JP2011069098A - Data center - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a complete data center which is composed of a building, and which can enhance the easiness, safety, flexibility and extensibility of operations, for example, in the updating of server equipment by enabling an infrastructure corresponding to characteristics to be supplied to a server group split depending on a utilization form, while sharing infrastructure equipment. <P>SOLUTION: This data center includes an infrastructure building 100 which has the infrastructure equipment, a CPU building 200 which is constructed in such a manner as to be connected adjacently to the infrastructure building 100 and which can be removed by itself, and an extension space 300 in which the CPU building 200 can be newly constructed. Each of the CPU buildings 200 houses a plurality of servers which are classified depending on the utilization forms. The infrastructure building 100 supplies a power source and a communication line to each of the CPU buildings 200, additionally supplies air conditioning to each of the CPU buildings 200 by an air-conditioning system, and enables the specifications of the power source, the communication line and the air conditioning for supply to be changed depending on the characteristics of the servers housed in each of the CPU buildings 200. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a data center construction technique, and more particularly to a technique that is effective when applied to a data center that can flexibly cope with various usage forms having different functions and purposes of use.

  In a data center, a large number of ICT (Information and Communication Technology) devices such as servers and routers (hereinafter sometimes simply referred to as “servers”) are stored using a server rack or the like. In general, a computer room or the like in which a large number of server racks are stored has a so-called large room specification, and server racks and other machines are arranged in a vacant space on a large floor. ing.

  In a large-scale data center, a large number of servers to be stored are, for example, in-house in-house systems and business systems for users (customers), and other companies and other customers have in-house systems and business systems. There are a wide variety of servers, such as those that have been outsourced for operation and management, and so-called housing services and cloud computing services. Therefore, a large-scale data center that accommodates these servers requires high flexibility and expandability.

  On the other hand, in contrast to the conventional data center as a building, in recent years, a so-called module (container) in which a network, a power supply, an air conditioner, etc. are incorporated into a container or prefabricated for transportation, and a server rack or the like is accommodated there. Type data centers are being used mainly in the United States. By configuring the data center as a module (unit) that can be easily moved and expanded in this way, the flexibility and expandability are greatly improved, and the data center can be quickly constructed and deployed. Further, since excessive space and equipment are not required, the environmental load can be reduced.

  Further, as a technique that can be used when constructing a large-scale data center using a plurality of such modular data centers, for example, US Pat. No. 7,278,273 (Patent Document 1) discloses a container for transportation. Describes a data center including one or more modules equipped with a plurality of computers and an air conditioning system for cooling. In addition, Non-Patent Document 1 describes the effective use of space, excellent environmental performance, expandability and flexibility, and construction period by vertically stacking modules (units) to adapt to the Japanese environment where the site area is small. Data center technology that enables shortening and the like is described.

US Pat. No. 7,278,273

“Nest Cloud”, [online], Taisei Corporation, [Search September 9, 2009], Internet <URL: http://www.taisei.co.jp/ss/it/menu01/menu012/ menu0123 / 1245629202176.html>

  As described above, in a conventional large-scale data center, since a large variety of servers are arranged in a large-room computer room, work efficiency when performing equipment update operations while the data center is in operation In addition, the length of wiring and piping increases, and the flow line becomes complicated, which often results in inefficiency and low flexibility and expandability.

  In addition, mixing various servers with different functions and purposes of use creates places with different local loads, and it is not possible to supply infrastructure such as power supplies and air conditioning uniformly in the data center. It is an efficient situation. For example, with the recent increase in the amount of heat generated by servers and the increase in the density of server storage, there is an increasing tendency that places that require a large air conditioning capacity locally are scattered in the data center.

  On the other hand, a so-called module (container) type data center can be easily constructed, for example, for each user (customer) or for each purpose of use, and has high flexibility and expandability. Therefore, it is possible to construct a large-scale data center using a plurality of these, for example, using a technique as described in Patent Document 1 and Non-Patent Document 1.

  However, the module type data center itself is not intended to be installed in a stationary manner, and is generally configured in a form suitable for small-scale use. On the contrary, it may be inefficient. For example, modular data centers each have or have to supply infrastructure equipment that should originally be shared in the data center, such as air conditioners and power supply / communication devices, to improve efficiency. In some cases, it may not be possible to take advantage of the benefits of a large-scale data center that operates efficiently by sharing resources such as infrastructure facilities.

