Method for interworking data centers

Abstract

A method for interworking data centers according to one embodiment of a first aspect of the present invention comprises the steps of: selecting, among a plurality of data centers, data centers approved for use of computing resources provided therein; checking information about available computing resources for each of the selected data centers; receiving a request for a predetermined computational task from a predetermined user; requesting the computational task from at least some of the selected data centers on the basis of the checked information; and receiving a result of the computational task from the at least some data centers and transmitting the result to the user.

Description

Data Center Integration Method

[0001] The technology described below relates to techniques for data center operators to interconnect and control data centers.

[0002] For reference, the present application claims priority based on Korean patent application filed on October 21, 2024 (Application No. 10-2024-0143897). The entire contents of the said application, which form the basis of this priority, are cited in the present application as reference.

[0003] A data center is a physical space or facility that houses computing systems and hardware equipment for building and running applications. Depending on its intended use, a data center can be built as an on-premise data center or a third-party data center. With the recent increase in artificial intelligence services, the demand for data centers is also growing significantly.

[0004] Users utilize data centers to build or execute specific applications. In this context, users are companies or individuals that provide specific services using data centers. By using data centers, users can provide their services to service users without the burden of management.

[0005] Data centers are constructed by individual operators. These heterogeneous data centers exhibit diverse communication specifications and protocols. Consequently, to utilize the resources of a specific data center, data center operators must either use dedicated programs provided by the center or develop their own programs for accessing it. As such, conventional data centers suffer from poor service scalability and interoperability because they require independent protocols and interfaces specific to each operator.

[0006] The technology described below aims to provide a unified protocol for information exchange and interoperability between data center operators and data centers.

[0007] A data center interlocking method according to an embodiment of the first aspect of the present invention comprises: a step of selecting a data center that has approved the use of computational resources provided within a plurality of data centers; a step of verifying information regarding available computational resources for each of the selected data centers; a step of receiving a request for a predetermined task operation from a predetermined user; a step of requesting the task operation to at least some of the selected data centers based on the verified information; and a step of receiving the result of the task operation from at least some of the data centers and delivering it to the user.

[0008] Each of the above multiple data centers may be assigned a unique ID. In this case, for each of the above multiple data centers, the use of computing resources may be approved only if the data center ID sent by the control device to each of the above multiple data centers matches the ID assigned to the corresponding data center.

[0009] The information regarding the above available computing resources may additionally include information on the status of the corresponding data center as follows: State 1-1, where additional resource allocation is possible; State 1-2, where it is normal but all resources are in use; State 1-3, where it is serviceable but under maintenance; and State 1-4, where it is serviceless.

[0010] The above method may further include the step of requesting additional computing resources other than the identified available computing resources from each of the selected data centers.

[0011] Among the selected data centers, there may be data centers that cannot allocate the requested additional computing resources. In this case, the method may further include the step of requesting the additional computing resources from a data center other than the selected data center among the plurality of data centers.

[0012] Information regarding the available computing resources may include (i) the type of computing device, the number of computing device cycles, and the storage space capacity, or (ii) the computing performance and storage space capacity of the computing device.

[0013] The above method may further include the step of transmitting a request for control of computing resources to at least some of the selected data centers; and the step of verifying information on available computing resources for each of the data centers that received the request for control.

[0014] A control device for a data center according to an embodiment of a second aspect of the present invention comprises a computing device and a communication device, wherein the computing device selects a data center that has approved the use of computing resources provided within it among a plurality of data centers, confirms that information regarding available computing resources for each of the selected data centers has been received through the communication device, confirms that a request for a specific task operation has been received from a specific user through the communication device, requests the task operation from at least some of the selected data centers through the communication device based on the confirmed information, and receives the result of the task operation from the at least some of the data centers through the communication device and transmits it to the user through the communication device.

[0015] A computer program stored in a non-transient computer-readable recording medium according to an embodiment of a third aspect of the present invention comprises instructions for the processor to perform a method, which, when executed by a processor, includes the steps of: selecting a data center that has approved the use of computational resources provided within a plurality of data centers; verifying information regarding available computational resources for each of the selected data centers; receiving a request for a predetermined task operation from a predetermined user; requesting the task operation from at least some of the selected data centers based on the verified information; and receiving the result of the task operation from the at least some of the data centers and delivering it to the user.

[0016] A non-transient computer-readable recording medium storing a computer-executable instruction according to an embodiment of the fourth aspect of the present invention, wherein the computer-executable instruction, when executed by a processor, comprises the steps of: selecting a data center that has approved the use of a computational resource provided internally among a plurality of data centers; verifying information regarding available computational resources for each of the selected data centers; receiving a request for a predetermined task operation from a predetermined user; requesting the task operation from at least some of the selected data centers based on the verified information; and receiving the result of the task operation from the at least some of the data centers and delivering it to the user.

[0017] The technology described below enables data center operators to facilitate service support for data centers. The technology described below allows operators to make unified service requests to different heterogeneous data centers. Therefore, the technology described below enables data center operators to flexibly utilize data centers in the process of providing services to users.

[0018] Figure 1 is an example of a system that provides data center services.

[0019] Figure 2 is an example of the interconnection process between a data center operator and a data center.

[0020] Figure 3 is an example of operation after interoperability between a data center operator and a data center.

