Cross-region computing power cooperation system and method

Abstract

The invention relates to a cross-region computing power cooperation system and method. The method comprises the following steps: acquiring computing power resource information of computing power cluster nodes of each region in a global range; obtaining a cluster pairing strategy according to the computing power resource information of the computing power cluster nodes of each region in the global range, and obtaining a logical computing power cluster group according to the cluster pairing strategy; computing power cluster nodes in the computing power cluster group are used for mutual cooperative computing; acquiring a task processing request, wherein information in the task processing request comprises a service attribute, a computing power demand and a data volume; and according to the task processing request and the computing power resource information of the computing power cluster nodes in the computing power cluster group, determining a target computing power cluster node matched with the service attribute, the computing power demand and the data volume from the computing power cluster group, and sending the task processing request to the target computing power cluster node. The method can improve the utilization rate of global computing power resources.

Description

Technical Field

[0001] The present application relates to the field of computing power collaboration technology, and in particular to a cross-regional computing power collaboration system and method. Background Art

[0002] With the development of technologies such as cloud computing, big data, and artificial intelligence, the demand for computing power continues to grow and is becoming more diversified and distributed. Traditional computing power is usually limited to a single data center or region, making it difficult to efficiently utilize computing resources around the world, resulting in low resource utilization. Summary of the invention

[0003] Based on this, it is necessary to provide a cross-regional computing power collaboration system and method that can improve resource utilization in response to the above technical problems.

[0004] In a first aspect, the present application provides a cross-regional computing power collaboration system, including a global resource discovery module, a dynamic cluster pairing module, and a distributed scheduler module, wherein:

[0005] The global resource discovery module is used to obtain computing resource information of computing cluster nodes in various regions within the global scope;

[0006] The dynamic cluster pairing module is used to obtain a cluster pairing strategy based on the computing power resource information of the computing power cluster nodes in various regions within the global scope, and obtain a logical computing power cluster group according to the cluster pairing strategy; the computing power cluster nodes in the computing power cluster group are used for mutual collaborative computing;

[0007] The distributed scheduler module is used to obtain a task processing request, the information in the task processing request includes business attributes, computing power requirements and data volume; and according to the task processing request and the computing power resource information of the computing power cluster nodes in the computing power cluster group, determine a target computing power cluster node that matches the business attributes, the computing power requirements and the data volume from the computing power cluster group, and send the task processing request to the target computing power cluster node.

[0008] In one of the embodiments, the computing power resource information of the computing power cluster node includes resource status, network delay, bandwidth information and service label of the computing power cluster node;

[0009] The dynamic cluster pairing module is also used to obtain a cluster pairing strategy based on the resource status, the network delay, the bandwidth information and the service label, pair the computing power cluster nodes in multiple regions according to the cluster pairing strategy to obtain a logical computing power cluster group, and assign the pairing label to the corresponding computing power cluster node in the computing power cluster group.

[0010] In one of the embodiments, the dynamic cluster pairing module is also used to obtain the overall operating status of the computing power cluster nodes in the computing power cluster group, and update the cluster pairing strategy according to the overall operating status, and modify the pairing labels of the corresponding computing power cluster nodes in the computing power cluster group according to the updated cluster pairing strategy.

[0011] In one of the embodiments, the distributed scheduler module is also used to obtain the task processing result corresponding to the task processing request fed back by the target computing power cluster node. When the task processing result indicates that the task processing request has not been executed, the task processing request is sent to the computing power cluster node in the computing power cluster group that is closest to the target computing power cluster node.

[0012] In one of the embodiments, the global resource discovery module includes a distributed database and a metadata management system; the global resource discovery module obtains computing power resource information of computing power cluster nodes in various regions in the global scope through an API interface or a message queue.

[0013] In one of the embodiments, the distributed scheduler module interacts with the computing power cluster nodes through an API interface.

[0014] In a second aspect, the present application provides a cross-regional computing power collaboration method, including:

[0015] Obtain computing power resource information of computing power cluster nodes in various regions globally;

[0016] A cluster pairing strategy is obtained according to the computing power resource information of the computing power cluster nodes in various regions within the global scope, and a logical computing power cluster group is obtained according to the cluster pairing strategy; the computing power cluster nodes in the computing power cluster group are used for mutual collaborative computing;

[0017] Obtain a task processing request, wherein the information in the task processing request includes business attributes, computing power requirements, and data volume; determine a target computing power cluster node that matches the business attributes, the computing power requirements, and the data volume from the computing power cluster group based on the task processing request and the computing power resource information of the computing power cluster nodes in the computing power cluster group, and send the task processing request to the target computing power cluster node.

