Novice users do not immediately think about such an important topic which is Internet traffic. They hardly know about this thing because such combination is necessary to pay only for time spent on the Internet, but not for such download.
This is understandable – that it can actually pump up a dial-up with the speed limit. Telephone line is not intended for downloading files-movies, music albums. Even the short songs dial-up pump difficulty, it is not the phone volume.
Another thing – a dedicated line or district network.
When a home user is considered as a dedicated channel – it immediately goes into several other consumer categories. Because now the consumption of the web is growing in several orders. What do you think how much download does a simple internet user do, on his computer when the Internet comes through fast local area network with the speed of about two hundred megabytes.
And that if he does not swing movies, is not fond of music in MP3 format (more accurately, the user is not fond of downloading stuff on to the computer from network) and not a big fan of all graphics and video. Two hundred megabytes – the usual thing, and it is all about the most common sites and the most common quantities.
In this case, Internet users already have to think about the fact of such traffic, and most importantly – how much of this traffic is spent. Because the monthly fee service provider usually includes a limited number of like prepaid, or rather free megabytes – it all depends on the tariff plans.
But very often it happens that the user does not just exceed the limit amount, and goes far beyond it. And then the user begins to treat traffic gently, especially figuring in how much does this traffic costs.
Do you think when you place your site on the Internet, you will not think about the traffic? Alas, in this case you have to remember the traffic. First, you should be done with a website and have a hosting package for your website.
If you think that you are not going to get massive traffic for your website and you pay for server hosting package that is not going to pass on any benefits then you will also need to develop your website, right?
And then you start on working over it, first, work hard, create new content, and besides, try to fill a site with different file contents.
Gradually the size of the site grows, it gains popularity, respectability, increase in attendance and downloads. And then the provider tells you – “Stop!”
This would mean that there is a lot of load on the hosted website, and your page traffic does not fit within the limits of hosting package, for which you pay two hundred to three hundred bucks a month. You have to think about that, or rent a server from your ISP, or just install your own server at your ISP (called colocation hosting). And then the term TRAFFIC confront you with all its inevitability.
Because in this case, there are same problems as like in the case of a dedicated internet network: the monthly fee includes some traffic (it is measured in gigabytes, because it is still on the server, not on your local computer) but as a rule, if free traffic per month is exceeded, you have to pay for it.
Some hosting providers prefer to pre-negotiate at such conditions or collect money for all traffic to avoid disagreements and recriminations.
Then you have a question – why, actually, you have to pay for traffic that uses your site? The users who visit your resource, pay their internet connection provider. And you pay for the traffic that comes from the users computer (on your site).
Then why you must pay for this traffic? It turns out that providers receive double payment – in the sense that the users take the money because they swing the information on your website, and you get the money for something that users pump data from your page. Two times to take money for the same thing – Is it fair? – You ask, and in this holy wrath of pure truth is almost one hundred percent.
Interestingly, such a paradoxical situation has actually occurred:
Providers have the money for the traffic, from users and site owners. Of course, one could say that users only pay for the time spent on web, and the website owner pay to keep their website up and running 24/4/365 days. So should not the creators of sites pay for traffic? I am deeply convinced.
Setting up a data center is a complicated and lengthy procedure. Would that help to standardize the process and reduce the risk of unreliable equipment and eliminate the need for additional investment in the elimination of defects?
According to the research, the world average volume of user data per one company that is 120 terabytes. In turn, estimated at IDC, last year for one hour in the world, dispatched 35 billion messages. Given these numbers, it turns out that one message has approximately 3.4 MB of information. Of course, most of the e-mails are not accompanied by voluminous reports. But not so little attachment contains a sample from corporate databases or presentations with graphic slides and video. The size of this investment – a few tens of megabytes. For the preparation of reports they utilized powerful computing resources. In addition, the database itself are handled by dozens of different kinds of applications, covering all aspects of the company.
Intensive data streams circulating in the information systems of many companies require a particular organization of the IT infrastructure. It must adapt the changing business requirements and, in particular, to ensure a steady increase in productivity of the decisions and maximize the effectiveness of their operation.
Concentration of resources
Solution to this problem may be the concentration of computing resources and their distribution functions between applications. This approach is known as “virtualization”. “Virtualization of resources” again sparked interest in the concept of centralized computing. The result of the implementation of this concept are data processing centers (DPC). Data Processing Center (DPC) – a group of premises whose main function is to place processing equipment and storage.
The option of data center represents all the leading manufacturers of computer and communications equipment, as well as software, which organizes the operation and management information systems. Typically, these solutions are used by the latest development companies. In turn, should pay special attention to IT infrastructure, within the centralization of computing resources.
Illustrating the need for a universal and productive infrastructure for data centers can be similar to the automotive industry.
The first cars on the design repeated the horse-drawn carriages, and only after some time they worked out universal principles of layout nodes, which remain unchanged. On the basis of these principles they kept improving the design of the car continuously, whose goal was to create an optimal functioning and is also more economical models.
