Datacenter: the stages of big life
Today we decided to pay attention to the topic of the “life cycle” of data centers, to talk about what it is and how to properly manage it allows companies to reduce not only current, but also capital costs. Given the current trends in the introduction of all new solutions (including hyperconvergent systems) and services, significant difficulties in this process may be associated with the change and adaptation of the physical infrastructure of data centers.
The data center life cycle includes a phased development of engineering infrastructure in accordance with the objectives of the business and IT processes and can be divided into several stages:
1. Preparatory : understanding the objectives of the creation, the formation of the concept of solutions, the choice of the site;
2. Design in accordance with the selected concept and characteristics of the site;
3. Construction ;
4. Operation ;
5. Conformity assessment and performance analysis.
')
Business requirements for IT services are constantly changing, and completing this cycle, conformity assessment just helps to understand how effectively the data center can solve current problems, and whether it can cope with new ones, is it ready for growth. This is followed by planning to refine or modernize the site for new tasks and all subsequent stages - the circle closes.
The Russian market has accumulated substantial competence in the field of design and construction of data centers, there are specialists of the appropriate qualifications, but it is not always possible to complete the project in accordance with the KPIs set by the client. Already in the course of the project, business requirements may change and are further incorrectly formulated and reflected in the terms of reference. As a result, the construction of the site is of high quality, but the main difficulty at the design stage is to understand what tasks the facility faces and how exactly the data center should be designed and equipped to meet all business requirements, which is not always possible.
Today it is required to clearly formulate the terms of reference, describing in it correct and optimal solutions that meet business objectives in the long term for at least 5-8 years. The training center of a vendor with appropriate competence can assist in the training of competent specialists. So the task of the Schneider Electri c Training Center is to familiarize partners and customers with the technically complex products of the company, the peculiarities of its use, installation and operation in real conditions.
A data center, as a rule, is not physically tied to consumers of its services, however, it is important to bear in mind that placing it in an inappropriate building dramatically increases the cost of the site infrastructure. It is worth remembering that the environment of the data center can affect its reliability.
Typical mistakes made during the preparatory stage:
1. Designing a data center as an office, while it is an industrial facility;
2. Errors in the reservation of systems, components, communications;
3. Errors in the calculation of the size of the allocated area for engineering systems and auxiliary facilities, with an assessment of the weight of the equipment and its dimensions, with ensuring the autonomy of the object.
It is important to understand the key metrics of the data center, the prospects for its development, as the engineering systems must “survive” 2-3 generations of IT equipment. Operational services are rarely involved in a project during the construction phase. Meanwhile, experts in power supply and cooling systems should be involved at the design stage and be present in the working groups.
No matter how well a project is thought out and correctly implemented, many problems arise at the operation stage. One of them is of an organizational nature and concerns the interaction between the administration and the IT department. When there is no such bundle, the solution as a whole may not work effectively or not function at all.
In practice, administrative departments sometimes turn out to be very far from understanding what a data center is and what its requirements are, and sometimes they simply become an insurmountable bureaucratic barrier, so it is advisable to show this service the concept of building a data center at their site, explain the principles of operation engineering systems and approximate equipment parameters.
The IT department may have its own internal problems. When different IT departments are responsible for different components (servers, storage systems, network equipment, cooling systems, etc.) and at the same time do not have clear regulations and division of responsibility areas, chaos occurs at the site.
In this case, it is easier to “move” to a new site than to carry out a deep modernization of the old one, and only then redo the existing data center without disrupting the work of the IT services operating at the new location. Another option - at the time of "overhaul" to rent space in a commercial data center. The optimal solution is chosen in each case.
In modern data centers for the continuous and efficient operation of the infrastructure, as a rule, a separate team of specialists is responsible. To automate monitoring, management, reporting, motion control, and performance monitoring, they use tools that are commonly referred to by the acronym DCIM (Data Center Infrastructure Management).
The tasks of monitoring events, the state of the equipment and the environment are often solved in the first place — for this purpose, tools such as NetBotz controllers and the Data Center Expert monitoring system are used. But now the most relevant tasks are the effective use of resources - from electricity to space in the machine rooms. The timely implementation of efficient regulations and control over the use of electricity allows not only reducing very significant operating costs and returning investments earlier, but also planning new acquisitions or construction, taking into account the assessment of real, not “paper”, equipment efficiency. As a platform for organizing this approach, you can use Data Center Operation - a modular solution with a set of all the necessary functions.
This Schneider Electric solution allows data center operators to control and more efficiently manage data centers and perform current tasks, while the integration of third-party IT services allows you to exchange relevant information between different vendor systems, resulting in a more complete picture of the performance and availability of the data center to increase its reliability and operational efficiency.