[0018] In one embodiment, the computing power resource information includes resource status, network delay, bandwidth information and service tags of computing power cluster nodes in various regions;

[0019] The step of obtaining a cluster pairing strategy based on the computing power resource information of the computing power cluster nodes in various regions within the global scope, and obtaining a logical computing power cluster group based on the cluster pairing strategy includes:

[0020] A cluster pairing strategy is obtained according to the resource status, the network delay, the bandwidth information and the service label, and computing power cluster nodes in multiple regions are paired and marked according to the cluster pairing strategy to obtain a logical computing power cluster group, and the pairing label is assigned to the corresponding computing power cluster node in the computing power cluster group.

[0021] In one of the embodiments, the method further includes: obtaining the overall operating status of the computing power cluster nodes in the computing power cluster group, updating the cluster pairing strategy according to the overall operating status, and modifying the pairing labels of the corresponding computing power cluster nodes in the computing power cluster group according to the updated cluster pairing strategy.

[0022] In one of the embodiments, the method further includes: obtaining a task processing result corresponding to the task processing request fed back by the target computing power cluster node, and when the task processing result indicates that the task processing request has not been executed, sending the task processing request to the computing power cluster node in the computing power cluster group that is closest to the target computing power cluster node.

[0023] The above-mentioned cross-regional computing power collaboration system and method, the aforementioned system includes a global resource discovery module, a dynamic cluster pairing module and a distributed scheduler module, wherein: the global resource discovery module is used to obtain the computing power resource information of the computing power cluster nodes in various regions within the global scope; the dynamic cluster pairing module is used to obtain the cluster pairing strategy according to the computing power resource information of the computing power cluster nodes in various regions within the global scope, and obtain the logical computing power cluster group according to the cluster pairing strategy; the computing power cluster nodes in the computing power cluster group are used for mutual collaborative computing; the distributed scheduler module is used to obtain the task processing request, and the information in the task processing request includes business attributes, computing power requirements and data volume; and according to the task processing request and the computing power resource information of the computing power cluster nodes in the computing power cluster group, the target computing power cluster node matching the business attributes, computing power requirements and data volume is determined from the computing power cluster group, and the task processing request is sent to the target computing power cluster node. In the embodiment of the present application, the system performs resource discovery in real time through the global resource discovery module, and reasonably performs cluster pairing through the dynamic cluster pairing module, which reduces the idle computing power resources in various regions within the global scope, so that the computing power cluster nodes can efficiently participate in computing tasks. The cross-regional computing power coordination system in this embodiment not only solves the problem of insufficient computing power resources in a single region, but also improves the comprehensive utilization rate of global computing power resources through intelligent resource scheduling and management, and promotes efficient coordination of computing power. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0025] Figure 1 A structural block diagram of a cross-regional computing power collaboration system in one embodiment;

[0026] Figure 2 A schematic diagram of a process of a cross-regional computing power coordination method in one embodiment;

[0027] Figure 3 This is an application environment diagram of a cross-regional computing power collaboration system in another embodiment;

[0028] Figure 4 A structural block diagram of a cross-regional computing power coordination device in one embodiment;

[0029] Figure 5 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0031] In an exemplary embodiment, Figure 1 As shown, a cross-regional computing power collaboration system is provided, and the system includes a global resource discovery module 102, a dynamic cluster pairing module 104 and a distributed scheduler module 106.

[0032] Specifically, the system in this embodiment can be applied to servers in a computing power network, which is an information infrastructure composed of network, computing power, storage and other technologies; wherein a computing power cluster is a computing power collection built by multiple computing power devices; a computing power cluster group formed by pairing computing power clusters is a collection of computing power clusters that are logically classified by logical labels based on Cluster Pair technology.

[0033] The global resource discovery module 102 is used to obtain computing resource information of computing cluster nodes in various regions within the global scope. The computing cluster nodes in various regions within the global scope may refer to a collection of computing resources in data centers or cloud platforms distributed in different geographical locations. These nodes may be servers, virtual machines or containers for performing computing tasks.

[0034] Among them, the global resource discovery module can be responsible for collecting and managing resource information from computing clusters in different regions, including computing power (such as CPU, GPU), memory, storage, network bandwidth, etc.; the global resource discovery module can provide a real-time updated resource list to help the system understand the current status and available resources of each cluster.

[0035] Exemplarily, the global resource discovery module can access computing cluster nodes in different regions to obtain computing resource information of each computing cluster node.

[0036] The dynamic cluster pairing module 104 is used to obtain a cluster pairing strategy based on the computing power resource information of the computing power cluster nodes in various regions in the global scope, and obtain a logical computing power cluster group based on the cluster pairing strategy. The computing power cluster nodes in the computing power cluster group are used for collaborative computing.