[0021] Figure 4 is another example of operation after interoperability between a data center operator and a data center.

[0022] Figure 5 is an example of an interconnection process between a data center operator and a data center considering the user.

[0023] Figure 6 is an example of a data center control device.

[0024] The technology described below is subject to various modifications and may have various embodiments, and specific embodiments are illustrated in the drawings and described in detail. However, this is not intended to limit the technology described below to specific embodiments, and it should be understood that it includes all modifications, equivalents, and substitutions that fall within the spirit and scope of the technology described below.

[0025] Terms such as first, second, A, B, etc., may be used to describe various components, but such components are not limited by the said terms and are used solely for the purpose of distinguishing one component from another. For example, without departing from the scope of rights of the technology described below, the first component may be named the second component, and similarly, the second component may be named the first component. The term "and / or" includes a combination of multiple related described items or any of the multiple related described items.

[0026] In terms used in this specification, singular expressions should be understood to include plural expressions unless the context clearly indicates otherwise, and terms such as “includes” should be understood to mean that the described features, number, steps, actions, components, parts, or combinations thereof exist, and not to exclude the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0027] Before providing a detailed description of the drawings, it is to clarify that the classification of components in this specification is merely based on the primary function each component is responsible for. That is, two or more components described below may be combined into a single component, or a single component may be divided into two or more components based on more subdivided functions. Furthermore, each component described below may additionally perform some or all of the functions of other components in addition to its own primary function, and it goes without saying that some of the primary functions of each component may be exclusively performed by other components.

[0028] Furthermore, in performing the method or operation method, each process constituting the method may occur differently from the specified order unless a specific order is clearly indicated in the context. That is, each process may occur in the same order as specified, may be performed substantially simultaneously, or may be performed in the reverse order.

[0029] The technology described below relates to access or interoperability techniques for using data centers. The technology described below can provide a unified communication protocol for accessing various data centers and exchanging information.

[0030] First, the terms used below will be explained.

[0031] A data center is a physical space or facility that accommodates computing systems and hardware equipment for building and running applications. Therefore, a data center may not refer exclusively to specific hardware devices. However, hereinafter, a data center is defined as a distributed processing unit that provides computing power. A data center includes computational processing units, storage devices (memory, hard disks, etc.), and communication devices for transmitting and receiving data.

[0032] Computational resources may consist of at least one of a GPU (Graphics Processing Unit), CPU (Central Processing Unit), memory (RAM), and auxiliary storage media (hard disk, SSD (Solid-State Drives), etc.).

[0033] A data center operator is an entity that operates data centers to provide specific computing resources to users. A data center operator provides services by utilizing multiple data centers. The control device of a data center operator provides computing resources to users by exchanging specific information with multiple data centers. Hereinafter, "data center operator" refers to the control device on the side of the data center operator. However, for the convenience of explanation below, it may be simply referred to as "data center operator."

[0034] Services provided by a data center operator may be referred to as data center services. Data center services include the provision of computational resources and computational processing requested by users.

[0035] A user is an entity that receives specific computing resource services through a data center operator. Users can be companies or individuals. Users can request computing resource services from the data center operator and receive such services through servers or terminals.

[0036]

[0037] Figure 1 is an example of a system (100) that provides data center services.

[0038] The user device (111, 112) is a device at the user end requesting data center services. The user device (111, 112) may be a device such as a user terminal, a server, etc. The user device (111, 112) may be a server providing a specific application (e.g., an AI-based service). Or the user device (111, 112) may be a personal terminal on which a specific application runs.

[0039] A data center operator (121, 122) receives a request for a certain data center service from a user device and provides the data center service. As previously described, the data center operator (121, 122) refers to a control device belonging to the data center operator. The control device may be implemented in the form of a server, a PC, a chipset with an embedded solution, etc.

[0040] FIG. 1 illustrates a plurality of data centers (151, 152, 153, 154, and 155). As previously mentioned, each data center is an independent device that provides a certain amount of computational processing.

[0041] A data center operator (121) receives a request for a specific data center service from a user device (111). The data center operator (121) finds the data center(s) required for the requested task. The data center operator (121) can connect with multiple data centers (e.g., 151, 152, 153, and 154) based on the available computing resources of the data centers. The data center operator (121) requests a computation from the connected data center(s) and receives the computation result. The data center operator (121) provides the data center service according to the computation result to the user device (111).

[0042] A data center operator (122) receives a request for a specific data center service from a user device (112). The data center operator (122) identifies the data center required for the requested task and links with a specific data center(s). The data center operator (122) requests a computation from the linked data center(s) and receives the computation result. The data center operator (122) provides the user device (112) with a data center service based on the computation result.

[0043] For the sake of convenience of explanation in the drawings below, data center operators are referred to as SP (service provider), and data centers are referred to as DC (data center).

[0044]

[0045] Figure 2 is an example of the interconnection process (200) between a data center operator and a data center.

[0046] Data center operators must secure computing resources to provide data center services for user devices. Data center operators can first interconnect with multiple data centers. By allocating computing resources from each data center, data center operators can manage the total resources available for the service.

[0047] Allocated resources are computing resources available to the data center operator. Allocated resources are not physically distinct resources, but rather refer to resources possessing the maximum amount required to perform the requested task.

[0048] Data center operators and data centers can exchange information as shown in Table 1 below.