At the same time, the vehicle, which is a cart on which the most modern car engine, of course, would not meet the performance characteristics and economic indicators. So with data centers are. Approaches for building the infrastructure of computing centers of the sixties and seventies, or even for server rooms, do not correspond to the equipment in the modern data center.
Standards for data centers adhere to the basic principle of constructing cable infrastructure topology “hierarchical star
Thus, there is need to develop a special standard that defines the design and construction of infrastructure for the data center. The modern concept of DPC is based on the definition of its functionality and services, implemented an information system within the enterprise, or services provided by the user.
In accordance with the purpose of data centers, they can be divided into two groups. First – it is private (or corporate) data centers that operate exclusively within a particular company. The second group includes common data centers, often owned by the suppliers of Internet services and are used to implement services such as web hosting, colocation hosting, rental applications, the deployment of e-commerce, etc.
But those and other centers primarily represent premises, which set the computers and network equipment and creates conditions for continuous and reliable operation of this equipment, and storage systems. Among the obligatory prerequisite for the normal functioning of the DPC include the presence of uninterrupted power supply, maintenance of temperature and implementation of network connections within the data center, as well as connections to the networks of telecommunications service providers.
Currently, there are two standards defining the principles of data center infrastructure. It was developed in the U.S. TIA/EIA-942 standard and European standard EN 50173-5. Both standards contain many similar provisions, but the scope of the American standard is much broader, since it determines not only the features of the organization cabling.
Number of European standard says that he belongs to a group of cable standards, and the number after the hyphen indicates the application in the respective areas. Thus, EN 50173-5 – is a European standard, largely determines the cable solutions for data centers.
Standard TIA/EIA-942 considering several options for building
American Standard considers the structure as a whole and contains not only general guidance on the organization of cable infrastructure, installation, mounting fixtures and identifying sites for laying cable. It also focuses on the design of the network, providing access, rules of placing data center, the architectural features of buildings, the organization of power, lighting, climatic conditions, to ensure the smooth operation of the equipment, fire safety and protection from moisture.
An important component of the standard is the requirement to ensure high operational readiness of equipment in the data center required to service requests that come from a large number of users. Given such a broad scope of issues related to the implementation of infrastructure data center, will continue to be considered primarily an American standard.
Approved in April 2005, the standards bodies, the U.S. standard TIA/EIA-942 Telecommunications Infrastructure Standard for Data Centers defines the requirements and the basic rules for the design and implementation of data centers and server rooms.
The original “starting point” of the standard is beginning design work prior to construction or reconstruction. Only at this stage we can fully appreciate all the architectural features of data center premises and ensure coordination of all technical systems. Therefore, guided by the standard should be primarily the designers, because they have to plan for the relationship building’s architecture, its technical systems and cabling infrastructure to the operation of a large number of computer equipment with a high density layout.
List of main components
Standard TIA / EIA 942 provides the mandatory provision of specialized facilities and organization of work areas. In particular, this space for liner external telecommunication systems (Entrance Room), computer hardware (Computer Room), telecommunications equipment (Telecommunications Room) and accommodation for engineering systems, eg, electrical, industrial space conditioning and ventilation systems, etc.
To monitor and manage data center (especially the center, which is mission-critical) is organized by the center of the current network management (Network Operations Center – NOC). Its function is to identify faults and develop action precluding such effects, as a possible simple computer equipment. In the NOC located equipment, it monitors the thermal regime, stop and monitor equipment malfunction with subsequent diagnosis of modules and blocks, which are out of order.
In addition to space for the installation of computer equipment in the building of a data center can allocate space for offices and support services, such as customer service centers or service training of data entry. These facilities include the switching points of horizontal wiring for office and support services.
The room for computer equipment allocated area of basic wiring (Main Distribution Area – MDA), horizontal wiring (Horizontal Distribution Area – HDA), zonal routing (Zone Distribution Area – ZDA) and the domain routing equipment (Equipment Distribution Area – EDA).
Possibility to add external systems (ER), MDA, HDA, ZDA and EDA although in many respects and are consistent with but not identical premises and areas specified in the standard TIA/EIA-568-V.1 (Entrance Facility, Equipment Room, Telecom Room, Consolidation Point and Work Area). In the European standard, EN 50173-5 are used with other names of elements of cable infrastructure. External telecommunications services are connected to an external network interface (External Network Interface-ENI), which connects to the main switching center (Main Distributor – MD) through a network access subsystem (network access cabling). In the zonal sub-system by wiring to sockets equipment (Equipment Outlets – EO), either directly or through the connection of local distribution points (Local Distribution Point – LDP). However, LDP is an optional element.
Location of premises and areas determined by the size of the data center, as well as the ability to install additional equipment and transition to more advanced communication technologies.
In the premises for the liner external telecommunication, systems are interfaces that connect structured cabling data center routes a group of buildings, as well as with cable equipment suppliers of telecommunications services. This may be a separate room (standard recommends that you make a separate room for security reasons), but allowed the union and with room for computer equipment. In general, the premises equipment, providing input of external telecommunications services, consolidated in the MDA.