[0037] The dynamic cluster pairing module can formulate cluster pairing strategies based on global resource information to form logical computing cluster groups; it can also decide how to connect and combine multiple computing clusters based on resource availability and task requirements to achieve the best computing synergy. Logical computing cluster groups can refer to virtual clusters formed by cluster pairing, which can effectively utilize multiple physically dispersed (such as in different regions) computing resources to perform common tasks.

[0038] Exemplarily, the dynamic cluster pairing module analyzes the capabilities of each computing cluster node based on the computing resource information of computing cluster nodes in various regions globally, pairs the computing cluster nodes with sufficient computing resources, and thereby obtains a computing cluster group.

[0039] Among them, the distributed scheduler module 106 is used to obtain a task processing request, the information in the task processing request includes business attributes, computing power requirements and data volume; and according to the task processing request and the computing power resource information of the computing power cluster nodes in the computing power cluster group, determine the target computing power cluster node that matches the business attributes, computing power requirements and data volume from the computing power cluster group, and send the task processing request to the target computing power cluster node.

[0040] Among them, the distributed scheduler module can refer to the module responsible for receiving task requests and performing resource scheduling; the distributed scheduler module can analyze the required resources according to the characteristics of the task processing request (such as business attributes, computing power requirements, and data volume), select the most suitable cluster nodes from the formed computing power cluster group for task processing, and dispatch tasks to these nodes.

[0041] Exemplarily, the distributed scheduler module obtains a task processing request, and then analyzes the characteristics of the request based on the business attributes, computing power requirements, data volume and other information contained in the task processing request, and then determines the target computing power cluster node that can meet the task requirements from the computing power cluster group, and then sends the task processing request to the target computing power cluster node. If the target computing power cluster node is multiple computing power cluster nodes, the multiple computing power cluster nodes start collaborative computing according to the assigned tasks.

[0042] In this embodiment, the system performs real-time resource discovery through the global resource discovery module, and reasonably performs cluster pairing through the dynamic cluster pairing module, reducing the idle computing resources in various regions within the global scope, so that the computing cluster nodes can efficiently participate in computing tasks. The cross-regional computing power collaboration system in this embodiment not only solves the problem of insufficient computing power resources in a single region, but also improves the comprehensive utilization rate of global computing power resources through intelligent resource scheduling and management, and promotes efficient collaboration of computing power.

[0043] In an exemplary embodiment, the computing power resource information of the computing power cluster node includes the resource status, network delay, bandwidth information and service label of the computing power cluster node; the dynamic cluster pairing module is also used to obtain a cluster pairing strategy based on the resource status, network delay, bandwidth information and service label, and pair the computing power cluster nodes in multiple regions according to the cluster pairing strategy to obtain a logical computing power cluster group, and assign the pairing label to the corresponding computing power cluster node in the computing power cluster group.

[0044] Among them, the computing power resource information of the computing power cluster node includes the resource status, network delay, bandwidth information and service tags of the computing power cluster node.

[0045] Among them, resource status can refer to the current load situation of the node (such as CPU / memory usage); network latency can refer to the time required for data to be transmitted between nodes; bandwidth information can refer to the capacity of the network connection, usually expressed as the amount of data that can be transmitted per unit time; business labels can refer to labels assigned to nodes or their resources, indicating specific types of tasks or business areas that the nodes can optimize for processing; for example, some nodes may be specifically optimized for big data processing, while other nodes may be more suitable for machine learning tasks.

[0046] Exemplarily, the dynamic cluster pairing module analyzes the acquired computing power cluster node resource status, network delay, bandwidth information and service label to obtain the characteristics of each computing power cluster node, and can pair computing power cluster nodes with similar characteristics; the computing power cluster nodes can also be classified according to the characteristics, and then some computing power cluster nodes are selected from each type of computing power cluster nodes for pairing, so that the computing power cluster group obtained can contain computing power resources with comprehensive characteristics; then, according to the obtained cluster pairing strategy, the computing power cluster nodes in multiple regions are paired and marked to obtain a logical computing power cluster group, and the pairing label is assigned to the corresponding computing power cluster node in the computing power cluster group. After obtaining the pairing label, the computing power cluster node can search for other computing power cluster nodes in the same computing power cluster group through the above-mentioned cross-regional computing power collaboration system.

[0047] In this embodiment, dynamic pairing is performed between computing cluster nodes in multiple regions, and resource characteristics and advantages of different regions can be utilized to achieve more efficient collaborative computing.

[0048] In an exemplary embodiment, the dynamic cluster pairing module is also used to obtain the overall operating status of the computing power cluster nodes in the computing power cluster group, and update the cluster pairing strategy according to the overall operating status, and modify the pairing labels of the corresponding computing power cluster nodes in the computing power cluster group according to the updated cluster pairing strategy.