[0049] Meaning of Components Configuration Data Center Operator ID (SP ID): The ID of the operator providing services to the user. This ID is unique globally. The Data Center Operator ID is a fixed value and is formed during initial configuration. Service Type: The type of service provided by the Data Center Operator. Examples of service types include: e.g., cloud, telecommunications, AI inference / computation, etc. Data Center ID (DC ID): The Data Center ID assigned to allocate resources to the user. This ID is an identifier that is unique across the same operator, geographical area, or globally. The Data Center ID is a fixed value and is formed during initial configuration. Required Resource for SP: The amount of computational resources required by the Data Center Operator for service operation. Computational resources are classified into CPU, GPU, NPU, Memory, and storage. Alternatively, they can be classified into FLOPS, TOPS, and Storage. Total Resource of Data Center Computational Resources (Total Resource of DC ) Total amount of available computing resources in the relevant data center. Computational resources are classified into CPU, GPU, NPU, Memory, and storage. Alternatively, they can be classified into FLOPS, TOPS, and Storage. Assigned resource for SP. Computational resources allocated to the data center operator by the relevant data center. Computational resources are classified into CPU, GPU, NPU, Memory, and storage. Alternatively, they can be classified into FLOPS, TOPS, and Storage.

[0050] FLOPS (Floating Point Operations Per Second), used to classify computational resources, is a unit primarily used to numerically represent computer performance. FLOPS stands for floating-point operations per second and is based on the number of floating-point operations a computer can perform in one second. FLOPS operations include basic arithmetic operations as well as operations such as root, log, and exponential, with each counted as one operation. TOPS (Tera Operations Per Second), also used to classify computational resources, is a unit used to numerically represent computer performance in the same way as FLOPS; however, while FLOPS is a unit related to floating-point operations, TOPS is a unit based on integer processing.

[0051] Specifically, the data center operator and the data center perform the following interconnection process. FIG. 1 illustrates an exemplary data center operator (121) and data centers (151, 152).

[0052] A data center operator (121) sends a request for usage approval to a data center (151) to be connected, including its own SP ID (identifier), service type, and target DC ID (201). Meanwhile, in some cases, the data center operator (121) may not send a request for usage approval to a specific target data center, but instead send a message to find a data center with available computing resources.

[0053] The message transmitted in step 201 can be configured as shown in Table 2 below.

[0054] Registration Request {Sender {Service provider ID {}Service type {}}Target Data center {Data center ID {}}} # end

[0055] The data center (151) can verify its DC ID (202) and send a message acknowledging the use of computing resources to the data center (121) (203). The message sent in step 203 (Registration Request confirm) can be configured as shown in Table 3 below.

[0056] Registration Request confirm {Confirm {Data center ID {}}Identified Sender {Service provider ID {}}} # end

[0057] Meanwhile, the data center operator (121) may send a request for approval of use to the data center (152), including its own SP ID, service type, and target DC ID (211). At this time, the DC ID is assumed to be the ID of the data center (151). The data center (152) may check the requested DC ID and confirm that the requested target is not itself (212). In this case, the data center (152) may send a message rejecting approval of use of computing resources (access rejection) to the data center operator (121) (213). At this time, the access rejection message may include the ID of the data center (152) that is the incorrect target. The message (Registration Request reject) sent in step 213 may be configured as shown in Table 4 below.

[0058] Registration Request reject {Wrong target {Data center ID {}}} # end

[0059] A data center operator (121) may request resource information from a data center (151) (221). The resource information requested at this time may include the total resources of the data center. Additionally, the resource information at this time may include resource information required for the task. The message (Resource Request) transmitted in step 221 may be configured as shown in Table 5 below. The resource information required for the task may be composed of CPU, GPU, NPU, memory, and storage space items. CPU, GPU, and NPU may be defined by the number of cores (x) and the number of cycles (Ghz). Memory and storage space may include the required storage capacity. Alternatively, the resource information required for the task may be composed of FLOPS, TOPS, and storage space items.

[0060] Resource Request {Total Resource of DC Request {}Required Resource for SP {CPU {x core, yz GHz}GPU {x core, yz GHz}NPU {x core, yz GHz}Memory {x Gbyte}Storage {x Gbyte}orFLOPSTOPSStorage {x Gbyte}}} # end

[0061] The data center (151) can check its total resources and available computing resources (222). The data center (151) can transmit information on allocatable computing resources to the data center operator (121) based on the information on currently available computing resources (223). The message (Assigned Resource) transmitted in step 223 can be configured as shown in Table 6 below.

[0062] Assigned Resource {Total Resource of DC {CPU {x core, yz GHz}GPU {x core, yz GHz}NPU {x core, yz GHz}Memory {x Gbyte}Storage {x Gbyte}}Assigned Resource for SP {CPU {x core, yz GHz}GPU {x core, yz GHz}NPU {x core, yz GHz}Memory {x Gbyte}Storage {x Gbyte}orTFLOPSTOPSStorage {x Gbyte}}} # end

[0063] The data center operator (121) can send a message to the data center (151) confirming information on available computing resources (224). The data center operator (121) can send a computing request message to the data center (151) along with information on necessary computing resources (231). At this time, the information on necessary computing resources may be all available computing resources, or the information on necessary computing resources may be a part of all available computing resources.