  In addition, modular data centers are generally movable structures, so in Japan, laws and regulations such as the Building Standards Act, the Road Traffic Act and the Fire Services Act must be cleared for installation and construction. It is said that there are cases where the above regulations increase, and as a result, there may be a case where it is not possible to enjoy the merit in terms of easy construction and cost.

  Therefore, an object of the present invention is to store a large number of diverse servers, share resources of basic infrastructure facilities, and to a group of servers divided according to usage forms classified according to functions and intended use. It is possible to supply infrastructure according to the characteristics of each load, etc., and improve the ease and safety of work, flexibility, and expandability when adding, removing, and updating server equipment and facilities. It is to provide a complete data center consisting of buildings that can be made.

  The above and other objects and novel features of the present invention will be apparent from the description of this specification and the accompanying drawings.

  Of the inventions disclosed in this application, the outline of typical ones will be briefly described as follows.

  A data center according to a representative embodiment of the present invention is connected to an infrastructure building having an infrastructure facility including an air conditioner, a power feeding device, and a communication device, and around the infrastructure building adjacent to the infrastructure building. One or more CPU buildings that house a plurality of servers and can be removed independently, and one or more first expansions that are sites where the CPU buildings can be newly constructed Each of the CPU buildings contains a plurality of the servers classified according to usage, and the infrastructure building receives power from one or more substation facilities and one or more communication carriers. From the power supply device and the communication device to supply power and communication lines to the CPU buildings, and the air conditioning device supplies air conditioning to the CPU buildings. The specifications of the power supply, the communication line, and the air conditioning to be supplied can be changed according to the characteristics of the usage mode of the server housed in each CPU building. It is.

  Among the inventions disclosed in the present application, effects obtained by typical ones will be briefly described as follows.

  According to a typical embodiment of the present invention, in a complete large-scale data center composed of buildings, resources of basic infrastructure facilities are shared by the infrastructure building, and classified according to function, purpose of use, and the like. By dividing the CPU ridge as a building that houses the server according to the usage form in an appropriate unit, it is possible to efficiently supply infrastructure according to characteristics such as load. In addition, by adding, removing, and updating server devices and facilities in units of CPU buildings, it becomes possible to greatly improve the ease and safety of work and the flexibility and expandability.

It is the figure which showed the outline | summary of the structural example of the data center which is one embodiment of this invention. It is the figure which showed the outline | summary of the structural example of the data center which is one embodiment of this invention. It is the figure shown about the example of the update cycle of the building in a data center, an installation, and IT apparatus. It is the figure which showed the outline | summary about the example of the update method at the time of updating a server during operation | movement of a data center.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. Note that components having the same function are denoted by the same reference symbols throughout the drawings for describing the embodiment, and the repetitive description thereof will be omitted.

[Overview]
In a conventional large-scale data center, as described above, various server devices are stored in a large-room computer room. Therefore, work efficiency when updating server devices while the data center is in operation, etc. In addition, the wiring and piping length of cables such as power supply and communication increase, and the flow line becomes complicated, so that there are many cases where inefficiency and flexibility and expandability are low.

  FIG. 3 is a diagram illustrating an example of an update cycle of buildings, facilities, and IT equipment in a data center. In the example of FIG. 3, the useful life of a data center building (building) is generally about 30 years, and is updated about 30 years after completion. Further, the useful life of infrastructure facilities (air conditioners, power supply / communication devices, etc.) in data centers is generally about 15 years. That is, the update operation is performed once during the life cycle of the data center building. Similarly, the service life of IT equipment (servers) is generally about 5 years, and in this case, the update operation is performed five times during the life cycle of the data center building.

  When renewing a data center building, for example, a new data center building is built on another site, and at the same time, new infrastructure equipment and IT equipment are installed and migrated, thereby operating the current data center. Update work can be performed without stopping.

