic parameters (temperature, humidity) and security (physical and informational). As a rule, for high availability, data centers are connected to several (channels of various telecom operators or nodes of the corporate network. The quality and width of channels affect the level of services provided.
The main characteristic of the data center is its level of reliability. According to the TIA-942 standard, there are four levels of reliability:
Tier I (N) - without reservation. Availability of 99.71%. This is the most unreliable data center in which in the event of malfunctions and equipment failure or at the start of repair work, all data exchange centers are stopped. Moreover, in such a data center there are no raised floors and additional power sources, and the engineering infrastructure is not reserved.
Tier II (N + 1) - reservation of critical nodes. Availability of 99.741%. In this case, there is a certain level of redundancy, in such a data center there are raised floors and additional power sources, but in the event of the start of repairs, the data center will suspend its work.
Tier III (N + 1) - reservation of critical nodes, ways of obtaining electricity and routes of coolant delivery. At the same time, it is possible to decommission any node for maintenance, while maintaining the full functionality of the facility as a whole. Availability 99.982%. In such data centers there is the possibility of repair work without stopping the data center. Engineering systems are once reserved and there are several channels for power distribution and cooling, but only one of them is constantly running.
Tier IV (2 (N + 1)) is the most fault-tolerant level, where one accident is allowed (and not the planned decommissioning of a node) at one time. Availability is 99.995%, which means idle for only 26 minutes per year. In data centers of this type it is possible to conduct all types of work without stopping its work. Engineering systems are reserved twice, the main and additional systems are duplicated.
Four levels of fault tolerance data center.
According to experts, a fully redundant data center usually costs 2.5 times as much as a simple data center, so it is extremely important to determine at the level of pre-project preparation which category the site should fit. Both underestimation and overestimation of the importance of the parameter for permissible downtime equally affect the company's budget. After all, financial losses are possible either due to downtime and disruptions in critical systems, or due to irrational investments.
Other characteristics
The main characteristic of the data center is fault tolerance, but no less important are the cost of operation, the indicator of energy consumption and energy efficiency (PUE). In determining the requirements for the data center, it is necessary to take into account the development trends of information systems and telecommunications, to provide options for expanding the capacity of the data center in accordance with the enterprise development plans.
The requirements for space and data center space are determined along with the overall performance and capacity of IT systems as well as such factors as calculated heat removal and the necessary level of redundancy for fault tolerance. An important parameter is the power consumption of the main equipment. For example, when designing a new data center, it is recommended to focus not on 5–6 kW per rack, but on connecting 10–15 kW. The power of uninterruptible power supplies is selected depending on the power consumption of the main equipment.
Driven to the data center should be approximately 250% of the power consumed by IT systems. Along with IT, it is consumed by air conditioning and power supply systems, as well as lighting, ventilation, fire suppression, access control, and video surveillance systems.
The dimensions and total weight of the racks determine the requirements for the minimum load capacity of floors and raised floors, and the composition and performance of information systems for cable systems and external cable entries for communication channels. Based on the capacity and types of equipment to be placed, as well as the required levels of service, requirements for auxiliary facilities are drawn up.
The requirements for the data center are also determined by the characteristics of the application systems and the need for the allocation of IT resources, legislative acts of state regulators of economic activity. In general, they depend on business needs and business problems to be solved, however, there are general requirements defined by standards.
Data Center Classification
Data centers can be divided into several types, which depend on some parameters. In terms of use, two types of data centers are usually distinguished: corporate and commercial (CSC). The purpose of creating a corporate data center is the automation of business processes, storage and processing of information of a company or organization. The main task of a commercial data center - making a profit from the rental of space, the provision of services and services.
According to its purpose in the information infrastructure of the data center can be primary or backup. The latter is necessary to replace the main in case of unforeseen accidents or preventive maintenance.
Data centers can also be divided into several types by size. Large data centers have their own building, which is created or selected in accordance with the requirements for the data center requirements. Average data centers, as a rule, rent a room of the required size, as well as channels with a certain bandwidth. Small data centers and server do not necessarily located in poorly adapted for this room, and the list of services presented at a minimum. In recent years, high-tech container data centers have become popular - specially designed and quickly deployable modular solutions. Finally, data centers are classified according to their reliability levels — the Tier I-IV mentioned above.
There are several types of commercial use of data centers. The first is virtual hosting, which implies the rental of a portion of disk space, processor time and RAM (one physical server is shared between multiple clients). The second - dedicated hosting - a similar scheme, but with guaranteed resources. The third is the placement of the server and other client equipment in the center (co-location).
Data Center Standards
There are two main documents that are most often mentioned when discussing data center standards: these are the TIA 942 standard and the level classification from the Uptime Institute. Both of these documents regulate the levels of reliability (Tier), which often leads to confusion: for example, the Tier III by TIA 942 and Tier III by the Uptime Institute are two big differences.
The ANSI / TIA / EIA-942 standard regulates the requirements for the location of data processing centers, external infrastructure, telecommunications space inside the data center, cable system and data center cable channels, and its infrastructure, depending on the required level of data center reliability. Data Centers I and II class can occupy part of a room, and objects of class III and IV are located only in separate buildings. In total, EIA / TIA-942 describes up to 200 different data center parameters.
The TIA 942 standard (Telecommunications Industry Association - Telecommunications Infrastructure Standard for Data Centers) was developed by the United States Telecommunications Industry Association, and primarily concerns the organization of structured cabling systems in the data center and fault tolerance and other engineering subsystems. It is a recommendation. The document developed by the Uptime Institute is not a standard, but rather a methodology for rationing the resiliency of the data center. It formulates the basic principles of design and approaches. Certified as a project (Tier Certification of Design Documents), and working platform (Tier Certification of Constructed Facility) and its operation (Operational Sustainability Certification). Certification is carried out only by the Uptime Institute.
