Resource scheduling method, system, device and equipment and storage medium

Abstract

The invention provides a resource scheduling method, system, device and equipment and a storage medium, and the method comprises the steps: receiving region identification information sent by an edge node, the region identification information being used for identifying a region resource pool to which a resource instance providing a service requested by a task request belongs; the regional resource pool comprises token buckets of resource instances providing services in the corresponding geographic regions; acquiring an idle token from a token bucket of a resource instance in a region resource pool to which the resource instance belongs based on the region identification information; and sending the token to the edge node, so that the edge node sends the task request to the resource instance indicated by the token for processing. According to the application, the edge node locates the regional resource pool corresponding to the requested service based on the service identifier carried by the task request and the locally stored resource partition configuration information, and directly requests the resource management center to inquire the idle token in the regional resource pool, thereby effectively shortening the time spent on obtaining the token, shortening the response time, and improving the efficiency. The treatment efficiency is improved.

Description

Technical Field

[0001] The present application belongs to the field of network technology, and specifically relates to a resource scheduling method, system, device, equipment and storage medium. Background Art

[0002] Resources such as models or services can be deployed in a distributed cloud-edge deployment mode. In the distributed network, there are multiple processing nodes and multiple edge nodes. Resource instances are deployed on the processing nodes, and task queues corresponding to each processing node are maintained on the edge nodes. After receiving a task request from the client, the edge node inserts the task request into the tail of a task queue. When the resource instance in the processing node corresponding to the task queue is idle, a task request is dequeued from the head of the task queue and sent to the processing node corresponding to the task queue for processing.

[0003] Since the client's task request needs to be queued at the edge node, the task request needs to wait for a long time before being processed, the processing efficiency of the task request is low, and the user waits for a long time. Summary of the invention

[0004] The present application proposes a resource scheduling method, system, apparatus, device and storage medium. It is not necessary to set up a queue corresponding to a resource instance on an edge node. Instead, the edge node locates the regional resource pool to which the resource instance of the requested service belongs based on the service identifier carried in the task request and the locally stored resource partition configuration information, and directly requests the resource management center to query the regional resource pool for free tokens, thereby effectively shortening the time spent on obtaining tokens, shortening the response time, and improving processing efficiency.

[0005] The first aspect of the present application provides a resource scheduling method, which is applied to a resource management center. The method includes:

[0006] Receiving regional identification information sent by an edge node, the regional identification information is used to identify a regional resource pool to which a resource instance providing a service requested by the task request belongs; the regional resource pool includes a token bucket of resource instances providing the service in a corresponding geographical area;

[0007] Based on the area identification information, obtaining an idle token from a token bucket of a resource instance in a resource pool of the corresponding area;

[0008] The token is sent to the edge node, so that the edge node sends the task request to the resource instance indicated by the token for processing.

[0009] A second aspect of the present application provides a resource scheduling method, which is applied to an edge node. The method includes:

[0010] Receiving a task request from a client, wherein the task request includes a service identifier of a requested service;

[0011] Acquire, from the resource partition configuration information, the regional identification information of the regional resource pool to which the resource instance providing the service indicated by the service identifier belongs; the resource partition configuration information includes the correspondence between the instance information of the resource instance in the geographical area to which the edge node belongs and the regional identification information of the regional resource pool corresponding to the geographical area;

[0012] Sending the region identification information to a resource management center so that the resource management center allocates a token to the task request based on the region identification information;

[0013] The token is received from the resource management center, and the task request is sent to the resource instance indicated by the token.

[0014] A third aspect of the present application provides a resource scheduling system, the system comprising: an edge node and a resource management center;

[0015] The edge node is used to receive a task request from a client, wherein the task request includes a service identifier of a requested service; obtain, from resource partition configuration information, the regional identification information of a regional resource pool to which a resource instance providing the service indicated by the service identifier belongs; the regional identification information is used to identify the regional resource pool to which the resource instance providing the service requested by the task request belongs; the resource partition configuration information includes a correspondence between instance information of a resource instance in a geographic region to which the edge node belongs and regional identification information of a regional resource pool corresponding to the geographic region; send the regional identification information to the resource management center, receive a token allocated by the resource management center based on the regional identification information, and send the task request to the resource instance indicated by the token;

[0016] The resource management center is used to receive the area identification information sent by the edge node; based on the area identification information, obtain an idle token from the regional resource pool to which it belongs; and send the token to the edge node.