  On the other hand, infrastructure facilities and servers must be updated in the current data center building. At this time, since the operation as a data center cannot be stopped, it is necessary to perform an update operation in parallel with the operation of the data center. Here, the infrastructure facilities generally have a redundant configuration such as providing a spare in addition to the dual / duplex in the data center. For example, the update operation is performed by sequentially updating one system at a time. It is possible to do it.

  FIG. 4 is a diagram showing an overview of an example of an update method used when updating a server while the data center is operating. In the state a in the figure, the CPU 1 and CPU 2 (for example, a unit having a plurality of server racks and partitioned by walls, cages, etc.) are operating, and the empty 1 and empty 2 areas (multiple server racks). And a space that can be partitioned by a wall or a cage) is prepared.

  Here, when updating CPU1 and CPU2, as shown in the state of b in the figure, first, CPU1 ′ that has updated CPU1 is newly established in, for example, an empty space 1 (S1), and then CPU1 is removed. (S2). Similarly, as shown in the state of c in the figure, the CPU 2 'updated with the CPU 2 is newly installed, for example, in the space 2 (S3), and then the CPU 2 is removed (S4). Further, the area where the CPU 1 is located is maintained as a vacant 1 (S5). In this way, the update operation is not performed at the same place, and the update is performed by moving to each unit composed of a plurality of servers.

  The data center according to an embodiment of the present invention can be easily relocated or removed for each server group classified according to function or purpose of use in order to facilitate the above update work. Then, it is divided and stored in the CPU building, which is an independent building. Each CPU building has an infrastructure building for supplying basic infrastructure, and each CPU building is constructed so as to be connected adjacent to the infrastructure building.

  In this way, server functions such as in-house systems, business systems for users (customers), systems that have been outsourced for operation and management from other customer companies, so-called housing services and services by cloud computing, etc. In addition to improving the flexibility and expandability by dividing the server into larger units than the module (container) type data center and storing it in the CPU building according to the usage form classified by the purpose of use etc. This enables efficient infrastructure supply to each CPU building.

  In addition, each CPU ridge is constructed to connect to the infrastructure ridge as a building rather than a modular data center that can be moved, and this unit can be moved and removed to improve flexibility and expandability. It is possible to solve the problems in the case where the modular data centers are integrated into a data center.

[Constitution]
FIG. 1 and FIG. 2 are diagrams showing an outline of a configuration example of a data center according to an embodiment of the present invention. FIG. 1 shows an example of a functional configuration (arrangement) of each building (building) constituting the data center of the present embodiment as viewed from above, and FIG. 2 shows a part (infrastructure). 2 shows an example of the configuration when the building and the CPU building are viewed from the side.

  In FIG. 1, the data center 1 has, for example, an infrastructure building 100 having infrastructure facilities such as an air conditioner, a power feeding device, and a communication device, and is connected to the periphery of the infrastructure building 100 adjacent thereto. In this form, it has one or more CPU buildings 200 for storing a plurality of servers, and one or more warehouses 210 having a space for storing articles, kitting work for servers stored in the CPU buildings 200, and the like.

  The infrastructure building 100 has a seismic / isolation structure, etc., similar to a general data center building, and has a function of supplying basic infrastructure to each CPU building 200. For this reason, the infrastructure building 100 receives power of different diameters from different substation facilities (not shown) by the power receiving path 1 (141) and the power receiving path 2 (142), for example. In addition, communication lines 1 (151) and communication carrier 2 (152) are used to enter communication lines from a plurality of carriers (not shown). As a result, the redundancy of the power supply / communication infrastructure required by the data center 1 can be enhanced. In the example of FIG. 1, two systems are input for both the power supply / communication infrastructure, but the number of input systems may naturally be increased.

  These power / communication infrastructures are supplied to each CPU building 200 by a power supply device in the infrastructure building 100 and a communication device (not shown). The infrastructure building 100 further includes an air conditioner (not shown), and air conditioning is similarly supplied to each CPU building 200. The infrastructure building 100 has specifications such as the capacity of the infrastructure (air conditioning, power supply, communication) to be supplied according to the characteristics such as the load in the usage form classified according to the function and purpose of use of the server stored in each CPU building 200. It can be changed or controlled as appropriate.

