As we all know, a comprehensive data center room construction project generally includes: comprehensive wiring, antistatic floor laying, roof wall decoration, partition decoration, UPS, special constant temperature and humidity air conditioning, room environment monitoring system, fresh air system, water leakage detection, ground wire system, lightning protection system, access control, monitoring, fire, alarm, shielding engineering.
Do you know what are the key problems in the design and construction of data center room? Today, let's learn something together.
1. Where will the data center be built
It is important to analyze the climatic conditions of the location of the new data center. This will help determine the most appropriate cooling measures, including water cooling units, mechanical refrigeration air conditioning, natural air cooling, indirect air cooling, adiabatic cooling, etc. for data center facilities to achieve their lower PUE goals.
2. How many racks are needed and what is the size
The number of racks determines the space requirements of the data center. Traditional IT racks are 600x1000mm (width x depth), while a room of 100 square meters can accommodate about 50 racks. Racks come in other sizes, of course, and knowing the rack size and number of racks makes it easy to estimate how much space your data center needs.
3. What level of data center is needed
The Tier level of a data center determines the redundancy requirements and power distribution path of the data center. For a Tier2 data center, the redundancy of power and cooling equipment deployed in the "N+1" architecture is sufficient. However, for a Tier3 data center facility, it is common to include "N+1" cooling equipment redundancy, with dual supply 2N power redundancy, and separate dual distribution racks. Therefore, knowing the level of the data center can determine the required redundancy design.
4. What is the average power density of each frame
It is important to note that when designing the power capacity of the data center, it should be based on average power density, not on the maximum power rating. The maximum IT load for a data center is the average power density of the racks multiplied by the number of racks. Then, add in the power of the data center's power systems, cooling systems and facilities like switches and storage equipment.
In order to avoid incurring expensive costs and to take full advantage of the power system so that the data center has the right size power supply and space, it is wise to adopt a modular strategy to reduce costs without sacrificing flexibility for future expansion.
5. Whether operation and maintenance personnel should participate in planning and design
The answer, of course, is yes! Normally, you should do the following:
a. Operation and maintenance participation in preliminary planning and design can make up for designers' lack of knowledge on system operation and management, improve design quality, and avoid or eliminate design deficiencies;
b. Operation and maintenance participate in the preliminary planning and design, so that the needs of operation and maintenance stage can be fully considered in the planning and design;
c. Operation and maintenance participate in the preliminary planning and design, so as to fully understand and master the structure, reliability weak links, remaining problems and potential risks of the maintained system, which is conducive to improving the quality of operation and maintenance, and formulating operation and maintenance plans and feasible reconstruction plans based on the basis.
6. Avoid being influenced by internal and external factors
For those who do not properly treat the distinctions between preferences, preferences, limitations and constraints, and who do not abide by the principles of scientific design, there are suggestions as follows:
a. To avoid being influenced by individual decision makers in the process of approval and decision-making, some key functions should be deleted or adjusted according to the opinions of individuals, so that the final delivered data center functions cannot meet operation and maintenance requirements;
b. Avoid being motivated by inclination, preference or profit. In the planning and design stage, some equipment manufacturers may affect the planning and design personnel's scheme formulation and equipment selection by exaggerating equipment performance and confusing concepts.
7. What level of backup battery does the AC or DC rack need
The server rack may need to deploy 100 percent DC power, 100 percent AC power, or a combination of both.
For example, if a data center is being built for hosting purposes, there may be a need for an AC (UPS) power supply system, whereas if it is to become a telecommunications facility, it may require the deployment of a DC power system. Knowing this, you will know the size and scale of the DC power or UPS power system you need.
When deploying backup batteries, you are advised to configure them according to the 15-minute standby discharge period. Deploying backup batteries in this way does not significantly increase capital expenditure, which would be difficult to justify, but it is more cost-effective. Companies need to increase the redundancy of spare generating sets rather than waste more money on adding more batteries.
8, avoid light planning and design, heavy construction
There is a current situation in the industry that planning and design are less important than construction, which is mainly reflected in:
a. Building the building structure first and then planning the data center, which brings insurmountable difficulties for the planning and design of the data center;
b. It is common to start the transformation after the construction of the machine room and the installation of the equipment;
c. Determine the equipment first and then the scheme. Because the functions of the purchased equipment do not meet the requirements of the scheme or the site does not meet the requirements of installation conditions, the equipment needs to be replaced before it is in operation;
d. The construction structure is difficult to meet the site layout requirements of the data center, resulting in unreasonable zoning planning of the equipment room; The air conditioner outdoor unit cannot be installed or is too far away. The power room is too far away from the main machine room, which increases transmission and management difficulties, increases costs, and reduces reliability.