[0017] A fourth aspect of the present application provides a resource scheduling device, which is applied to a resource management center. The device includes:

[0018] A receiving module, configured to receive regional identification information sent by an edge node, wherein the regional identification information is used to identify a regional resource pool to which a resource instance providing the service requested by the task request belongs; the regional resource pool includes a token bucket of resource instances providing the service in a corresponding geographical area;

[0019] A token acquisition module, used to acquire an idle token from the regional resource pool to which the token belongs based on the regional identification information;

[0020] The sending module is used to send the token to the edge node, so that the edge node sends the task request to the resource instance indicated by the token for processing.

[0021] A fifth aspect of the present application provides a resource scheduling device, which is applied to an edge node. The device includes:

[0022] A receiving module, used for receiving a task request from a client, wherein the task request includes a service identifier of a requested service;

[0023] an acquisition module, configured to acquire, from resource partition configuration information, the regional identification information of the regional resource pool to which the resource instance providing the service indicated by the service identifier belongs; the resource partition configuration information includes a correspondence between instance information of the resource instance in the geographical area to which the edge node belongs and regional identification information of the regional resource pool corresponding to the geographical area;

[0024] A sending module, used for sending the area identification information to a resource management center, so that the resource management center allocates a token to the task request based on the area identification information;

[0025] The receiving module is further used to receive the token from the resource management center;

[0026] The sending module is further used to send the task request to the resource instance indicated by the token.

[0027] An embodiment of the sixth aspect of the present application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method described in the first or second aspect above.

[0028] The embodiment of the seventh aspect of the present application provides a computer-readable storage medium on which a computer program is stored, and the program is executed by a processor to implement the method described in the first or second aspect above.

[0029] The technical solution provided in the embodiments of the present application has at least the following technical effects or advantages:

[0030] In the embodiment of the present application, the edge node can locate the regional resource pool to which the resource instance of the requested service belongs based on the service identifier carried by the task request and the locally stored resource partition configuration information, and directly request the resource management center to query the idle tokens in the regional resource pool, effectively shortening the time spent on obtaining tokens, shortening the response time, and improving processing efficiency. In addition, the resource management center manages the tokens of the resource instance through the token bucket of the resource instance, and can limit the flow of task requests to access the resource instance, thereby protecting the resource instance.

[0031] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] By reading the detailed description of the preferred embodiment below, various other advantages and benefits will become clear to those of ordinary skill in the art. The accompanying drawings are only used for the purpose of illustrating the preferred embodiment and are not considered to be limitations of the present application. In addition, the same reference symbols are used to represent the same components throughout the accompanying drawings.

[0033] In the attached picture:

[0034] Figure 1 A signaling interaction diagram of a resource scheduling method provided by an embodiment of the present application is shown;

[0035] Figure 2 An operation flow chart of a resource management center in a resource scheduling method provided in an embodiment of the present application is shown;

[0036] Figure 3 An operation flow chart of an edge node in a resource scheduling method provided by an embodiment of the present application is shown;

[0037] Figure 4 A schematic diagram of a resource scheduling system provided by an embodiment of the present application is shown;

[0038] Figure 5 A structural block diagram of a resource scheduling device applied to a resource management center provided by an embodiment of the present application is shown;

[0039] Figure 6 A structural block diagram of a resource scheduling device applied to an edge node provided by an embodiment of the present application is shown;

[0040] Figure 7 A schematic diagram of the structure of an electronic device provided by an embodiment of the present application is shown;

[0041] Figure 8 A schematic diagram of a storage medium provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0042] The exemplary embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present application and to fully convey the scope of the present application to those skilled in the art.

[0043] It should be noted that, unless otherwise specified, the technical terms or scientific terms used in this application should have the common meanings understood by technicians in the field to which this application belongs.

[0044] CDN (Content Delivery Network) is a distributed network service. Its main purpose is to distribute static or dynamic resource content such as websites, applications, videos, audio, etc. to server nodes deployed globally, so that users can obtain content from the node closest to them, thereby improving access speed and stability, reducing network congestion and latency, and also reducing the pressure on the source station.