In the data center there may be some room for liner external telecommunication systems, which allows to observe restrictions on the length of lines, as well as implement various service units. Field of basic routing (MDA) – a location of main switching center cabling data center. MDA is the most suitable place to install the routers and switches, the core of the local area network data center and network storage. In addition, this region can be integrated distribution points horizontal wiring servicing equipment in the immediate vicinity of the MDA.
Areas of horizontal wiring allocated for the implementation of distribution points horizontal subsystem, cable lines that reach the area of routing equipment. Therefore, HDA is considered as the location of switches, local area network and network storage, as well as KVM-Switch (allow you to control multiple servers through one set of “keyboard-video-mouse), which serve facilities in the respective EDA.
Additional switching centers zonal cable equipment, which correspond to the region zonal wiring with this optional element. They are placed between the HDA and EDA, where necessary frequent reconfiguration of cable equipment, or used as a means of providing additional flexibility in horizontal solutions. The horizontal cables that are suitable for ZDA, in zonal outlet or consolidation point. Further arrangement by means of switching cords.
In the zonal wiring, it is not recommended to install the active equipment, except for decisions on the organization of power in the twisted pair.
In the field of EDA implemented network, connections are needed for computer equipment.
Allowed additional connections between the HDA (including the purpose of redundancy) or connect to cable equipment rooms liner external telecommunication systems (when released a few such facilities).
Standard TIA/EIA-942 considering several options for constructing the cable infrastructure. The basic topology is best suited for any data center – and for corporate, located on the same site, and for the center of common use, dispersed across multiple platforms.
In distributed data centers, there is a allocated several rooms for liner external telecommunication systems. This is done for security reasons or if necessary to cover large distances in comparison with the maximum communication range.
The simplified topology of the data center involves the combination of HDA and MDA. In smaller data centers, these switching points can also be combined with equipment TR and ER. This system allows for increasing the length of the horizontal line on the optical fiber to 300 m.
In the data center with a centralized topology is realized an optical system with a central administration. In this system, all the electronics concentrates in MDA and the EDA, and the horizontal subsystem is missing.
Cable Equipment Data Center are the horizontal cabling, backbone wiring and switching equipment in the relevant fields.
Backbone wiring connects the Entrance Room, MDA and HDA. The horizontal wiring – is part of the SCS from the point of termination to the EDA to the horizontal distribution point in one area of HDA. As the backbone wiring actually aggregated traffic coming from the horizontal lines, it must have the appropriate bandwidth.
The implementation of horizontal and backbone postings, according to the standards implemented for the data center on a constant basis for SCS hierarchical star. It allow only one level of hierarchy for backbone wiring, which implies that one distribution point.
However, the standards provide redundant wiring. For this purpose we introduce the region of secondary wiring, switching point which is connected to the distribution points of the horizontal wiring, duplicating compounds with the basic wiring.
Standard EN 50173-5 includes the use of cable equipment to twisted pair and optical fiber. The minimum capacity of copper systems is determined by the class E. The minimum capacity for optics – Class OF-300, implemented fiber OM-2 or OM-3. The standard provides the length of optical fiber cable channels in the OM-1 OM-2 and OM-3.
TIA/EIA-942 Standard provides for the application of cable equipment, switching equipment and wiring cables in accordance with AN-SI/TIA/EIA-568-B.2 and VZ. This is a transmission medium of a twisted pair to the characteristic impedance of 100 ohms, a multimode fiber (62,5 / 125 or 50/125) and single-mode fiber.
In addition, a number of applications allow the use of coaxial cable with impedance of 75 ohms.
Process Longevity for the solutions that are implemented in the newly developed projects, provide some types of cable. In TIA/EIA-942 recommended twisted pair is sixth category and optimized for laser fiber. Given that at the time of TIA/EIA-942 absent standard twisted-pair sixth category is possibly an indication of the type of cable that is preferred for use in data centers.
The length of the cable channels (with patch cords) in the horizontal wiring for cable of any type should not exceed more than 100 m. For the solutions with a singular line, in which HDA is integrated with MDA, the length of optical cable ducts can reach 300 m.
Standard TIA/EIA-942 puts forward a number of requirements and recommendations concerning the organization of cabling. All computer and switching equipment is housed in enclosures and racks.
Raised floor has become an architectural foundation for building modern data centers (DC), but their use is not limited only by machine halls of data centers. Easy upgrade, as well as the ability to quickly add or remove cables, provided with modern modular system makes attractive use of raised floors and other premises.
Today, designers seek to establish Raised floor in different rooms. Raised floor makes it easy to upgrade the installation of various equipment, including ease of installation of new redundant cables, redevelopment of premises for other needs and contributes to their popularity among end users and customers. Raised floor has a broader consumer market. Over the past few years, the SAC upgraded quite significantly, and the construction of raised floors has remained unchanged. Raised floor makes sense to use when you need to run a large number of cables. They are more efficient and much less expensive than systems mounted near the ceiling, using which we must remember that the temperature near the ceiling is higher than the floor.