  Each CPU building 200 is configured as an independent building that is not a movable unit such as a module (container) type data center. At this time, for example, one CPU building 200 has a server group for its own in-house system, another CPU building 200 has a server group that is outsourced and operated and managed by a client company, and another CPU building. 200 is a server group that provides services as cloud computing, such as server functions, purpose of use, load characteristics, grades for operation and management, etc. And store. As a result, the infrastructure building 100 can efficiently supply infrastructure to each CPU building 200 according to characteristics such as load.

  Further, since the CPU ridge 200 is a building independent of the infrastructure ridge 100, when the stored server device is updated by the update method as shown in FIG. By newly constructing on the site of the expansion space 300) or by removing only the target CPU ridge 200 alone to make a new expansion space 300, other CPU ridges 200 (operating in other usage forms) The update operation can be performed without affecting the server.

  Also, various constructions and maintenance work for the CPU building 200 can be performed independently without affecting the infrastructure building 100 and other CPU buildings 200. Each CPU building 200 may have a different seismic isolation / seismic isolation structure or the like according to the characteristics of the server to be stored, or may have a different building size.

  The same applies to the warehouse 210. In response to the establishment or removal of the CPU ridge 200, the warehouse 210 is newly constructed on the site of an empty area (additional space 310) or only the target warehouse 210 is removed alone. It is possible to make a new extension space 310. As described above, since the CPU ridge 200 and the warehouse 210 can be newly established or removed on a ridge basis, the data center 1 can have high flexibility and expandability.

  Furthermore, around the infrastructure building 100, for example, a backyard 110 and an entrance 120 are arranged at opposing positions. The backyard 110 is an entrance for carrying in servers, other devices, articles, and the like, and the entrance 120 is an entrance for a person such as a worker who enters from an approach 130 connected to the entrance 120. These may be a structure independent of the infrastructure building 100, or may be a part of the infrastructure building 100.

  Articles brought in from the backyard 110 and people entering from the entrance can reach the CPU buildings 200 and the warehouse 210 through the passage 101 provided on the outer periphery of the infrastructure building 100, for example. Each CPU ridge 200 and warehouse 210 are configured to be able to come and go with the infrastructure ridge 100 via, for example, a door having a security function. In this way, by separating the entry / exit port for goods and the like and the entrance / exit for humans, it is possible to prevent these flow lines from being unnecessarily crossed.

  Further, for example, one side of the approach 130 on the side facing the infrastructure building 100 across the entrance 120 has one or more office buildings 220 that perform operations such as management and office work for each unit such as a customer. Similarly to the CPU ridge 200 and the warehouse 210, the office ridge 220 is newly constructed on the site of the vacant area (additional space 320) in units of the office ridge 220, or the target office ridge 220 alone is removed and newly created. It is possible to make an additional space 320.

  In FIG. 2, the infrastructure building 100 and each CPU building 200 are each configured as a structure fastened to the ground, and each CPU building 200 is, for example, a door having a security function with respect to the infrastructure building 100. It is shown that they are constructed so as to be connected to each other via (not shown).

  In the CPU building 200, for example, a plurality of server racks 202 are accommodated on a double floor structure constituted by a so-called free access floor or the like. In these server racks 202, a large number of servers having some common usage forms such as functions and usage purposes are stored in multiple stages. Further, for example, a distribution board 201 for supplying power to these servers is also housed.

  Infrastructure (air conditioning, power supply, communication) is supplied from the infrastructure building 100 to the CPU building 200 and the devices (such as the server rack 202 (server) and the distribution board 201) in the CPU building 200. For example, the infrastructure building 100 includes an air conditioner 102, and supplies cool air for cooling the servers in the server rack 202 via an air supply duct 103 or the like disposed under the floor. Further, the hot air discharged after the server is cooled is collected through the exhaust duct 104 or the like disposed on the ceiling.

  Further, the power distribution / communication infrastructure inputted from the substation facility or communication carrier is supplied to the distribution board 201 by the power supply / communication cable 105 wired under the floor via the power supply device and the communication device (not shown) in the infrastructure building 100. Or the server rack 202 (and server).