[0045] A large model refers to a machine learning model with a large number of parameters and complex structure. It can process massive data and complete various complex tasks, such as natural language processing, computer vision, and speech recognition.

[0046] In the related technology, the large model can be deployed in a distributed cloud-edge deployment manner, that is, the large model is deployed on a node in a distributed network, which includes multiple processing nodes and multiple edge nodes. The large model can be deployed on the processing node, and the edge node maintains a task queue corresponding to each processing node. After receiving the task request from the client, the edge node determines the processing node requested by the task request, inserts the task request into the tail of the task queue corresponding to the processing node, and when the large model in the processing node corresponding to the task queue is idle, a task request is dequeued from the head of the task queue, and the dequeued task request is sent to the large model in the processing node for processing.

[0047] In addition to large models, some algorithms, engines, software, etc. can be deployed on processing nodes in a distributed network. In the embodiments of the present application, these large models, algorithms, engines, software and other services deployed on processing nodes are collectively referred to as resource instances.

[0048] In the above-mentioned related technologies, since the task request of the client needs to be queued in the task queue in the edge node, the task request needs to wait for a long time before being processed, the processing efficiency of the task request is low, and the user waits for a long time.

[0049] Based on this, the embodiment of the present application provides a resource scheduling method, in which the resource management center maintains the token bucket of the resource instance. After receiving the task request, the edge node obtains the token of the service requested by the task request from the resource management center. The token is obtained from the regional resource pool to which the resource instance of the service requested by the task request belongs. The regional resource pool includes the token bucket of the resource instance providing the service in the corresponding geographical area. After obtaining the token, the edge node forwards the task request to the processing node where the corresponding resource instance is located.

[0050] The method does not set a task queue on the edge node. The resource management node is responsible for issuing tokens of resource instances. The resource management node issues tokens based on the token buckets of all resource instances that provide services requested by the task request in the regional resource pool. This can greatly increase the probability of the existence of idle tokens, so that the task request can be allocated a token in a very short time, shortening the waiting time of the task request, improving the response efficiency of the task request, and improving the processing efficiency of the task request.

[0051] The following describes a resource scheduling method, system, device, equipment and storage medium proposed according to an embodiment of the present application in conjunction with the accompanying drawings.

[0052] See also Figure 1 The signaling interaction diagram of the resource scheduling method shown is completed by the interaction between the resource management center and the edge node. Among them, the resource management center is the core part of the distributed network, responsible for global load balancing management, content management, equipment detection and fault alarm in the distributed network. The embodiment of the present application uses the resource management center to perform task scheduling on the resource instances in each processing node in the distributed network. The edge node is a node located at the edge of the network in a distributed network, close to the user or data source, and is used to provide storage, computing, network and other resources. The distributed network may also include multiple processing nodes, and resource instances are provided on the processing nodes, and the resource instances are used to provide computing or storage services to users. The edge node or resource management center provided with the resource instance may also belong to these multiple processing nodes. The above-mentioned resource instances may include but are not limited to virtual machine instances, container instances, cloud server instances, etc., and these instances may also be deployed with artificial intelligence models for providing services such as natural language processing, image processing, and video processing.

[0053] like Figure 1 As shown, the resource scheduling method includes the following steps 101-107.

[0054] Step 101: The edge node receives a task request from a client, where the task request includes a service identifier of the requested service.

[0055] When a user wants to use a service, he can send a task request for the service through his client, and the task request carries the service identifier of the service. The service described in the embodiment of the present application may refer to a service provided by a resource instance deployed in a processing node in a distributed network, and a service identifier may be used to uniquely identify a specific resource instance. Alternatively, a service identifier may also be used to uniquely identify a service, and multiple resource instances may provide the same service, so these multiple resource instances all correspond to the same service identifier.

[0056] The service identifier may be expressed in the form of a URL (Uniform Resource Locator), a URI (Uniform Resource Identifier), etc. of the service corresponding to the service identifier.

[0057] Step 102: The edge node obtains the regional identification information of the regional resource pool to which the resource instance providing the service indicated by the service identifier belongs from the resource partition configuration information; the resource partition configuration information includes the correspondence between the instance information of the resource instance in the geographical area to which the edge node belongs and the regional identification information of the regional resource pool corresponding to the geographical area.