Optimization - improving the parameters of the data center subsystems - requires a preliminary assessment stage - auditing, analyzing the effectiveness of their work, creating a mathematical model of the site and eliminating problem points based on monitoring data.
A high-quality monitoring system, created specifically for solving data center operation tasks, equipment operation modeling system, which allows to obtain a clear model of the data center operation, are important tools through which operation engineers can track the state of the data center in real time.
Various improvements, such as building systems for isolating air corridors, reorganizing the raised floor plates, eliminating places with incorrect circulation of air streams in the racks, and then reconfiguring air conditioning systems, reduce the energy consumption of the data center in real projects by 10-20%. Such optimization (if only energy consumption is taken into account) for 3-3.5 years pays off.
Another optimization effect is improving the performance of cooling systems when previously installed equipment is used more efficiently , for example, it becomes possible to install more servers in a rack. "The efficiency of the data center" can be increased from 70-75% to 90-95%. Such optimization pays off in about half a year.
Data analysis helps to find a specific solution to reduce the cost of operating devices at the customer’s site, select the most energy-efficient mode, or optimize resource utilization. Or understand that it's time to upgrade.
The modernization of the data center does not always mean the installation of new equipment and in the first place represents an increase in the efficiency of the systems and the efficiency of their use.
Assessing the situation helps to understand what needs to be done in order for the data center to work effectively again. One of the customers solved this problem by creating closed cold corridors and blocking air flows where they reduced the efficiency of heat removal. And the work was done without stopping the data center systems.
In many projects, work is carried out at the junction of the stages of operation and evaluation. Today, many customers have new IT tasks, information is being introduced, business is being digitalized, and various operational problems have to be solved. The help of experts at the assessment stage and the audit of the data center help to cope with them. The Regional Application Center (Regional Application Center) created by Schneider Electric in Russia is engaged in integrated solutions for data centers. Its main task is to create and support a sustainable and efficient functioning of the data center at all stages of its life cycle.
Schneider Electric experts are involved in the design, construction, integrated testing and commissioning of facilities, in the subsequent support and modernization of the data center. At each stage, the customer is provided with fault tolerance and data center efficiency. With the growth of the scale and speed of digital technologies, the solutions we offer will help companies, providers of cloud, telecommunication services and commercial sites to look to the future with confidence.
The data center life cycle includes a phased development of engineering infrastructure in accordance with the objectives of the business and IT processes and can be divided into several stages:
1. Preparatory : understanding the objectives of the creation, the formation of the concept of solutions, the choice of the site;
2. Design in accordance with the selected concept and characteristics of the site;
3. Construction ;
4. Operation ;
5. Conformity assessment and performance analysis.
')
Business requirements for IT services are constantly changing, and completing this cycle, conformity assessment just helps to understand how effectively the data center can solve current problems, and whether it can cope with new ones, is it ready for growth. This is followed by planning to refine or modernize the site for new tasks and all subsequent stages - the circle closes.
Design and construction
The Russian market has accumulated substantial competence in the field of design and construction of data centers, there are specialists of the appropriate qualifications, but it is not always possible to complete the project in accordance with the KPIs set by the client. Already in the course of the project, business requirements may change and are further incorrectly formulated and reflected in the terms of reference. As a result, the construction of the site is of high quality, but the main difficulty at the design stage is to understand what tasks the facility faces and how exactly the data center should be designed and equipped to meet all business requirements, which is not always possible.
Today it is required to clearly formulate the terms of reference, describing in it correct and optimal solutions that meet business objectives in the long term for at least 5-8 years. The training center of a vendor with appropriate competence can assist in the training of competent specialists. So the task of the Schneider Electri c Training Center is to familiarize partners and customers with the technically complex products of the company, the peculiarities of its use, installation and operation in real conditions.
A data center, as a rule, is not physically tied to consumers of its services, however, it is important to bear in mind that placing it in an inappropriate building dramatically increases the cost of the site infrastructure. It is worth remembering that the environment of the data center can affect its reliability.
Typical mistakes made during the preparatory stage:
1. Designing a data center as an office, while it is an industrial facility;
2. Errors in the reservation of systems, components, communications;
3. Errors in the calculation of the size of the allocated area for engineering systems and auxiliary facilities, with an assessment of the weight of the equipment and its dimensions, with ensuring the autonomy of the object.