Instead of pulling cables from the ceiling, installers makes it easy to run cables under raised floors. Existing solutions in this area provide efficient cooling, reducing the number and extent of hidden cables, consolidation of physical ports, and fewer cords to connect the equipment.
Raised floor is a floor consisting of removable tiles that are installed on the supporting structure, because of which, the flooring and floor-base can fit for different purposes.
Raised floor space used for cabling
Among other things, under raised floors form ventilation system that provide cooling wind direction where it is needed. Raised floor is ideal for data centers, where ever there is new equipment or it is rearranged from one place to another, it is easy to reinstall. It also ensure proper maintenance of cable management: The cable is pulled under the floor, and it can be easily accessed, they are also easy to upgrade or shift.
Raised floor can be installed if the customer is implementing or planning to implement new technologies that requires reconfiguration of existing equipment or if he would need in the future. One of the main characteristics of raised floors is quick and easy access to cabling. You can quickly raise the deck and get access to the underground space.
The design of raised floor and the types of raised floors :
Pedestal raised floor construction Floating (movable) raised floor.
In the pedestal design, it support removable tiles that are used at column-fixed-height pedestals. Raised floor tiles are usually made of steel, aluminum or wood treated with flame retardant.
Flooring should settle at a height of not less than 150-300 mm from the floor base (concrete floor).
“Float” design should be built at a height of 460 mm from the floor. Vertical seismic load is absorbed by the dampers and springs that are installed inside the support cylinder. With regard to horizontal displacements, they are neutralized with steel shoes that are Teflon-coated and they are wore at the base of cylinders. Shoes can skim the surface of tiles and can be clamped to the floor. In the typical raised floor systems, they use Slub tile of 0.6 x 0,6 m and 0,36 square area.m size.
In order to supply cooling air to the places of high heat in front of the enclosure, perforated tiles are installed. Tiles for service under the raised floors systems can be removed from the floor and they can be shifted to other places.
Close to the particularly strong heating equipment, the devices that guides the air flow into the holes of perforated raised floor tiles can be installed. When the premises have to change something, tiles are removed from the floor and they are moved to another location.
In the initial arrangement of the room they all want to ensure optimal, as it seems, the balance of computer and communications equipment. But people often do not know at what point they will need this technique in the future, therefore they do not know what changes need to be made in the scheme of its deployment. It is therefore desirable to have a solution that allows you to quickly cost-shift the equipment, introduce new technologies to replace the equipment. If you are completely sure that you will never make any changes and do not need to introduce new technologies, raised floor is not for you.
Raised floor is usually associated with data centers, since a large amount of equipment is much easier to organize the flow of cooling air under the floor rather than from above. But the most difficult problem in data centers is the organization of cooling and access to cables. Often in the field of concentration equipment arise where equipment is overheating, so-called “hot spots”.
This problem is only exacerbated by the extensive use of blade servers. Typical data centers built before 2000, calculated on the heat, is equal to 5 kW per rack. Now, in today’s data centers, this figure could be range from 7 kW to 35 kW per server rack, or even higher because there are a lot of data centers who place more servers in small area, which give correspondingly more heat, so they often have hot spots.
Design Problems with Raised Floor
When you install the raised floor system, designers face a number of standard tasks, including – to set the tiles, to ensure reliable grounding and protection from electrical interference. In areas that are in front of places of racks, stacked perforated tiles. These metal surfaces should not be open, we must exclude them into direct contact with installed equipment. Metal tiles and raised floor is a supporting structure and must be well grounded. Openings for dragging cables to avoid mechanical damage to the latter should not have sharp edges.
When designing the raised floor, you have to make sure that:
* You have a sufficient powerful air conditioning system for cooling all the equipment;
* You have to consider the limitations of mechanical load on the raised floor and concrete floor;
* You should check that there is enough free space to install the equipments available;
* You should address the needs of areas for future development
When the layout of equipment will be ready on the floor, arrange appropriate stencils surfaces for installation of equipment and markup the input and output of cables.
With the development of cable technology and expanding the scope of raised floors, designers are facing new challenges. Twenty years ago the premises of data centers were the largest of those premises, which used Raised floor.
Now raised floors are installed in huge halls like a library, “clean rooms” for different industries. The use of raised floors allows customers to abandon expensive ceiling mounted systems, cables, besides, it does not require punching holes in the building constructions to liner cables.
That is what made Raised floor an attractive technical solution, and brought them outside the computer rooms of data centers and server rooms. Now the technology of raised floors, which was developed mainly for the latter, is largely growing demand from owners of office for general purposes.
Choose the flexibility to install
The modern office environment is characterized by frequent change of place. Update equipment and implement new technologies quickly and conveniently.
In the 90 years of last century, the main motivation of the raised floor is to access the power cable, telephone and computer system as fast as possible. Today more and more users are turning to such solutions for arranging other facilities. They clearly understand all the benefits that are promised by improving the quality of ventilation and air conditioning, as well as the ability to quickly rebuild the cable infrastructure, which in turn leads to a reduction in operating costs.