[0049] Among them, the overall operating status may refer to the overall performance and health of the computing cluster nodes, which may include status information such as CPU usage, memory consumption, network latency, fault status, and task load.

[0050] Exemplarily, the dynamic cluster pairing module can periodically collect overall operating status information from each node in the computing power cluster group. The dynamic cluster pairing module updates the cluster pairing strategy based on the overall operating status. For example, cluster B fails and needs to be removed from the current pairing strategy. Clusters A and C are in good condition, but the CPU load is high, and task load balancing may be required. Then, the pairing labels of the corresponding computing power cluster nodes in the computing power cluster group are modified according to the updated cluster pairing strategy.

[0051] In this embodiment, the dynamic cluster pairing module continuously monitors and adjusts the operating status in the computing power cluster group, and updates the cluster pairing strategy and labels in real time, thereby achieving more flexible and efficient resource coordination and use.

[0052] In an exemplary embodiment, the distributed scheduler module is also used to obtain the task processing result corresponding to the task processing request fed back by the target computing power cluster node. If the task processing result indicates that the task processing request has not been executed, the task processing request is sent to the computing power cluster node in the computing power cluster group that is closest to the target computing power cluster node.

[0053] The task processing result may refer to the execution result returned by the target computing cluster node after receiving the task processing request. The computing cluster node closest to the target computing cluster node may refer to the node with the smallest physical or logical distance to the target computing cluster node in the network topology corresponding to the computing cluster group.

[0054] In this embodiment, when the target node cannot complete the task processing request, the distributed scheduler module sends the task processing request to the computing cluster node in the computing cluster group that is closest to the target computing cluster node. When the task processing fails, the task can be flexibly adjusted to migrate it to other computing cluster nodes; by migrating to the closest computing cluster node, the delay of the migration process can be reduced and the efficiency of task migration can be improved.

[0055] In an exemplary embodiment, the global resource discovery module includes a distributed database and a metadata management system; the global resource discovery module obtains computing power resource information of computing power cluster nodes in various regions in the global scope through an API interface or a message queue.

[0056] In this embodiment, the global resource discovery module realizes the rapid acquisition and effective management of resource information of computing power cluster nodes in various regions by combining a distributed database with a metadata management system and utilizing an API interface or a message queue.

[0057] In an exemplary embodiment, the distributed scheduler module interacts with the computing power cluster nodes through an API interface.

[0058] In this embodiment, using the API interface, the distributed scheduler module can quickly submit tasks, request resource status and obtain processing results, thereby realizing instant task scheduling, reducing the delay in task execution and improving the response rate of the system.

[0059] In an exemplary embodiment, Figure 2 As shown, a cross-regional computing power collaboration method is provided. This embodiment uses the method applied to a server as an example for illustration. It can be understood that the method can also be applied to a terminal, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0060] Step S202, obtaining computing power resource information of computing power cluster nodes in various regions within the global scope.

[0061] Among them, the computing power cluster nodes in various regions globally can refer to a collection of computing resources distributed in data centers or cloud platforms in different geographical locations. These nodes can be servers, virtual machines or containers used to perform computing tasks.

[0062] The computing resource information may refer to detailed information describing the computing resources available to the computing cluster nodes.

[0063] Exemplarily, the server accesses computing power cluster nodes in different regions, obtains computing power resource information of each computing power cluster node, and obtains computing power resources of computing power cluster nodes in various regions globally.

[0064] Step S204: obtain a cluster pairing strategy based on the computing power resource information of the computing power cluster nodes in various regions in the global scope, and obtain a logical computing power cluster group based on the cluster pairing strategy, wherein the computing power cluster nodes in the computing power cluster group are used for mutual collaborative computing.

[0065] Among them, the cluster pairing strategy can be used to determine how to combine or pair cluster nodes with different computing power to achieve collaborative computing.

[0066] Among them, the logical computing power cluster group can refer to a virtual cluster formed by cluster pairing, which can effectively utilize multiple physically dispersed (such as in different regions) computing resources to perform common tasks.

[0067] Exemplarily, the server analyzes the capabilities of each computing cluster node based on the computing resource information of computing cluster nodes in various regions globally, pairs the computing cluster nodes that currently have sufficient computing resources, and thereby obtains a computing cluster group.

[0068] Step S206, obtain the task processing request, determine the target computing cluster node that matches the business attributes, computing power requirements and data volume from the computing power cluster group based on the task processing request and the computing power resource information of the computing power cluster nodes in the computing power cluster group, and send the task processing request to the target computing power cluster node.

[0069] The information in the task processing request includes business attributes, computing power requirements, and data volume.