  It is assumed that the air conditioner 102, the power supply device, and the communication device can control the specifications of the supply level, contents, and the like for each CPU building 200 according to the required infrastructure capacity and the like. Accordingly, it is possible to efficiently supply infrastructure to each CPU building 200 while sharing basic infrastructure facilities among the CPU buildings 200. At this time, a plurality of infrastructure facilities such as an air conditioner 102, a power feeding device, and a communication device may be provided as needed.

  As described above, according to the data center which is an embodiment of the present invention, functions and usage are shared in the infrastructure building 100 while sharing resources of basic infrastructure facilities in the large-scale data center 1. The server is divided into units larger than the case of the module (container) type data center and stored in the CPU ridge 200 for each usage form classified according to purpose. Thereby, it is possible to efficiently supply infrastructure to each CPU building 200 according to characteristics such as load.

  In addition, by constructing each CPU ridge 200 so as to be connected to the infrastructure ridge 100 as a building rather than a movable modular data center, it is possible to add / remove / update server devices and facilities, etc. It is possible to do with. This greatly improves the ease and safety of server device update operations and the flexibility and expandability of the data center, and the modular data center is integrated into the data center as a complete data center. The problems in building a center can be solved.

  As mentioned above, the invention made by the present inventor has been specifically described based on the embodiment. However, the present invention is not limited to the embodiment, and various modifications can be made without departing from the scope of the invention. Needless to say.

  For example, the arrangement and configuration of each function (building) shown in FIGS. 1 and 2 are merely examples, and the infrastructure building 110 is constructed such that the CPU building 200, which is an independent building, is connected. In addition, as long as the configuration has an additional space 300 that can newly construct the CPU ridge 200 in such a form, the shape of the infrastructure ridge 110, the arrangement position and size of other buildings, etc. It can be changed appropriately according to site conditions.

  The present invention relates to a technique that is effective when applied to a data center that can flexibly cope with various usage forms having different functions and purposes of use.

1 ... Data center,
DESCRIPTION OF SYMBOLS 100 ... Infrastructure building, 101 ... Passage, 102 ... Air conditioning apparatus, 103 ... Air supply duct, 104 ... Exhaust duct, 105 ... Power supply / communication cable, 110 ... Backyard, 120 ... Entrance, 130 ... Approach, 141 ... Power receiving path 1 142 ... Power receiving path 2, 151 ... Communication carrier 1, 152 ... Communication carrier 2,
200 ... CPU building, 201 ... distribution panel, 202 ... server rack, 210 ... warehouse, 220 ... office building,
300, 310, 320 ... additional space.

Claims (5)

A data center that houses a large number of servers,
An infrastructure building having infrastructure facilities including an air conditioner, a power supply device, and a communication device;
One or more CPU buildings that house a plurality of servers that are built around and connected to the infrastructure building and can be removed independently;
One or more first expansion spaces that are sites where the CPU building can be newly constructed,
Each of the CPU ridges stores a plurality of the servers classified according to usage patterns,
The infrastructure building receives power from one or more substation facilities, enters communication lines from one or more communication carriers, and supplies power and communication lines to the CPU buildings via the power supply device and the communication device. Furthermore, the air conditioning is supplied to each CPU building by the air conditioner, and the specifications of the power source, the communication line, and the air conditioning to be supplied are characteristics of the usage mode of the server stored in each CPU building. A data center characterized in that it can be changed according to the situation. In the data center according to claim 1,
One or more warehouses built around and connected to the infrastructure building and capable of being removed alone;
A data center having one or more second expansion spaces, which are sites where the warehouse can be newly constructed. In the data center according to claim 1 or 2,
Further, a backyard that is built around the infrastructure building so as to be connected adjacent to the infrastructure building, and serves as a loading / unloading port for articles including the server,
A data center, which is constructed so as to be connected adjacent to the infrastructure building and has an entrance serving as a doorway for people. In the data center according to claim 3,
Furthermore, one or more office buildings that are built on the side facing the infrastructure building across the entrance and can be removed independently;
A data center comprising one or more third expansion spaces that are sites where the office building can be newly constructed. In the data center according to any one of claims 1 to 4,
When updating the server stored in the CPU building,
The CPU building for storing the updated server is newly built in the first expansion space, the CPU building for storing the server before the update is removed, and the new first expansion is added. A data center that can be updated by using space.

Images