[0064] A data center (151) can perform a certain operation using computational resources upon request (232). A data center operator (121) receives the operation result from the data center (151) (233). A data center operator (121) can provide a service based on the operation result to a user who requested the data center service (241).

[0065] Meanwhile, at each step of FIG. 2, if a reply is not received from the other party within a certain time t, the data center operator or data center may resend the message. The time t can be set in units such as 1 second, 2 seconds, 5 seconds, 10 seconds, etc. The maximum number of resends can also be set to various values ​​such as 2, 4, 8, 10, 20 times. If there is no reply even after the data center operator or data center has resent the message up to the maximum number of resends, the resends may be stopped and the connection status may be reset. Meanwhile, after the data center operator resets the connection, an attempt to connect to the same data center may be made after a certain time y has elapsed. At this time, the time y can be set to various values ​​such as 10 minutes, 20 minutes, 30 minutes, 1 hour, etc.

[0066]

[0067] Data center operators must ensure the maintenance of service quality by receiving updates on information such as whether the data center is operating without issues and whether there are changes in available resources, and by periodically renewing interoperability certifications. Data center operators may check the interoperability status periodically or non-periodically after connecting with a specific data center. Data center operators may update interoperability information periodically or non-periodically after connecting with a specific data center. Data center operators may also maintain interoperability by performing re-certification procedures periodically or non-periodically after connecting with a specific data center.

[0068] FIG. 3 is an example of operation (300) after linkage between a data center operator and a data center. FIG. 3 assumes that the data center operator (121) is linked to the data center (151) and the data center (152).

[0069] A data center operator can check whether to continue maintaining the connection periodically or non-periodically after connecting with a specific data center. In this case, the fixed interval can be set to various values, such as 10 minutes, 30 minutes, 1 hour, 2 hours, 4 hours, 6 hours, 12 hours, or 24 hours. Depending on the situation (e.g., an emergency), the data center operator may check whether to maintain the connection at an arbitrary time rather than at the fixed interval.

[0070] The data center operator (121) can send a message to the currently connected data center (151) to check the current operational status (normal / abnormal) of the data center (301).

[0071] The message (Status Report Request) transmitted in step 301 can be configured as shown in Table 7 below.

[0072] Status Report Request {Sender {Service provider ID {}Service type {}}Target Data center {Data center ID {}}} # end of Status Report Request

[0073] The data center (151) checks its available computing resources and operation schedule (302). The data center (151) transmits the DC operation status, including the confirmed information, to the data center operator (121) (303). At this time, the operation status may include whether the data center is functioning normally and available resources. The message (Status Report Response) transmitted in step 303 may be configured as shown in Table 8 below.

[0074] Status Report Response {Data center ID {}DC Status {}Connection Status {}} # end of Status Report Response

[0075] DC status{} can be any one of Good (normal status and additional resource allocation available), Full capacity (normal but full resource usage), Under maintenance (service available but under inspection), or Unavailable (service unavailable). Connection Status{} can be one of Good (connection normal), Need to disconnect (request to disconnect), Need to pause (request to temporarily disconnect), or Terminated (forcibly terminated connection from DC). The data center operator (121) may request additional computing resources from the data center (151) (311). The additional computing resource request message used at this time may be composed of information in a form similar to Table 5.

[0076] The data center (151) can check the currently available computing resources (312) and send an additional computing resource approval message to the data center operator (121) (313). At this time, the additional computing resource approval message may be composed of allocatable resource information similar to Table 6.

[0077] The data center operator (121) may send an additional computation request message to the data center (151) along with computation resource information (321). At this time, the computation resource information may be the entire allocatable resource. Or, the computation resource information may be a part of the entire allocatable resource.

[0078] The data center (151) can perform certain additional operations using computational resources upon request (322). The data center operator (121) receives the results of the operations from the data center (151) (323). The data center operator (121) can provide services based on the results of the operations to the user who requested the data center service.

[0079] The data center operator (121) can send a message to the currently connected data center (152) to confirm the current operational status (normal / abnormal) of the data center (331). At the same time, the data center operator (121) can request additional computing resources from the data center (152) (331).

[0080] The data center (152) checks its available computing resources and operation schedule (332). The data center (152) transmits the DC operation status, including the checked information, to the data center operator (121) (333). At this time, the operation status may be an abnormal state or an inability to allocate additional resources. If a problem occurs in the data center's operation status or if it becomes impossible to provide resources, the data center (152) may transmit a message to stop the connection or temporarily suspend the connection.

[0081] The data center operator (121) can stop the connection and stop the service to the user by receiving a connection stop / temporary connection suspension message from the data center (152) (341).

[0082] A data center operator (121) sends a request for usage approval to another data center (153), including its own SP ID, service type, and target DC ID (351). The data center (153) can verify its own DC ID (352) and send a request for usage approval message for computing resources to the data center (121) (253). The data center operator (121) can send a request for computing to the data center (153) along with information on necessary computing resources, taking into account available resources (354). Subsequently, the data center operator (121) can perform additional computing through the data center (153) and provide services to the user based on the results of the computing.