It is important to understand the key metrics of the data center, the prospects for its development, as the engineering systems must “survive” 2-3 generations of IT equipment. Operational services are rarely involved in a project during the construction phase. Meanwhile, experts in power supply and cooling systems should be involved at the design stage and be present in the working groups.
Exploitation
No matter how well a project is thought out and correctly implemented, many problems arise at the operation stage. One of them is of an organizational nature and concerns the interaction between the administration and the IT department. When there is no such bundle, the solution as a whole may not work effectively or not function at all.
In practice, administrative departments sometimes turn out to be very far from understanding what a data center is and what its requirements are, and sometimes they simply become an insurmountable bureaucratic barrier, so it is advisable to show this service the concept of building a data center at their site, explain the principles of operation engineering systems and approximate equipment parameters.
The IT department may have its own internal problems. When different IT departments are responsible for different components (servers, storage systems, network equipment, cooling systems, etc.) and at the same time do not have clear regulations and division of responsibility areas, chaos occurs at the site.
In this case, it is easier to “move” to a new site than to carry out a deep modernization of the old one, and only then redo the existing data center without disrupting the work of the IT services operating at the new location. Another option - at the time of "overhaul" to rent space in a commercial data center. The optimal solution is chosen in each case.
In modern data centers for the continuous and efficient operation of the infrastructure, as a rule, a separate team of specialists is responsible. To automate monitoring, management, reporting, motion control, and performance monitoring, they use tools that are commonly referred to by the acronym DCIM (Data Center Infrastructure Management).
The tasks of monitoring events, the state of the equipment and the environment are often solved in the first place — for this purpose, tools such as NetBotz controllers and the Data Center Expert monitoring system are used. But now the most relevant tasks are the effective use of resources - from electricity to space in the machine rooms. The timely implementation of efficient regulations and control over the use of electricity allows not only reducing very significant operating costs and returning investments earlier, but also planning new acquisitions or construction, taking into account the assessment of real, not “paper”, equipment efficiency. As a platform for organizing this approach, you can use Data Center Operation - a modular solution with a set of all the necessary functions.
This Schneider Electric solution allows data center operators to control and more efficiently manage data centers and perform current tasks, while the integration of third-party IT services allows you to exchange relevant information between different vendor systems, resulting in a more complete picture of the performance and availability of the data center to increase its reliability and operational efficiency.
Evaluation and optimization
Optimization - improving the parameters of the data center subsystems - requires a preliminary assessment stage - auditing, analyzing the effectiveness of their work, creating a mathematical model of the site and eliminating problem points based on monitoring data.
A high-quality monitoring system, created specifically for solving data center operation tasks, equipment operation modeling system, which allows to obtain a clear model of the data center operation, are important tools through which operation engineers can track the state of the data center in real time.
Various improvements, such as building systems for isolating air corridors, reorganizing the raised floor plates, eliminating places with incorrect circulation of air streams in the racks, and then reconfiguring air conditioning systems, reduce the energy consumption of the data center in real projects by 10-20%. Such optimization (if only energy consumption is taken into account) for 3-3.5 years pays off.
Another optimization effect is improving the performance of cooling systems when previously installed equipment is used more efficiently , for example, it becomes possible to install more servers in a rack. "The efficiency of the data center" can be increased from 70-75% to 90-95%. Such optimization pays off in about half a year.
Data analysis helps to find a specific solution to reduce the cost of operating devices at the customer’s site, select the most energy-efficient mode, or optimize resource utilization. Or understand that it's time to upgrade.
The modernization of the data center does not always mean the installation of new equipment and in the first place represents an increase in the efficiency of the systems and the efficiency of their use.
Assessing the situation helps to understand what needs to be done in order for the data center to work effectively again. One of the customers solved this problem by creating closed cold corridors and blocking air flows where they reduced the efficiency of heat removal. And the work was done without stopping the data center systems.
In many projects, work is carried out at the junction of the stages of operation and evaluation. Today, many customers have new IT tasks, information is being introduced, business is being digitalized, and various operational problems have to be solved. The help of experts at the assessment stage and the audit of the data center help to cope with them. The Regional Application Center (Regional Application Center) created by Schneider Electric in Russia is engaged in integrated solutions for data centers. Its main task is to create and support a sustainable and efficient functioning of the data center at all stages of its life cycle.
Schneider Electric experts are involved in the design, construction, integrated testing and commissioning of facilities, in the subsequent support and modernization of the data center. At each stage, the customer is provided with fault tolerance and data center efficiency. With the growth of the scale and speed of digital technologies, the solutions we offer will help companies, providers of cloud, telecommunication services and commercial sites to look to the future with confidence.
Source: https://habr.com/ru/post/335654/
All Articles