9. Avoid neglecting system maintainability and repairability design
System construction is divided into three parts and maintenance is divided into seven parts. Any equipment may fail. Quick repair has become a key indicator to improve availability.
The neglect of maintainability and repairability is manifested in:
a. Later maintenance channels and Spaces are not considered in the planning and design phase. For example, the device is too close to the wall, batteries are placed against the wall, the cable layout is unreasonable, the cable pipeline or cable tray blocks the upper weak wire slot, resulting in later maintenance failure and difficult to use maintenance tools;
b. When a fault occurs, emergency materials and spare parts cannot be quickly transported, and there is no maintenance operation space for the replacement of faulty equipment and components, which delays the time limit for troubleshooting and even leads to major accidents;
c. The redundancy guarantee capability of the system when the equipment is maintained without considering the failure in the later stage;
d. Use automatic methods as far as possible to minimize manual operations by maintenance personnel and reduce uncertainties and uncontrollability caused by manual operations.
10. Avoid a usability design that lacks science
System availability is the most important index of data center planning and design, but it lacks scientific basis in planning and design. The main manifestations are as follows:
a. During the planning and design of data centers, reliability calculation of various systems will be carried out. However, at present, the calculation basis and data source of reliability by different design institutes and different designers have not been unified, resulting in different definitions and different results for the design level and reliability of the same data center;
b. There are cases of usability planning, design and construction first, and then retrograde design after completion of construction, and promote the retrograde standard to data center users. This is the inverted design, often because of some key defects in the design, although most of the design meet the grade requirements, but because of the existence of key defects cause the grade decline;
c. Only pay attention to the availability of equipment or links (subsystems), and ignore the impact of the correlation between subsystems in the whole system on availability.
11. Avoid setting high targets in isolation from actual needs and feasibility
At the initial stage of planning and design, the functional indexes of data centers are determined subjectively, and the indexes of large scale, high availability level, high rack power density, and low energy consumption are pursued unrealistically.
However, in the specific planning and design, detailed demonstration was not made in strict accordance with the planning principles and procedures, and the specific scheme and practical measures did not match the overall planning. The result is:
a. Due to unclear actual demand and lack of necessary conditions for feasibility, repeated design and modification not only result in waste of cost, but also greatly prolong the construction cycle;
b. The machine room that has been built and put into use cannot be fully utilized, and the reason may be that there is no expected business demand or the conditions of the machine room do not meet the needs of users, so it must be re-optimized and reformed before it can be used;
c. The planned functions cannot be realized, for example, the system availability does not reach the planned level, the cooling solution does not support the planned high power density of the rack, the generator does not support the continuous running condition of the system, and the system energy consumption indicator (PUE) is too high due to excessive planning.
12, avoid the misunderstanding of heavy equipment light system
In the industry, especially in the planning and design personnel, there is a widespread phenomenon of emphasizing equipment and undervaluing system and neglecting the overall situation, which is mainly manifested in:
a. First determine the equipment specifications, models and even manufacturers, and then select or modify the design scheme according to the equipment;
b. The power supply system is designed with 2N redundancy of the highest available level. However, the result of the design is that only the UPS system has 2N functions, while the entire power supply system has single-path points of failure;
c. The whole system is designed according to the highest level of redundancy fault-tolerant system, but the air conditioning and refrigeration equipment is single path power supply;
d. The alternating-current backup energy diesel generator is configured for the system, but the generator does not have automatic starting function, because the planner lacks the understanding that continuous refrigeration is also an important condition for continuous operation of the system.
13. Pay attention to overall design
It is very important to improve the overall design ability of the system to complete the task of planning and design.
a. In the process of data center construction, a large number of delivered data centers fail to meet business and maintenance requirements due to insufficient consideration of phased and specialized implementation projects in planning and design, as well as the connection between construction of different specialties. Some problems even require a large amount of investment to complete the transformation;
b. Different designers only pay attention to the content they are responsible for, and lack overall consideration on whether other specialties are perfectly connected, and whether there are contradictions and conflicts between them;
c. Planners and designers lack an accurate grasp of future business development and insufficient consideration of capacity management and expansion in the later stage;
d. Planners and designers are not familiar with the surrounding resource environment and physical environment, and the designed scheme may have poor implementation feasibility or bring insurmountable difficulties to the later operation and maintenance work.
summarize
There are many other issues to consider in the construction of new data centers, but industry experience tells us that understanding these 13 key points in the construction and design process of data centers can help ensure that the final construction result is as close to the real needs of users as possible.