[0083] Meanwhile, at each step of FIG. 3, if a reply is not received from the other party within a certain time t, the data center operator or data center may resend the message. The time t can be set in units such as 1 second, 2 seconds, 5 seconds, 10 seconds, etc. The maximum number of resends can also be set to various values ​​such as 2, 4, 8, 10, 20 times. If there is no reply even after the data center operator or data center has resent the message up to the maximum number of resends, the resends may be stopped and the connection status may be reset. Meanwhile, after the data center operator resets the connection, an attempt to connect to the same data center may be made after a certain time y has elapsed. At this time, the time y can be set to various values ​​such as 10 minutes, 20 minutes, 30 minutes, 1 hour, etc.

[0084]

[0085] FIG. 4 is another example of operation (400) after linkage between a data center operator and a data center. FIG. 4 is an example in which the data center first reports its operational status and checks whether to maintain linkage. In FIG. 4, it is assumed that the data center operator (121) is linked with the data center (151).

[0086] The data center (151) checks its operational status (401).

[0087] The data center (151) can check whether to continue the connection periodically or non-periodically after connecting with the data center operator (121). At this time, the set period can be set in various ways, such as 10 minutes, 30 minutes, 1 hour, 2 hours, 4 hours, 6 hours, 12 hours, 24 hours, etc. Depending on the situation (e.g., emergency situation), the data center (151) can check whether to maintain the connection at any time other than the set period.

[0088] The data center (151) can send a message to the currently linked data center operator (121) to confirm the current operational status (normal / abnormal) of the data center (402). The message (Status Report) sent in step 402 can be configured as shown in Table 9 below. Each status information in the message below is as described above.

[0089] Status Report {Data center ID {}DC Status {}Connection Status {}} # end

[0090] Meanwhile, the operation status message may further include information about the currently assigned resource (Assigned Resource for SP) and the changed resource status (Changed Resource for SP), as shown in Table 10 below.

[0091] Assigned Resource {Assigned Resource for SP {CPU {x core, yz GHz}GPU {x core, yz GHz}NPU {x core, yz GHz}Memory {x Gbyte}Storage {x Gbyte}orTFLOPSTOPSStorage {x Gbyte}}Changed Resource for SP {CPU {x core, yz GHz}GPU {x core, yz GHz}NPU {x core, yz GHz}Memory {x Gbyte}Storage {x Gbyte}orTFLOPSTOPSStorage {x Gbyte}}} # end of Assigned Resource

[0092] The data center operator (121) may send a request message for adjusting computing resources to the data center (151) based on the status of the data center (151) or a new computing request (411). At this time, the message may include information about the computing resources currently required. The data center (151) may check its available computing resources (412) and send information about available computing resources to the data center operator (121) (413). The data center (151) may maintain a certain amount of available computing resources based on the requested computing resource adjustment request (414). Alternatively, depending on the case, the data center (151) may no longer allocate all or part of the current computing resources to the data center operator (121). The data center operator (121) may delete the connectivity information when it receives a connection suspension message or a forced connection termination message from the data center (151).

[0093] When a data center operator (121) receives a message from a data center (151) to temporarily suspend the connection, the connection with the data center may be maintained, but the resources of the data center may be excluded from the service.

[0094] If the data center operator (121) lacks the currently required computing resources, it may request the use of computing resources from a new data center (153). The data center operator (121) sends a request for usage approval message to another data center (153) including its own SP ID, service type, and target DC ID (421). The data center (153) checks its own DC ID (422) and can send a message approving the use of computing resources to the data center (121) (423). The data center operator (121) can send a request for computing to the data center (153) along with information on the required computing resources, taking into account available resources (424). Subsequently, the data center operator (121) can perform additional computing through the data center (153) and provide services to the user based on the results of the computing.

[0095] Meanwhile, at each step of FIG. 4, if a reply is not received from the other party within a certain time t, the data center operator or data center may resend the message. The time t can be set in units such as 1 second, 2 seconds, 5 seconds, 10 seconds, etc. The maximum number of resends can also be set to various values ​​such as 2, 4, 8, 10, 20 times. If there is no reply even after the data center operator or data center has resent the message up to the maximum number of resends, the resends may be stopped and the connection status may be reset. Meanwhile, after the connection is reset by the data center operator, an attempt to connect to the same data center may be made after a certain time y has elapsed. At this time, the time y can be set to various values ​​such as 10 minutes, 20 minutes, 30 minutes, 1 hour, etc.

[0096]

[0097] In order for communication between a data center operator and a data center, it must first be possible to efficiently manage information regarding the user who requested the service. FIG. 5 is an example of an interoperability process (500) between a data center operator and a data center that takes the user into consideration. A user (user device, 111) applies to a data center operator (121) for subscription to data center services and is assigned their own identifier (ID). Subsequently, the data center operator (121) can perform certain data center services based on the user ID. The information used in this process is as shown in Table 11 below.