[0070] Exemplarily, the server obtains a task processing request sent by a business system or other terminal, and then analyzes the characteristics of the request based on the business attributes, computing power requirements, data volume and other information contained in the task processing request, and then determines the target computing power cluster node that can meet the task requirements from the computing power cluster group, and then sends the task processing request to the target computing power cluster node. If the target computing power cluster node is multiple computing power cluster nodes, the multiple computing power cluster nodes start collaborative computing according to the assigned tasks.

[0071] In the above cross-regional computing power coordination method, through real-time resource discovery and reasonable cluster pairing, the idle computing power resources in various regions on a global scale are reduced, so that the computing power cluster nodes can efficiently participate in computing tasks. It not only solves the problem of insufficient computing power resources in a single region, but also improves the comprehensive utilization rate of global computing power resources through intelligent resource scheduling and management, and promotes efficient coordination of computing power.

[0072] In an exemplary embodiment, the computing power resource information includes the resource status, network delay, bandwidth information and service tags of the computing power cluster nodes in each region; the above step S204 may specifically include: obtaining a cluster pairing strategy based on the resource status, network delay, bandwidth information and service tags, pairing and marking the computing power cluster nodes in multiple regions according to the cluster pairing strategy to obtain a logical computing power cluster group, and assigning the pairing tags to the corresponding computing power cluster nodes in the computing power cluster group.

[0073] Among them, the computing power resource information of the computing power cluster node includes the resource status, network delay, bandwidth information and service tags of the computing power cluster node.

[0074] Among them, resource status can refer to the current load situation of the node (such as CPU / memory usage); network latency can refer to the time required for data to be transmitted between nodes; bandwidth information can refer to the capacity of the network connection, usually expressed as the amount of data that can be transmitted per unit time; business labels can refer to labels assigned to nodes or their resources, indicating specific types of tasks or business areas that the nodes can optimize for processing; for example, some nodes may be specifically optimized for big data processing, while other nodes may be more suitable for machine learning tasks.

[0075] Exemplarily, the server analyzes the acquired computing power cluster node resource status, network delay, bandwidth information and service labels to obtain the characteristics of each computing power cluster node, and can pair computing power cluster nodes with similar characteristics; the computing power cluster nodes can also be classified according to the characteristics, and then some computing power cluster nodes are selected from each type of computing power cluster nodes for pairing, so that the computing power cluster group obtained can contain computing power resources with comprehensive characteristics; then, according to the obtained cluster pairing strategy, the computing power cluster nodes in multiple regions are paired and marked to obtain a logical computing power cluster group, and the pairing labels are assigned to the corresponding computing power cluster nodes in the computing power cluster group. After obtaining the pairing labels, the computing power cluster nodes can search for other computing power cluster nodes in the same computing power cluster group through the above-mentioned cross-regional computing power collaboration system.

[0076] In this embodiment, dynamic pairing is performed between computing cluster nodes in multiple regions, and resource characteristics and advantages of different regions can be utilized to achieve more efficient collaborative computing.

[0077] In an exemplary embodiment, the above-mentioned cross-regional computing power coordination method may also include: obtaining the overall operating status of the computing power cluster nodes in the computing power cluster group, and updating the cluster pairing strategy according to the overall operating status, and modifying the pairing labels of the corresponding computing power cluster nodes in the computing power cluster group according to the updated cluster pairing strategy.

[0078] Among them, the overall operating status may refer to the overall performance and health of the computing cluster nodes, which may include status information such as CPU usage, memory consumption, network latency, fault status, and task load.

[0079] Exemplarily, the server can periodically collect overall operating status information from each node in the computing power cluster group, and the dynamic cluster pairing module updates the cluster pairing strategy according to the overall operating status. For example, cluster B fails and needs to be removed from the current pairing strategy. Clusters A and C are in good condition, but the CPU load is high, and task load balancing may be required. Then, the pairing labels of the corresponding computing power cluster nodes in the computing power cluster group are modified according to the updated cluster pairing strategy.

[0080] In this embodiment, the dynamic cluster pairing module continuously monitors and adjusts the operating status in the computing power cluster group, and updates the cluster pairing strategy and labels in real time, thereby achieving more flexible and efficient resource coordination and use.

[0081] In an exemplary embodiment, the above-mentioned cross-regional computing power collaboration method may also include: obtaining the task processing result corresponding to the task processing request fed back by the target computing power cluster node, and when the task processing result indicates that the task processing request has not been executed, sending the task processing request to the computing power cluster node in the computing power cluster group that is closest to the target computing power cluster node.

[0082] The task processing result may refer to the execution result returned by the target computing cluster node after receiving the task processing request. The computing cluster node closest to the target computing cluster node may refer to the node with the smallest physical or logical distance to the target computing cluster node in the network topology corresponding to the computing cluster group.