The increasingly widespread use of raised floors is logically correct for the tendency of enterprises. If you need to update equipment or alter the jobs in offices, but there is no raised floor, you have to deal with time-consuming and expensive process of transfer of cable lines. It is often required to drill floor and walls for dragging cables in new places, as well as withdraw hung ceiling for shifting air ventilation and air conditioning.
But if you have a raised floor, you only raise tile flooring, change the routing of telephone cables, data transmission network and power supply, then put the tiles on the space, enter the new office furniture or rearranges old , plugging the plug – and you are done. It would be unfair architecture, if you designed a raised floor and not drawing attention to other systems, an organic combination that provides a cost-effective operation of the building.
There are many different systems and therefore, in order to regulate costs, it is important to choose the right product. In order to present the conditions that make the building cost effective, they need to use flexible systems that can be reused or reconfigured.
Complex design issues with raised floors
As the scope of raised floors beyond the data center, it complicates their design and installation. If there is a need for reorganization or modernization, we propose that a system should be able to reconfigured. Additional difficulties are technological change and cable preferences.
In the data centers, raised floor can be installed without problems. But when installing such floors throughout the building shifted doors and elevators. We need to reconstruct the entire building to install the raised floor, and sometimes it can be done, and sometimes it can not.
The challenges of providing customer service and the restructuring of structured cabling systems (SCS).
As Raised floor is used in other areas of the building, installers are facing with the challenge of providing technical services and reorganization of the SCS in the office premises. One possible solution is to use raised floors and quickly plug cable systems, which reduces the time of the initial arrangement of office space, and enable rapid restructuring of jobs.
We need to know how long it takes for a permutation, the introduction of new technology, or changes to reconfigure the equipment.
Contemporary Raised floor must be equipped with removable grille, which could be moved during the permutation, extensions or changes. These systems must have strong support structures that will support the floor tile raised floor and provide easy access to cable household.
In the design of cable channels of the office premises, it is important to consider different outer diameter and the physical characteristics of pull cables. We need to lay cables in a certain way, because we can not just throw the cables on the floor. The outer diameter of standard Category 6 cable is much larger than the diameter cable category 5e. Shielded Category 6 cable can have a diameter of 0.251 inches, and unshielded cable enhanced category 6A could have a diameter of 0.315 inches.
Interferences under the raised floors and electrostatic electricity in data center
Another problem is the pickup, when the cables are twisted and the pair is laid under the floor. Static electricity is also produced in the underground systems. These problems may occur because of the abundance of microscopic whisker crystals of zinc, a side effect of the molecular pressure. Filamentary crystals of zinc are produced by friction on the metal surface, such as frames for raised floors.
They break the electrical contact, conduct electricity and can disrupt electronic devices, entering the electrical components and causing a short circuit. If the standard protection system could not deal with these electrical problems, it is necessary to take appropriate measures for grounding raised floors.
Data Center Processing is a complex engineering structure, intended for centralized deployment and maintenance of data center equipment.
There are many external factors that are affecting the quality of resources. Especially important and noteworthy is the negative impact of the environment and the human factor. To avoid this problem, the wide range of high-tech engineering solutions are used.
With Various solutions and methods of designing and building data centers, each in its own way is correct and usually corresponds to the norms and rules adopted in the construction and telecommunications. In this article I want to share the operation of data centers which includes power supply, cooling system and security.
Uninterruptible Power Supply
Data Center is one of the energy-intensive industries. Modern equipment, mounted in a 19 – inch rack unit height of 42, can consume up to 3 – 4 kW of electricity and provide a lot of heat. For its removal, you have to install air-conditioning, which consumes up to 50% of all power. As a result, equipment in data centers typically consume more kilowatts of electricity per square meter.
Power supply systems should be organized in two geographically separated transformer substations. Cable lines should route independently, each must be connected to “their” transformer. You will have to use automatic selection (ATS), exercising choice and switching between the main and backup lines.
Despite the existence of redundant power, it is desirable to use diesel generator power plant. In the scheme of power they are after you. That is, in case of a complete power outage or its non required parameters (voltage, frequency, and “purity”) is run automatically Data Encryption Standard, and the load is transferred to them.
DES, as a rule, it have a supply of fuel for up to 8 hours of continuous work and fills without stopping. Diesel Power Module can continuously work up to 3 – 4 months, subject to an established supply of fuel.
Next, install an uninterrupted power supply. It is absolutely necessary equipment, as the most dangerous power surges are short, lasting 2-3s. There are such negative factors, causing significant damage, as the excess voltage, frequency variation, grounding violation, the interphase potential, etc. UPS is operating in the mode of on-line, is an ideal Isolated, consuming “dirty” power from the city or DES, and issuing an absolutely clean power, the frequency of 50 Hz, 220 – 380 watts, without extra harmonics.
Creating such a serious electrical power supply system, data center requires a thorough approach. Design and construction perform specialized design and installation in organization, operation is carried out in accordance with the rules and regulations of its own engineering service company that owns the data center.