[0058] In an embodiment of the present application, in the resource management center, regional resource pools are divided according to geographical regions. The geographical regions can be divided according to provinces, cities, districts, etc., and resource instances that provide the same service in the same proxy area are divided into the same regional resource pool. Furthermore, the regional resource pools can also be divided into levels according to the scope of the geographical region. For example, if the scope of geographical region A is larger than the scope of geographical region B, and the scope of geographical region A covers the scope of geographical region B, then the resource instances that provide a certain service in geographical region B can be divided into a first-level regional resource pool, and the resource instances that provide the service in geographical region A can be divided into a second-level resource pool. Obviously, the second-level resource pool includes the resource instances in the first-level resource pool.

[0059] After dividing one or more regional resource pools corresponding to the service in the above manner, the resource management center also assigns regional identification information to each regional resource pool, which is used to uniquely identify the regional resource pool to which it corresponds. For each resource instance belonging to a regional resource pool, the resource management center maintains a token bucket corresponding to the resource instance in the regional resource pool, and the token bucket is used to store the token corresponding to the resource instance. The token is used to indicate that the resource instance is idle and a task can be assigned to the resource instance.

[0060] The above-mentioned regional resource pool can be understood as a data set, and the token bucket can be understood as a queue for storing tokens. The regional resource pool is equivalent to a mapping relationship between instance information of multiple resource instances and token buckets. The instance information may include the service identifier of the service provided by the resource instance and the instance identifier of the resource instance. Among them, the instance identifier may include but is not limited to the number or name of the resource instance, and the token bucket of the resource instance may include one or more tokens, indicating that the resource instance is idle and tasks can be scheduled to the resource instance. Alternatively, there may be no token in the token bucket of the resource instance, indicating that the resource instance is currently busy and tasks cannot be scheduled to the resource instance.

[0061] The resource management center also determines the edge nodes in the geographical area corresponding to the regional resource pool, and synchronizes the resource partition configuration information to each edge node in the geographical area, wherein the resource partition configuration information includes the corresponding relationship between the instance information of the resource instance in the geographical area and the regional identification information of the regional resource pool corresponding to the geographical area. The edge node receives the resource partition configuration information synchronized by the resource management center and stores the resource partition configuration information.

[0062] After the edge node receives the task request sent by the client, it can determine the instance information of the service identifier carried by the task request from the locally stored resource partition configuration information, and obtain the regional identification information corresponding to the instance information from the resource partition configuration information. The regional identification information is the regional identification information of the regional resource pool to which the service requested by the client belongs.

[0063] Step 103: The edge node sends the area identification information to the resource management center.

[0064] Step 104: The resource management center receives the regional identification information sent by the edge node, where the regional identification information is used to identify the regional resource pool to which the resource instance providing the service requested by the task request belongs; the regional resource pool includes a token bucket of resource instances providing services in the corresponding geographical area.

[0065] Step 105: Based on the regional identification information, the resource management center obtains an idle token from the token bucket of the resource instance in the regional resource pool.

[0066] Step 106: The resource management center sends the idle tokens to the edge nodes.

[0067] Step 107: The edge node receives the token from the resource management center, and sends the task request to the resource instance indicated by the token.

[0068] The method of the embodiment of the present application divides the regional resource pool according to the geographical area on the resource management node side, and maintains the token bucket of the resource instances that belong to the same geographical area and provide the same service in the regional resource pool, and the resource management node synchronizes the resource partition configuration information to the edge node. In this way, it is achieved that the task queue is not set on the edge node to wait for the allocation of tokens, but the resource management node is responsible for issuing tokens. The resource management node issues tokens based on the token buckets of all resource instances in the regional resource pool that provide the requested service of the task request, which can greatly increase the probability of the existence of idle tokens, so that the task request can be allocated a token in a very short time, shortening the waiting time of the task request, and improving the response efficiency and processing efficiency of the task request.

[0069] In some embodiments of the present application, when a service provider requests to register a new resource instance, the resource management center performs a registration operation for the resource instance, creates a corresponding token bucket for the resource instance, and each token stored in the token bucket includes address information of the resource instance; based on the geographical area where the resource instance is located, the multi-level regional resource pool to which the resource instance belongs is determined; and the instance information of the resource instance is correspondingly associated with the token bucket of the resource instance to each level of regional resource pool to which the resource instance belongs.