Although in essence these levels are similar, the principles of assessment vary considerably. Uptime Institute verifies the result - how the data center works, rather than strict compliance with the documents. In Russia, there is no single standard, however, data centers focus on the requirements developed for communication facilities, on TIA-942, and also on GOSTs. There is also a relatively new standard BICSI 002 2010 Data Center Design and Implementation Best Practices, which appeared in 2010 and was updated in 2011. It was created to complement existing standards. European standard EN 50173-5 is also close to completion, on the basis of which in the future it is planned to form the corresponding Russian GOST.
SAFEDATA data centers
Data processing centers are the core business of SAFEDATA , the operator of its own distributed data center network and IT outsourcing service provider. The SAFEDATA group, which today includes the Data Storage Center, the Center for Virtualization Technology, the CDN operator NGENIX and the operator of the traffic exchange platform MSC-IX, owns a network of data centers built in accordance with the requirements of the international standards TIA-942 (Tier III).
According to iKS-Consulting, the structure of Rostelecom in 2015 will occupy 12.5% ​​of the Russian data center market. Now the state-owned company has launched 3081 racks, including the MMTS-9 and SAFEDATA data centers. At the end of 2014, the income of its data center is estimated at 1.67 billion rubles.
SAFEDATA data centers - modern and energy-efficient engineering solutions. They are designed to function in virtually any environment. The data center SAFEDATA Moscow-I with a total area of ​​more than 2000 sq. M is located in Ostapovsky passage.
The data center Moscow-I began its work in September 2007. Since the start of the data center has not been a single emergency stop.
The infrastructure of the data center, designed for 500 racks, built in accordance with the standard TIA-942.
This 4 MW data center directly feeds from Mosenergo TPP-8. The data center provides uninterrupted power supply category 1, has independent power inlets, redundant diesel generator sets. For uninterrupted operation of the data center, four FG Wilson DGIs with a capacity of 1,250 kVA each, and four UPSs with a capacity of 800 kVA are involved. DGU automatically starts when the voltage from the transformer substation is lost and lasts up to 8 hours without refueling.
The data center has a VOLS to MMTS-9, MMTS-10, and customers are given the opportunity to enter their own FOCL.
The operation of the power supply system is monitored round-the-clock by SAFEDATA staff
In the Moscow-I data center there are installed precision cabinet air conditioners operating according to the (N + 1) scheme with a cooling capacity of 100 kW each. The precision air conditioning system is based on Hiref cabinet air conditioners and Geoclima refrigerators. From the air conditioners, the cooled air is fed under the raised floor and further, through the perforated raised floor tiles, into the “cold” corridors.
Guntner external condensing units are located on the roof of the building, connected to the air conditioners by two independent tracks.
On the open technological sites of the data center, two chillers with free cooling technology of 700 kW each have been installed, as well as two chillers with Turbocool technology. The total cooling capacity of the two circuits of the independent cooling system of the data center has been increased to 2,600 kW. A 40% ethylene glycol solution is used as a refrigerant. The data center is equipped with automatic gas fire extinguishing systems (AGPT) based on the gas mixture Freon 227.
Data Center Moscow-II , located at pl. Academician Kurchatov, 1, p.19, built in the framework of cooperation with the Russian Scientific Center "Kurchatov Institute".
Data Center Moscow-II with a total area of ​​3500 sq. M. built in accordance with the standard TIA-942 (Tier III) and put into operation in 2014.
It is designed for 900 racks, and the allocated electrical power is 10 MW.
Like Moscow-I, this data center provides its own fiber optic links to MMTS-9 and MMTS-10, and large customers are given the opportunity to enter their own fiber-optic links.
The data center has uninterrupted power supply category 1, independent power inputs duplicated by diesel-rotary uninterruptible power supplies. It is equipped with eight diesel rotor uninterruptible power supplies Hitec Power Protection with a capacity of 1,680 kVA each, reserved according to the N + 1 scheme.
Special backup capacity ensures the autonomous operation time of the entire diesel-rotary UPS system for at least 20 hours (prompt delivery of diesel fuel ensures uninterrupted operation of the data center, even with a long-term lack of power to the city power network).
Maintaining the required climatic parameters by using precision air conditioners Hiref. In the Moscow-II data center, several dozens of precision cabinet air conditioners operating under the (N + 1) scheme with a cooling capacity of 100 kW each are installed. The entire air conditioning system is reserved according to the N + 1 scheme for each engine room. From the air conditioners, the cooled air is supplied under the raised floor and further, through the perforated tiles of the raised floor to the “cold” corridors.
The data center is equipped with automatic gas fire extinguishing systems (AGPT) based on Freon 125 gas mixture. AGPT includes light and sound alarms, warning of gas start-up, which is integrated with the cooling system to turn off the air-conditioning equipment in case of an accident, and a system of physical access control.
The security system and access restrictions of the Moscow-I and Moscow-II data centers are organized on several levels and include a protected area, organizational procedures for controlling access to the data center, physical access control based on "proximity" cards, video surveillance system inside, outside the data center and in each module with thirty-day archiving of records. At the request of the customer can be installed additional security features.
The main feature of the modern data center is reliability. It does not stop even in case of accidents in the power systems of the city. The failed equipment and communication channels replace the backup systems. Emergency situations will not prevent SAFEDATA data centers - in case of any disaster there is an answer.
Source: https://habr.com/ru/post/265311/
All Articles