The task of cooling is maintained within the data center operating temperature range from 19 to 24 ° C and humidity from 40 to 80%. Typically, data centers of medium size (area 100 – 200 square meters. M) use cabinet-Precision freon air conditioners, air takes heat from the top room and force cold air under the raised floor. Calculation of power system cooling is a factor 1, ie, 1 kW power equipment is require 1 kW refrigeration. Reservation system is carried out by the scheme N +1. What does this mean? When we require refrigeration system cooling total of 100 kW and are available in cabinet driers freon air conditioners to 25 kW cold, you should install 5 cases: 4 main producing 100 kW, and a backup in case of failure of one of them.
Experience shows that this is a very effective method, especially if all five boxes are connected into a single management system. In this case, the software rotates the role of spare cabinet, which allows more efficient cooling system to expend resources in general. If the total heat generation data center equipment exceeds 150 – 200 kW, and is expected to further increase capacity step by step, it is advisable to install liquid coolant.
Use the following scheme. The street is established powerful chiller, the cooling water-based coolant to 16 ° C. Coolant circulates through the pipe by placing the data center, where a cabinet-type air-conditioners are circulating air. Heated agent returns to the chiller for cooling, the circle closes. Capacities are based chillers might be restricted to financial capacity.
19-inch cabinet, totally filled with modern server equipment, capable to provide up to 20 kW of heat. It should be remembered that due to structural and physical features, the usual way is almost impossible to remove more than 5 kW of heat, ie, blowing cold air from under the raised floor.
To resolve this problem, there are several approaches, in particular, the organization of “hot” and “cold” corridors. Under the corridor means a passage between rows of cabinets. In the “hot” corridor sent fans, blow-hot air from the servers, but from the “cold” to take the cold air blown out from under the raised floor through the lattice. This scheme allows to significantly raise the efficiency of refrigeration.
Also required to organize a flow of fresh air from outside. The fact that the air is constantly circulating through the computer cabinets and air conditioners, “fade” and requires updating. Influx through a custom install, heating and desiccating air of the street. In addition, it creates extra pressure inside the data center, which prevents penetration of dust.
For humidification steam generators are used. Dry air is not very effective for cooling systems because of the physical principles of air conditioning. With decreasing humidity of the electrostatic potential increases, which may cause equipment failure. Cooling system – a complex and delicate mechanism. As practice shows, this is the most critical and unreliable component of a complex data center. If it stop at 30 minutes, it can lead to heat rooms up to 60 – 70 ‘C, which entails the failure of equipment.
Raised floor is a necessary component of the data center. Under a pressurized cold air under it are power cables and low-power infrastructure. Typically, a raised floor is made of MDF tiles with a metal base with laminated cover, size 600 x 600 mm. Height above floor level – from 100 to 800 mm, for the most optimal data center 350 – 500 mm.
Early Detection Of Fire and Gas Fire
Maximum system efficiency gas fire should work in an earlier stage of development of fire detection, ie, when there is a corruption of heating elements or the initial inflammation, and in less than one minute extinguish the heat (and) fire. The complex is warning about fire and fire fighting should inform about the potential for ignition much earlier than would have to use extinguishing system.
This is achieved by installing a large number of high-sensitivity smoke, optical, chemical, spectral and other fire alarms tied into a single intelligent fire alarm and suppression, as well as the complex institutional arrangements. It includes the constant visual inspection of equipment, compliance with fire regulations, and rules of operation of electrical installations. Very effective is the system of super-early fire warning VESDA (Very Early Smoke Detection Aspirator), which allows you to detect a potential source of fire, long before its occurrence.
There are many types of fire extinguisher for automatic fire extinguishing systems, suitable for extinguishing fires in data centers. These include a variety of gases – halons, inert, carbon dioxide. It also use fine-dispersed water and powders. In this case, there is a principle – the more expensive solution, the more will remain in good serviceable after activation of the fire. The most expensive, but the most loyal to the equipment, are extinguishing mixtures based on Freon or inert gases.
CCTV and differentiation of physical access – the most important attributes of a modern data center.
The most effective is the system of access control based on proximity-cards. It consists of a server management, system controllers and readers, as well as maps (keys). This solution is inexpensive but effective measure of protection. In conjunction with CCTV, it is able to provide a reasonable level of security for data center.
Each door is equipped with a installed card reader, lock and a video camera. At an official request employees and customers are issued by a private key, which is the entry ticket to the territory of the physical perimeter of the data center. Typical attributes of key-picture of the owner, his personal details and the name of the company where he works. The key is always a staff member and provides easy access to the necessary facilities. Their list allow the access time prescribed in management of dedicated servers at institution account and linking it to a particular key.
Along with the system of separation of physical access, video surveillance system is a mandatory part of the organizational and technical measures of data center. It consists of an N-th number of video cameras installed so as to monitor virtually all the technical premises, entrances, exits, passages and hidden squares data center becomes easier. The video image comes to security monitors and and they are archived on digital media which creates another “virtual” level of security for data center services.