[0070] Among them, the registration operation of executing the resource instance includes but is not limited to allocating computing power, storage, network and other resources to the resource instance, determining the processing node deployed by the resource instance, etc. After the resource management center completes the registration process of the resource instance, it creates a token bucket of the resource instance and stores a preset number of free tokens in the token bucket, and the preset number can be 5 or 8, etc. Each token stored in the token bucket includes the address information of the resource instance and a character sequence used to distinguish it from other tokens. The address information can be the URL or URI of the resource instance, etc.

[0071] The resource management center determines the regional resource pool to which the resource instance belongs based on the geographical area where the processing node where the resource instance is deployed is located. The determined regional resource pool may include a primary resource pool, a secondary regional resource pool, or more levels of regional resource pools. The resource management center associates the instance information of the resource instance with the token bucket of the resource instance to the regional resource pools of each level to which the resource instance belongs.

[0072] After each resource instance is registered, the resource management center stores the mapping relationship between the instance information of the resource instance and the token bucket in the regional resource pools at all levels corresponding to the geographical area where the resource instance is deployed. Task scheduling and token allocation for the resource instance are implemented according to the geographical area, so that when a user subsequently requests the service of the resource instance, the resource management center can allocate tokens, and the user's task request does not need to wait too long at the edge node to obtain an idle token. In addition, the token bucket of the resource instance is associated with the regional resource pools at all levels in the geographical area to which it belongs, so that more reasonable management of resource instances is achieved. For the user's task request, the resource instance close to the user can be preferentially scheduled to provide services to the user on the premise that the resource instance has an idle token, which helps to improve service efficiency and service quality.

[0073] In some embodiments of the present application, after the resource management center associates the instance information of the resource instance and the token bucket of the resource instance to the regional resource pools at all levels to which the resource instance belongs, it also includes: determining the instance information of the resource instance and the regional identification information of the regional resource pools at all levels to which the resource instance belongs as the resource partition configuration information corresponding to the resource instance; determining the target geographical area corresponding to the regional resource pool with the smallest coverage in the regional resource pools at all levels; and sending the resource partition configuration information to the edge nodes within the target geographical area.

[0074] Assuming that the regional resource pool to which the resource instance belongs includes a secondary resource pool a and a primary resource pool b, only the resource partition configuration information corresponding to the resource instance is sent to the edge node in the geographical area corresponding to the primary resource pool b.

[0075] For the edge node side, the edge node receives the resource partition configuration information sent by the resource management center and stores the resource partition configuration information.

[0076] In this way, the edge node in the target geographical area corresponding to the regional resource pool with the smallest coverage in the regional resource pools of all levels to which the resource instance belongs can, when receiving a task request for the service of the resource instance, first request a token from the regional resource pool with the smallest coverage in the resource management center based on the resource partition configuration information. If the token can be requested, a resource instance close to the user can be allocated to the task request to provide services, thereby improving the response speed and service quality. If a token cannot be requested from the regional resource pool with the smallest coverage, an idle token can also be requested from a higher-level regional resource pool, thereby increasing the probability of obtaining an idle token and reducing the waiting time for the user's task request. Moreover, by maintaining regional resource pools at all levels in the resource management center and synchronizing the resource partition configuration information to the edge node, relative to the method of allocating tokens by setting task queues corresponding to each resource instance on the edge node, the method of the embodiment of the present application can save storage space of the edge node and improve the flexibility and speed of token allocation.

[0077] In some embodiments of the present application, a failure event of a resource instance is detected, and the instance information and token bucket of the resource instance are deleted from the multi-level regional resource pool to which the resource instance belongs; a configuration update instruction is sent to the edge node in the target geographic area, and the configuration update instruction is used to instruct the edge node to delete the configuration information corresponding to the resource instance from the resource partition configuration information.

[0078] The edge node receives a configuration update instruction sent by the resource management center. The configuration update instruction is sent by the resource management center when it detects a failure event of a resource instance. The configuration update instruction is used to instruct the edge node to delete the configuration information corresponding to the resource instance from the resource partition configuration information; the edge node responds to the configuration update instruction and updates the resource partition configuration information.

[0079] The failure events of resource instances may include resource instance deregistration, failure, or offline.