[0098] Meaning of Components Configuration User ID (User ID) Global User ID. This ID is unique to the same data center service users, specific geographical areas, or worldwide. Composed of Service Provider ID + In-service User ID Data Center Operator ID (SP ID) ID of the provider offering services to the user. This ID is unique to the world. The Service Provider ID is a fixed value and is formed during initial configuration. Indicated User ID (In-service User ID) An ID assigned by the provider to distinguish users within the service while accommodating them. An ID assigned by the service provider to the user, fixed after initial setup. Required Computational Resources (Required Resource for user) When the service provider receives a task request from a user, the provider calculates the computational resources required for that task and allocates the total amount of necessary resources. Computational resources are classified into CPU, GPU, NPU, Memory, and storage. Alternatively, they can be classified into FLOPS, TOPS, and Storage. Assigned Computational Resources (Assigned Resource for user) The amount of computational resources assigned. Computational resources are classified into CPU, GPU, NPU, Memory, and storage. Alternatively, they can be classified into FLOPS, TOPS, and Storage. Assigned Data Centers: A set of data centers assigned to allocate resources to a specific user. The first assigned data center is 1, the next is 2, and so on, managing subordinate data centers in this order. Data Center ID (DC ID): The ID of the data center assigned to allocate resources to a specific user. This ID is unique on a global basis. The Data Center ID is a fixed value and is formed during initial configuration. Amount of Computational Resources Allocated from a Data Center: The amount of computational resources allocated to a user from that data center. Computational resources are classified into CPU, GPU, NPU, Memory, and Storage. Alternatively, they can be classified into FLOPS, TOPS, and Storage.

[0099] A user ID can be composed of a combination of a data center operator ID and an in-service user ID assigned to the user by the data center operator. The user ID within a data center operator is information that identifies users at the level of a specific data center operator. The data center operator ID is a globally unique ID that can be defined by the data center operator during the initial business establishment. For example, a data center operator ID can be defined as a country code + a serial number (e.g., when registering as a data center operator in Korea, Operator A: 820000001, Operator B: 820000002).

[0100] The required amount of computing resources can be determined by calculating the amount of computing resources required for a given operation when the data center operator receives computing operations from a user.

[0101] The allocated amount of computing resources refers to the total sum of computing resources allocated to a data center operator from the data center for the corresponding user.

[0102] An allocation data center refers to a set of data centers used by a data center operator to allocate computing resources to users.

[0103] The assigned data center ID is a globally unique ID that can be defined during initial deployment.

[0104] The amount of computing resources allocated in an assigned data center refers to the resources allocated by that data center for that user.

[0105] The user (111) sends a user registration request message (Service Registration Request) to the data center operator (121) (501). The message in step 501 can be defined as shown in Table 12 below.

[0106] Service Registration Request {Target Service provider {Service provider ID {}}} # end

[0107] The data center operator (121) can set a user ID within the service (502). The data center operator (121) can transmit the user ID within the service to the user (111) (503). The message (Service Registration Response) of step 503 can be defined as shown in Table 13 below.

[0108] Service Registration Response {Confirm {Service provider ID {}}User ID {Service provider ID {}Indicated User ID {}}} # end

[0109] The user (111) can send a message to the data center operator (121) that the user ID within the service has been verified (504). The message of step 504 (Service registration complete) can be defined as shown in Table 14 below.

[0110] Service registration complete{User ID {Service provider ID {}Indicated User ID {}}} # end

[0111] A data center operator (121) can define and register a user ID (505). As described above, the user ID may be information that combines or combines a user ID within the service and a service provider ID. Subsequently, the user (111) can request data center services from the data center operator (121) (511). The data center operator (121) can determine the necessary resources based on the task requested by the user (111) (512).

[0112] The data center operator (121) sends a request for usage approval to the data center (154), including its SP ID, service type, and target DC ID (521). The message (Service Request) in step 521 can be defined as shown in Table 15 below.

[0113] Service Request {User ID {Service provider ID {}Indicated User ID {}}Target Service {Work type{}}} # end

[0114] The data center (154) can verify its DC ID and available resources (522) and send a message acknowledging the use of computing resources to the data center (121) (523). The data center operator (121) can send a computing request message to the data center (154) along with information on necessary computing resources, taking into account available resources (531). At this time, the data center operator (121) can send a confirmation message to the user (111) that the work is being performed (541). The message (Service response) of step 541 can be defined as shown in Table 16 below. That is, the data center operator (121) may provide the user (111) with information on computing resources currently being used for computing.

[0115] Service response{User ID {Service provider ID {}Indicated User ID {}}Service status{Resources of user {CPU {x core, yz GHz}GPU {x core, yz GHz}NPU {x core, yz GHz}Memory {x Gbyte}Storage {x Gbyte}}orTFLOPSTOPSStorage {x Gbyte}}} # end

[0116] The data center (154) can perform the requested operation (551) and transmit the operation result to the data center operator (121) (552). The data center operator (121) can transmit the service result based on the operation result to the user (111) (553).

[0117] FIG. 6 is an example of a control device (600) of a data center. The control device (600) is a device that controls computational operations using the data center(s) by a data center operator. The control device (600) can be physically implemented in various forms. For example, the control device (600) can take the form of a network server, a chipset dedicated to data processing, etc.

[0118] The control device (600) may include a storage device (610), a memory (620), a computing device (630), an interface device (640), and a communication device (650).

[0119] The storage device (610) may be a storage medium such as a hard disk, flash memory, or SSD (Solid-State Drives).

[0120] The storage device (610) can store a program for data center operation and control.

[0121] The storage device (610) can store information of available data centers (data center ID, computational resource information, etc.).

[0122] The storage device (610) can store information of currently linked data centers (data center ID, available computing information, operational status, etc.).

[0123] The storage device (610) can store applications, artificial intelligence models, and data for computations transmitted by the user.

[0124] The storage device (610) can store the results of operations performed by the data center.

[0125] The memory (620) can store data and information generated during the process of the control device connecting to, linking with, and controlling the data center.