Cloud computing is an emerging technology being used by businesses in many different ways in order to help them save money but also so that they can update their IT systems to reflect the changes in attitude and technology. As well as cloud web hosting services being available for businesses that require added reliability, cloud computing services are also available. The main idea of cloud computing is to allow users to store all of their data in one location on the internet that can then be accessed by the many different devices that they may be using to view their data such as desktop computers and mobile devices.
A cloud computing service will allow you to keep all of your data in synchronization constantly meaning that if you upload a new file to your cloud space from one computer, it will be instantly viewable on another computer that is also part of your cloud – granted that the second computer is connected to the internet. Another good example of this will be with emails since if you have your email accounts setup on two devices, say a computer and your mobile device, then any changes that you make to the emails in your account will automatically be reflected on the other device; for example if you receive a new email and read it then this change of state will be reflected on all devices other than the one that you are using. This type of synchronization is useful for those that are using different devices since for example it will mean that you won’t need to have constantly dock your mobile device with your desktop computer to reflect any changes.
If you choose a cloud web hosting service then you will realize that cloud computing can also offer increased reliability over traditional single-server configurations. Reliability is an issue that is increasingly threatening to the web hosting industry these days as customers are expecting web hosting companies to be able to provide them with at least 99.9% guaranteed uptime. A cloud web hosting environment is one that will be making use of a large number of servers and each server will be designed for a specific role, with each role having its own dedicated server cluster; this type of configuration allows web hosting companies to guarantee high levels of reliability as it means that if the primary server fails, there will be at least one ‘fallback’ server available to take up the role of the affected server.
Data security is something else that cloud computing can offer as the data that you upload to your cloud space is likely to be replicated to more than one location meaning that even if a primary storage cluster does fail, the company providing you with your cloud computing services should be able to recover your data for you. If a business is going to rely solely on cloud computing in order for its IT requirements to be met then it is going to want the guarantee that any data, including emails and calendar events, are going to be kept safe and secure and that there will be no data loss in any situation.
If a company wishes to offer cloud computing services then it is going to have taken extra security precautions to help safeguard the data of its users as if anyone gets in trouble for lost or stolen data, it is going to be the company that is hosting it. There will be multiple security measures in place on both servers and the wider network in a cloud hosting environment as this will help to secure all data stored on servers as well as any information that may be passing around the cloud internally. Encryption is also a popular method available as it will mean that the data stored will not be readable unless a key is provided to decrypt it; although this is something that web hosting providers may not wish to do themselves, it may be a good idea for you to consider encrypting anything that you choose to upload to the cloud.
By choosing cloud services for your major business requirements including collaborative services such as calendaring and email, you will be able to relieve certain duties from your IT staff since you are going to be outsourcing such services to a third-party. If you are currently using your own servers to deal with services such as calendaring and emails then by outsourcing you will be able to scrap these services, thus saving you money. Furthermore by reducing the amount of work that your IT staff have to do, you may need to lay off some of your staff; although this is going to have a negative impact on them, it will allow you to save more money in the long term.
It is a good idea to only consider larger companies that are offering cloud computing services for the time being as they are the only ones that are going to have the capital and resources that will allow their cloud computing services to be effective. A cloud computing environment requires a large number of servers if it is to be effective and successful in meeting its primary principle, although smaller web hosting companies have the power to higher their own dedicated servers it is going to work better when a company has direct access to its servers. Smaller companies also won’t have the funds to allow them to purchase the numbers of servers that are required.
In conclusion, cloud computing has many advantages that it can provide businesses with both in the short term and long term. Although reliable cloud computing requires significant investment, if you have the money and are able to justify the cost then you will find cloud based services such as cloud hosting well worth the investment. It is a good idea to research the various cloud computing providers there are available if you are going to be spending a good amount of money as this will provide you with the best chance of finding a package that suits your requirements.
A dedicated server is probably going to be the most expensive form of web hosting that you will ever purchase and so as well as making sure that you purchase the right dedicated server for your requirements, you need to be sure that your server will be managed properly. Whilst most users of dedicated servers will have the experience that will allow them to manage their dedicated servers properly, for more inexperienced users there are managed dedicated server hosting services available from certain web hosting providers. A managed dedicated server hosting package will allow you to focus in more important tasks as the management of your server will be taken care of by a professional support team that is likely to have a wide range of experience in more or less every issue that could arise with your dedicated server. It is recommended that you choose a larger web hosting company if you are going to use a managed dedicated server as this will provide you with the benefit of having access to a highly knowledgeable support team that can be contacted 24×7; although smaller web hosting providers may be able to offer you a more personal level of service, in the case of server hosting you want to be sure that any issues can be sorted quickly.
There are many different issues that could affect the reliability of your dedicated server which you may not be able to see to yourself and by using a managed dedicated server you won’t have to worry about these issues being a problem as you can simply contact your web hosting provider’s support team and have them sort out any issues for you. Most dedicated support teams have the benefit of being able to physically access the servers that they are maintaining which bodes well in your favour since it means that if there is a problem with the hardware of your server, this can be swapped out and replaced as necessary.