[0083] In this embodiment, when the target node cannot complete the task processing request, the distributed scheduler module sends the task processing request to the computing cluster node in the computing cluster group that is closest to the target computing cluster node. When the task processing fails, the task can be flexibly adjusted to migrate it to other computing cluster nodes; by migrating to the closest computing cluster node, the delay of the migration process can be reduced and the efficiency of task migration can be improved.

[0084] With the development of technologies such as cloud computing, big data, and artificial intelligence, the demand for computing power continues to grow, and it is characterized by diversification and distribution. Traditional computing power is usually limited to a single data center or region, making it difficult to efficiently utilize computing resources around the world, resulting in low resource utilization and failure to meet the requirements of real-time and load balancing. Cross-regional computing power collaboration integrates computing resources in multiple geographical locations to achieve dynamic allocation of tasks and resource sharing. The development of related technologies covers distributed scheduling algorithms, edge computing, 5G networks, and intelligent resource management, with the goal of optimizing network latency, improving bandwidth utilization, and ensuring data security and task consistency. In recent years, with the rise of edge computing, the potential of cross-regional computing power scheduling has further emerged, becoming a key technology for improving computing power resource utilization efficiency, reducing computing costs, and supporting real-time applications. The cross-regional computing power collaboration technology proposed in this application belongs to the field of cloud computing and distributed computing, and aims to solve the problem of uneven distribution of computing power resources. Current edge computing mainly focuses on localized processing and lacks a cross-regional scheduling mechanism. Existing distributed task scheduling systems are usually used for resource management between nodes in the same cluster or local area network, and do not deeply solve cross-regional computing power collaboration and computing power fluctuations.

[0085] In order to solve the above problems, in an application example, this application also provides a cross-regional computing power collaboration system, which consists of the following modules: global resource discovery module, distributed scheduler module, dynamic cluster pairing (ClusterPair) module and control center module. These modules work together to achieve cross-regional computing power collaboration. Among them:

[0086] The global resource discovery module is responsible for collecting and maintaining global computing resource information, including resource status, network latency, bandwidth, business tags, etc. of computing clusters (computing cluster nodes) in various regions. This module provides a global view of computing resources and provides data support for computing collaboration. The global resource discovery module can be composed of a distributed database and metadata management system, running on each computing cluster node and the control center server. The data of the global resource discovery module can be derived from the resource monitor of each computing node, and collected to the resource directory module through the API interface or message queue for unified storage and update.

[0087] The distributed scheduler module is used to select the optimal computing node to allocate tasks based on the computing resource information, task requirements and network status in the global resource discovery module. The distributed scheduler module can run on the control center server and interact with the computing cluster nodes through the API interface. The distributed scheduler can obtain the current computing resource information from the resource discovery module, receive task requirements from the business system (requirements represented by task processing requests), generate task allocation plans, and send scheduling instructions to the corresponding computing nodes.

[0088] The dynamic cluster pairing module is used to dynamically pair computing clusters in multiple regions and establish a logical "pairing" relationship based on factors such as network latency, bandwidth, and computing load, so that computing clusters marked as the same pair can collaborate with each other. The dynamic cluster pairing module can be managed by the control center, which is responsible for dynamically updating cluster labels and pairing relationships. The dynamic cluster pairing module can generate a pairing strategy based on the status information of each cluster provided by the resource discovery module. According to the pairing strategy, the computing cluster node group can be obtained, and the pairing label can be assigned to each computing cluster; when the scheduling strategy needs to be adjusted, the cluster pairing target is updated by modifying the label.

[0089] The control center module is used to perform global scheduling control and policy adjustment, monitor the overall operating status of computing resources, and update scheduling policies based on system load, business needs, and policies. The control center module can serve as a master node to centrally manage global computing scheduling tasks. The control center module can receive data from all modules, including resource lists, pairing information, and scheduling status, and coordinate the work of each module. Among them, the dynamic cluster pairing module can be included in the control center module.

[0090] In this embodiment, the data flow and execution sequence of the cross-regional computing power coordination system can be combined Figure 3 The specific steps may include:

[0091] Step S1, resource status collection. Each computing cluster node reports the resource status to the global resource discovery module in real time through the resource monitor, including CPU, memory, network delay, bandwidth utilization and other information.

[0092] Step S2: Dynamic pairing and label update: The dynamic cluster pairing module of the control center generates or updates the pairing relationship of the clusters according to the global resource information and the current network status, and marks the logical cluster pairs.

[0093] Step S3, receiving a task scheduling request. The distributed scheduling module receives a task scheduling request (task processing request) sent by the business system. The task scheduling request includes information such as the business attributes, computing power requirements, and data volume of the task.