[0126] The interface device (640) is a device that receives certain commands and data from the outside.

[0127] The interface device (640) can receive commands for data center linkage and control.

[0128] The interface device (640) can receive data center service requests from a physically connected input device or external storage device.

[0129] The interface device (640) can receive information about available data centers (data center ID, computational resource information, etc.).

[0130] The interface device (640) may also receive the result of an operation performed by the data center.

[0131] Meanwhile, the interface device (640) may refer to a device configuration that transmits data received through the communication device (650) below into the control device.

[0132] The communication device (650) refers to a configuration that receives and transmits certain information through a wired or wireless network.

[0133] The communication device (650) can receive a service subscription request from the user device.

[0134] The communication device (650) can receive certain feedback information from the user device. The information transmitted by the user device to the control device is as described in FIG. 5.

[0135] The communication device (650) can receive a data center service request from a user device.

[0136] The communication device (650) can transmit a request for approval to use computing resources to the data center. At this time, the communication device (650) can transmit an SP ID, DC ID, user ID, etc.

[0137] The communication device (650) may receive certain feedback information from the data center during the data center interlocking process. The information transmitted by the data center to the control device is as described in FIGS. 2 to 5.

[0138] The communication device (650) can receive approval for the use of computational resources from the data center.

[0139] The communication device (650) can transmit a request for resource information of a specific data center to the data center. In response to this, the communication device (650) can receive the total amount of computing resources, the amount of available computing resources, etc. of the specific data center.

[0140] The communication device (650) can transmit a request for computational resources required for a specific task to a data center. In response to this, the communication device (650) can receive information on computational resources available for allocation from a specific data center.

[0141] The communication device (650) can transmit a computation request requiring a certain amount of computational resources to the data center.

[0142] The communication device (650) can receive computation results from a specific data center.

[0143] The communication device (650) can transmit a request to a specific data center to check the operational status of the data center after connecting to the data center. The communication device (650) can transmit status information of the data center periodically or non-periodically. In response to this, the communication device (650) can receive status information of the specific data center. At this time, the status information may include normal / abnormal status and / or available resource information.

[0144] The communication device (650) can transmit a request for additional computation processing to a specific data center. In response to this, the communication device (650) can receive additional computation results from the specific data center.

[0145] The communication device (650) can receive status information of a specific data center periodically or non-periodically after connecting to the data center. At this time, the status information may include normal / abnormal status and / or available resource information.

[0146] The communication device (650) can transmit the computation result of the data center to the user device. Alternatively, the communication device (650) can transmit the service result based on the computation result of the data center to the user device.

[0147] The computing device (630) can set a user ID within the service of the user device according to the user device's subscription application. The computing device (630) can generate a unique user ID of the user device by combining or combining the user ID within the service of the user device and the SP ID.

[0148] The computing device (630) can determine the computing resources required for a task based on a certain computing task request from a user device. At this time, the computing resources can be expressed as a specific number or capacity of computing resources as defined in Table 5, etc.

[0149] The computing device (630) can generate various requests to be transmitted to the data center according to a predefined format. Tables 2, 5, and 7 are examples of message formats.

[0150] The computing device (630) can determine the data center(s) capable of processing the current task among the available data centers. The computing device (630) can determine the data center(s) that satisfy the total amount of computing resources required to perform the task.

[0151] The computing device (630) can control the process of linking with the data center required for computing.

[0152] The computing device (630) can control the interruption or suspension of the data center's connection according to the status information of the specific data center.

[0153] The computing device (630) can determine whether additional computing resources are required depending on changes in work content or changes in the state of a specific data center. If additional computing is required, the computing device (630) can determine the computing resources required for the additional computing. The computing device (630) can control the process of requesting resources for additional computing from the currently linked data center or a new data center.

[0154] The computing device (630) may be a device such as a processor, AP, or a chip with a program embedded in it that processes data and performs certain operations.

[0155]

[0156] The methods according to the embodiments described in the specification of the present disclosure may be implemented in the form of hardware, software, or a combination of hardware and software.

[0157] When implemented in software, a computer-readable storage medium may be provided for storing one or more programs (software modules). One or more programs stored in the computer-readable storage medium are configured for execution by one or more processors within an electronic device. One or more programs include instructions that cause the electronic device to execute methods according to embodiments described in the specification of this disclosure.

[0158] In addition, the method of interoperability or authentication between a data center and a data center operator as described above may be implemented as a program (or application) including an executable algorithm that can be executed on a computer. The program may be provided by storing it on a non-transitory computer-readable medium.

[0159] A non-transient readable medium refers to a medium that stores data semi-permanently and can be read by a device, rather than a medium that stores data for a short period of time, such as a register, cache, or memory. Specifically, the various applications or programs described above may be stored and provided on a non-transient readable medium such as a CD, DVD, hard disk, Blu-ray disc, USB, memory card, ROM (read-only memory), PROM (programmable read-only memory), EPROM (Erasable PROM, EPROM), EEPROM (Electrically EPROM), or flash memory.

[0160] Transient readable media refers to various types of RAM such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDR SDRAM), Enhanced SDRAM (ESDRAM), Synclink DRAM (SLDRAM), and Direct Rambus RAM (DRRAM).