When choosing a managed dedicated server hosting service for your requirements, it is important to evaluate the features of the web hosting company that you may be looking at as well as those of the dedicated server that you are reviewing. It is important for any web hosting provider that is offering managed dedicated servers to be able to provide 24×7 support to all customers as if there are any server issues, it means that they will be dealt with promptly regardless of what day it is and what time of day it is. Furthermore it is important for your chosen web hosting company to employ knowledgeable support staff who will be able to deal with any issues that face them; if a support team contains staff members that aren’t able to answer even the most basic of technical questions then such a support team is going to be useless in providing helpful and reliable web hosting support. Well established web hosting companies are the best to consider in such a situation since they are likely to have a large number of happy, past customers whom you can rely on to provide an honest account of their web hosting provider’s level of service and performance.
A fully managed dedicated server hosting service is likely to provide you with a service that will offer a higher level of uptime rather than if you were to manage your dedicated server yourself. A support team will be able to professionally optimize any applications that you have installed on your server to use less system resources which will free more up to be used by the applications that you are using mainly on your server as these are the ones that are going to be important to you. If your server does happen to crash for whatever reason then the support team will be alerted and will response as they see fit to help minimize the level of downtime that could be incurred and therefore minimizing the overall impact of the situation – if you were to manage your dedicated server yourself then you would need to employ your own monitoring services and would need to notify your web hosting company yourself if your server was unavailable.
The specifications of your dedicated server are also important, as they will dictate the overall performance of your server. A high specification dedicated server is going to be the most expensive option and is only recommended for businesses that have a distinct need for performance. A medium specification dedicated server is probably more than suitable for most users if it is managed properly as a majority of server applications can be optimized to make more conservative use of system resources. For those on a restricted budget there is nothing wrong with using a low specification dedicated server, although depending on what your requirements are you may find that it isn’t powerful enough for your needs.
Making the most of managed dedicated server hosting is an easy task – all you need to do is leave everything related to the management of your dedicated server to the support team that your web hosting provider offers. If you are making the most of your web hosting service then you can be sure that you are getting good value for money and because of this it isn’t worth considering managed dedicated server hosting if you feel that you have the skills that will allow you to do a good job of managing your server on your own.
In conclusion, a managed dedicated server is the best route to go if you don’t have any previous experience of using either VPS server hosting or dedicated server hosting. Any major web hosting company should be able to provide managed dedicated servers, although you need to realize that you are going to be paying an added premium for the right to have access to a support team when you require it. There should be no difference in the level of support available for Linux servers and Windows servers, although if a web hosting provider specializes in one over the other then there may be slight differences.
Co-locations centres are a type of data centre where many customers host their servers on powerful and fast networks. Co-location centres are often referred to as ‘carrier hotels’ as multiple telecommunication providers tend to offer connection peering points at co-location centres. Co-location centres are often made up of several data floors that are simply large rooms for full of data racks and are normally designed to host thousands of servers, providing larger companies with the space that they require in order to make the most of their IT infrastructure. By using co-location server hosting you are able to make the most of server hosting whilst owning the equipment without having the expense of running your own data centre.
Advantages of Outsourcing Server Hosting and Data Center Functions
By choosing co-location server hosting, you are effectively outsourcing your server hosting operations to a third-party as another company will be providing you with the bandwidth, power and server space that you require in order for a server to be of any use to use. If you have chosen a co-location provider that will provide you with full server support then you will effectively have no power over your server yourself as anything that you require of it will have to go through the support team that you have been assigned to. The main benefits of outsourcing in this way include:
Physical Features of a Co-location Centre
Co-location centres generally share certain features to help ensure a high standard across all data centres that are constructed.
Almost every data centre will have many different levels of security in place to help ensure that only authorized personnel are granted access to confidential and sensitive areas of the building, namely the data floors. Popular security features include:
Data Floor Features
Data floors are the large rooms, often spanning entire floors that servers are hosted in. Most data centres have a large number of data floors to help keep things separate and in some situations clients may require their own server rooms. Most data floors feature:
A co-location centre will normally provide you with a connection to a network that peers with several different providers to help bring external traffic into the data centre and to deliver it to the correct servers. Using multiple peering providers will allow you to ensure that your website will load fast from anywhere in the world – by just using a single connection you can never be sure that all of your visitors will be provided with a satisfactory browsing experience when using your website.
Multiple connection providers will also provide you with the guarantee of not having to worry about your server’s availability if one of the connections fails. Traffic can simply be routed via another connection of one of the main connections fails, although this will be rather slow. However, having the guarantee of traffic always being able to reach your server is an important must for most companies.
In conclusion, it is important for you to do your research when choosing a co-location centre as there are many factors that need to be considered. If you are going to be hosting a large number of servers then you may find it to be of benefit to you to visit data centres that are local to you so that you can get a real idea of the level of service that you are going to receive as well as what you are paying for.