[0094] Step S4, task allocation and execution. The distributed scheduling module obtains resource status and pairing strategy from the global resource directory and pairing information, and selects the best computing node from the computing power cluster group to execute the task. After the task allocation is completed, the scheduler sends the task execution instruction to the corresponding computing power node.

[0095] Step S5: Task result transmission and feedback: After the corresponding computing node completes the task, the computing result is transmitted back to the business system by the computing power cluster node through the control center, and the control center updates the resource status and task log.

[0096] In this embodiment, cross-regional computing power collaborative scheduling is realized. By integrating computing cluster resources in multiple regions, tasks are dynamically assigned to the optimal regional node for execution, thereby breaking through the limitations of existing technologies in scheduling within a single data center or local area network, and improving the utilization rate of global computing power. This solution introduces the dynamic labeling mechanism of cluster pairing for the first time. It can dynamically pair and mark computing power clusters in multiple regions through different factors (such as network delay, cluster business labels, etc.). The marked computing power clusters are logically identified as a whole. The computing power resources within the whole can cooperate with each other and are not affected by the global scheduling strategy. When the control center is performing global computing power coordination, it only needs to change the cluster pairing target by modifying the label of the computing power cluster. This dynamic cluster pairing technology not only retains the flexibility of computing power scheduling in the entire region, but also reduces the impact of global scheduling on the fluctuation of computing power within the paired cluster, thereby improving the stability of the operation of the global computing power cluster. After sending the task processing request to the target computing cluster node, if the target computing cluster node is unable to continue processing the task processing request, it can return the processing result representing the incomplete request to the control center, and then the control center can send the task processing request to the computing cluster node closest to the target computing cluster node in the computing cluster group; if the target computing cluster node is unable to continue processing the task processing request, the target computing cluster node can also use the pairing label to find the computing cluster node closest to itself in the computing cluster group, and then migrate the task processing request to the nearest computing cluster node. Through the above-mentioned task migration method, that is, through the dynamic Cluster Pair technology, the global computing power fluctuations that may occur during global computing power coordination are solved.

[0097] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0098] Based on the same inventive concept, the embodiment of the present application also provides a cross-regional computing power collaboration device for implementing the cross-regional computing power collaboration method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in the one or more cross-regional computing power collaboration device embodiments provided below can refer to the above limitations on the cross-regional computing power collaboration method, which will not be repeated here.

[0099] In an exemplary embodiment, Figure 4 As shown, a cross-region computing power coordination device 900 is provided, including: a resource information acquisition module 901, a cluster pairing module 902 and a target node determination module 903, wherein:

[0100] Resource information acquisition module 901, used to obtain computing resource information of computing cluster nodes in various regions in the global scope;

[0101] The cluster pairing module 902 is used to obtain a cluster pairing strategy based on the computing power resource information of the computing power cluster nodes in various regions in the global scope, and obtain a logical computing power cluster group according to the cluster pairing strategy; the computing power cluster nodes in the computing power cluster group are used for mutual collaborative computing;

[0102] The target node determination module 903 is used to obtain a task processing request, the information in the task processing request includes business attributes, computing power requirements and data volume; based on the task processing request and the computing power resource information of the computing power cluster nodes in the computing power cluster group, a target computing power cluster node that matches the business attributes, computing power requirements and data volume is determined from the computing power cluster group, and the task processing request is sent to the target computing power cluster node.

[0103] In an exemplary embodiment, the computing power resource information includes the resource status, network delay, bandwidth information and service tags of the computing power cluster nodes in each region; the above-mentioned cluster pairing module 902 is also used to obtain a cluster pairing strategy based on the resource status, network delay, bandwidth information and service tags, and pair the computing power cluster nodes in multiple regions according to the cluster pairing strategy to obtain a logical computing power cluster group, and assign the pairing tag to the corresponding computing power cluster node in the computing power cluster group.

[0104] In an exemplary embodiment, the above-mentioned cross-regional computing power collaboration device also includes a cluster group update module, which is used to obtain the overall operating status of the computing power cluster nodes in the computing power cluster group, and update the cluster pairing strategy according to the overall operating status, and modify the pairing labels of the corresponding computing power cluster nodes in the computing power cluster group according to the updated cluster pairing strategy.

[0105] In an exemplary embodiment, the above-mentioned cross-regional computing power collaboration device also includes a task migration module, which is used to obtain the task processing result corresponding to the task processing request fed back by the target computing power cluster node. When the task processing result indicates that the task processing request has not been executed, the task processing request is sent to the computing power cluster node in the computing power cluster group that is closest to the target computing power cluster node.