[0161] Additionally, the program may be stored on an attachable storage device that can be accessed via a communication network such as the Internet, Intranet, LAN (local area network), WAN (wide area network), or SAN (storage area network), or a combination thereof. Such a storage device may be connected to a device performing an embodiment of the present disclosure through an external port. Additionally, a separate storage device on a communication network may be connected to a device performing an embodiment of the present disclosure.

[0162] The embodiments and drawings attached to this specification merely clearly illustrate a part of the technical ideas included in the aforementioned technology, and it is self-evident that all variations and specific embodiments that can be easily inferred by a person skilled in the art within the scope of the technical ideas included in the specification and drawings of the aforementioned technology are included within the scope of the rights of the aforementioned technology.

Claims

1. A data center interlocking method performed by a control device of a data center operator, wherein A step of selecting a data center that has approved the use of internal computing resources among multiple data centers; A step of verifying information on available computing resources for each of the selected data centers above; A step of receiving a request for a specific operation from a specific user; Based on the above-mentioned confirmed information, a step of requesting the above-mentioned work operation from at least some of the above-mentioned selected data centers; and A step comprising receiving the result of the above-mentioned operation from at least some of the data centers and delivering it to the user. Data Center Interconnection Method 2. In Paragraph 1, Each of the above multiple data centers is assigned a unique ID, and In each of the aforementioned multiple data centers, The use of computing resources is approved only if the data center ID sent by the control device to each of the plurality of data centers matches the ID assigned to the corresponding data center. Data Center Interconnection Method 3. In Paragraph 1, The information regarding the above available computing resources includes, Information regarding the status of the corresponding data center is additionally included in the 1-1 state where additional resources can be allocated, the 1-2 state where it is normal but full resources are in use, the 1-3 state where it is serviceable but under maintenance, and the 1-4 state where it is serviceless. Data Center Interconnection Method 4. In Paragraph 1, The step of requesting additional computing resources other than the identified available computing resources from each of the selected data centers further Data Center Interconnection Method 5. In Paragraph 4, Among the selected data centers mentioned above, there are data centers that cannot allocate the requested additional computing resources, and The above method is, The method further includes the step of requesting the additional computing resources from a data center other than the selected data center among the plurality of data centers. Data Center Interconnection Method 6. In Paragraph 1, The information regarding the above-mentioned available computing resources includes (i) the type of computing device, the number of computing device cycles, and the storage space capacity, or (ii) the computing performance and storage space capacity of the computing device. Data Center Interconnection Method 7. In Paragraph 1, The step of sending a request message for control of computing resources to at least some of the selected data centers; and The method further includes the step of verifying information regarding available computing resources for each data center that received the above adjustment request message. Data Center Interconnection Method 8. A non-transient computer-readable recording medium storing at least one computer-executable instruction, wherein, when the at least one instruction is executed by a processor, A step of selecting a data center that has approved the use of internal computing resources among multiple data centers; A step of verifying information on available computing resources for each of the selected data centers above; A step of receiving a request for a specific operation from a specific user; Based on the above-mentioned confirmed information, a step of requesting the above-mentioned work operation from at least some of the above-mentioned selected data centers; and A method in which the processor performs a step including receiving the result of the above-mentioned work operation from at least some of the data centers and delivering it to the user. Non-transient computer-readable recording medium.

9. In Paragraph 8, Each of the above multiple data centers is assigned a unique ID, and In each of the aforementioned multiple data centers, The use of computing resources is approved only if the data center ID sent by the data center operator's control device to each of the aforementioned multiple data centers matches the ID assigned to the corresponding data center. Non-transient computer-readable recording medium.

10. In Paragraph 8, The information regarding the above available computing resources includes, Information regarding the status of the corresponding data center is additionally included in the 1-1 state where additional resources can be allocated, the 1-2 state where it is normal but full resources are in use, the 1-3 state where it is serviceable but under maintenance, and the 1-4 state where it is serviceless. Non-transient computer-readable recording medium.

11. In Paragraph 8, The above method is, The step of requesting additional computing resources other than the identified available computing resources from each of the selected data centers further Non-transient computer-readable recording medium.

12. In Paragraph 11, Among the selected data centers mentioned above, there are data centers that cannot allocate the requested additional computing resources, and The above method is, The method further includes the step of requesting the additional computing resources from a data center other than the selected data center among the plurality of data centers. Non-transient computer-readable recording medium.

13. In Paragraph 8, The information regarding the above-mentioned available computing resources includes (i) the type of computing device, the number of computing device cycles, and the storage space capacity, or (ii) the computing performance and storage space capacity of the computing device. Non-transient computer-readable recording medium.

14. In Paragraph 8, The above method is, The step of sending a request message for control of computing resources to at least some of the selected data centers; and The method further includes the step of verifying information regarding available computing resources for each data center that received the above adjustment request message. Non-transient computer-readable recording medium.

15. Arithmetic unit; and As a control device for a data center including a communication device, The above computing device is, Select the data center that has approved the use of internal computing resources among multiple data centers, and It is confirmed that information regarding available computing resources for each of the above-mentioned selected data centers has been received through the communication device, and It confirms that a request for a specific operation from a specific user has been received through the communication device, Based on the above-mentioned confirmed information, the above-mentioned work operation is requested to at least some of the above-mentioned selected data centers through the above-mentioned communication device, and The result of the above operation is received from at least some of the data centers via the communication device and transmitted to the user via the communication device. Data center control unit.

Images