[0106] Each module in the above cross-regional computing power collaboration device can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each module above.

[0107] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Figure 5 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a cross-regional computing power collaboration method is implemented.

[0108] Those skilled in the art will understand that Figure 5 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0109] In one embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above method embodiments when executing the computer program.

[0110] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0111] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0112] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0113] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0114] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A cross-regional computing power collaboration system, characterized in that: The system includes a global resource discovery module, a dynamic cluster pairing module and a distributed scheduler module, wherein: The global resource discovery module is used to obtain computing resource information of computing cluster nodes in various regions within the global scope; The dynamic cluster pairing module is used to obtain a cluster pairing strategy based on the computing power resource information of the computing power cluster nodes in various regions within the global scope, and obtain a logical computing power cluster group according to the cluster pairing strategy; the computing power cluster nodes in the computing power cluster group are used for mutual collaborative computing; The distributed scheduler module is used to obtain a task processing request, the information in the task processing request includes business attributes, computing power requirements and data volume; and according to the task processing request and the computing power resource information of the computing power cluster nodes in the computing power cluster group, determine a target computing power cluster node that matches the business attributes, the computing power requirements and the data volume from the computing power cluster group, and send the task processing request to the target computing power cluster node.

2. The system according to claim 1, characterized in that The computing power resource information of the computing power cluster node includes the resource status, network delay, bandwidth information and service label of the computing power cluster node; The dynamic cluster pairing module is also used to obtain a cluster pairing strategy based on the resource status, the network delay, the bandwidth information and the service label, pair the computing power cluster nodes in multiple regions according to the cluster pairing strategy to obtain a logical computing power cluster group, and assign the pairing label to the corresponding computing power cluster node in the computing power cluster group.

3. The system according to claim 2, characterized in that The dynamic cluster pairing module is also used to obtain the overall operating status of the computing power cluster nodes in the computing power cluster group, and update the cluster pairing strategy according to the overall operating status, and modify the pairing labels of the corresponding computing power cluster nodes in the computing power cluster group according to the updated cluster pairing strategy.

4. The system according to claim 1, characterized in that The distributed scheduler module is also used to obtain the task processing result corresponding to the task processing request fed back by the target computing power cluster node. If the task processing result indicates that the task processing request has not been executed, the task processing request is sent to the computing power cluster node in the computing power cluster group that is closest to the target computing power cluster node.

5. The system according to claim 1, characterized in that The global resource discovery module includes a distributed database and a metadata management system; the global resource discovery module obtains computing power resource information of computing power cluster nodes in various regions in the global scope through an API interface or a message queue.

6. The system according to claim 1, characterized in that The distributed scheduler module interacts with the computing power cluster nodes through an API interface.

7. A cross-regional computing power coordination method, characterized in that: The method comprises: Obtain computing power resource information of computing power cluster nodes in various regions globally; A cluster pairing strategy is obtained according to the computing power resource information of the computing power cluster nodes in various regions within the global scope, and a logical computing power cluster group is obtained according to the cluster pairing strategy; the computing power cluster nodes in the computing power cluster group are used for mutual collaborative computing; Obtain a task processing request, wherein the information in the task processing request includes business attributes, computing power requirements, and data volume; determine a target computing power cluster node that matches the business attributes, the computing power requirements, and the data volume from the computing power cluster group based on the task processing request and the computing power resource information of the computing power cluster nodes in the computing power cluster group, and send the task processing request to the target computing power cluster node.

8. The method according to claim 7, characterized in that The computing power resource information includes resource status, network delay, bandwidth information and service tags of computing power cluster nodes in various regions; The step of obtaining a cluster pairing strategy based on the computing power resource information of the computing power cluster nodes in various regions within the global scope, and obtaining a logical computing power cluster group based on the cluster pairing strategy includes: A cluster pairing strategy is obtained according to the resource status, the network delay, the bandwidth information and the service label, and computing power cluster nodes in multiple regions are paired and marked according to the cluster pairing strategy to obtain a logical computing power cluster group, and the pairing label is assigned to the corresponding computing power cluster node in the computing power cluster group.

9. The method according to claim 7, characterized in that: The method further comprises: The overall operation status of the computing power cluster nodes in the computing power cluster group is obtained, and the cluster pairing strategy is updated according to the overall operation status, and the pairing labels of the corresponding computing power cluster nodes in the computing power cluster group are modified according to the updated cluster pairing strategy.

10. The method according to claim 7, characterized in that The method further comprises: Obtain a task processing result corresponding to the task processing request fed back by the target computing power cluster node. If the task processing result indicates that the task processing request has not been completed, send the task processing request to the computing power cluster node in the computing power cluster group that is closest to the target computing power cluster node.

Images