[0080] After a resource instance fails, the relevant information of the resource instance is deleted from the resource pools at all levels of the resource management center, and the configuration information corresponding to the resource instance is also deleted on the edge node. On the one hand, this can free up the storage space of the resource management center and the edge node, and on the other hand, it can avoid the situation where task requests are still scheduled to the resource instance after the resource instance fails.

[0081] Through the above process, the regional resource pools of various levels to which each resource instance belongs are maintained in the resource management center, and the resource partition configuration information corresponding to each resource instance is also stored on the edge node. Afterwards, when the edge node receives a task request from the client, the task request carries a URL, and the service identifier of the requested service is parsed from the URL. Based on the service identifier, it is queried whether there is instance information containing the service identifier in the instance information of the resource instance included in the locally stored resource partition configuration information. If so, the regional identification information corresponding to the instance information containing the service identifier is obtained from the resource partition configuration information. The regional identification information is the identification information of the regional resource pool to which the resource instance of the service requested by the current task request belongs. And in the case where the resource instance of the service corresponds to a multi-level regional resource pool, the regional identification information is the lowest level regional resource pool in the multi-level regional resource pool corresponding to the resource instance of the service, and the lowest level means that the scope of the geographical area corresponding to the regional resource pool is the smallest in the multi-level regional resource pool.

[0082] After the edge node obtains the regional identification information of the regional resource pool to which the resource identifier of the service requested by the task request belongs, the regional identification information is sent to the resource management center. Based on the regional identification information, the resource management center obtains an idle token from the token bucket of the resource instance in the regional resource pool to which it belongs. Specifically, based on the regional identification information, the resource management center traverses the token bucket corresponding to the resource instance in the multi-level regional resource pool to which the resource instance belongs in the order of the level of the regional resource pool from low to high to see if there is an idle token; when the traversal finds an idle token, the traversal operation is stopped and the traversed idle token is obtained; wherein, in any two adjacent levels of regional resource pools in the multi-level regional resource pool, the coverage of the higher level regional resource pool is greater than the coverage of the lower level regional resource pool. Coverage refers to the scope of the geographical area corresponding to the regional resource pool. For example, the coverage of the higher level regional resource pool can be a province, while the coverage of the lower level regional resource pool can be a city in the province.

[0083] The resource management center queries whether there is an idle token in the token cylinder of the resource instance in the regional resource pool indicated by the regional identification information. If there is, an idle token is obtained and sent to the edge node.

[0084] If the resource management center does not find an idle token in the regional resource pool indicated by the regional identification information sent by the edge node, it indicates that the resource instances in the regional resource pool are all busy. The resource management center will query whether there are idle tokens in the upper-level regional resource pool corresponding to the regional resource pool. If there are, the idle tokens will be sent to the edge node. If there are still no idle tokens in the upper-level regional resource pool, continue to query the upper-level regional resource pool until an idle token is found. Or after querying the regional resource pools at all levels corresponding to the regional resource pool indicated by the regional identification information, no idle tokens are found. In this case, the resource management center sets up a waiting queue, inserts the regional identification information sent by the edge node into the waiting queue, and when idle tokens appear in the regional resource pools at all levels corresponding to the regional resource pool indicated by the regional identification information, the regional identification information is dequeued from the waiting queue and the idle tokens are sent to the edge node.

[0085] By allocating tokens to the client's task request through the above process, there is no need to set up queues corresponding to each resource instance on the edge node, which can save the storage space of the edge node. After receiving the task request, the edge node can locate the regional resource pool to which the resource instance of the requested service belongs based on the service identifier carried by the task request and the resource partition configuration information stored locally, and directly request the resource management center to query the idle tokens in the regional resource pool, which can effectively shorten the time spent on obtaining idle tokens, shorten the response time of the task request, and improve the processing efficiency of the task request.

[0086] Moreover, the regional identification information sent by the edge node indicates the regional resource pool with the smallest geographical area in the regional resource pools of all levels to which the resource instance of the requested service belongs. The processing nodes deployed by the resource instances belonging to the regional resource pool all belong to the same geographical area as the edge node. Therefore, if the resource management center queries an idle token from the regional resource pool, the resource instance allocated for the task request belongs to the same geographical area as the client, which shortens the data transmission distance and is conducive to improving processing efficiency and service quality. If the resource management center does not query an idle token from the regional resource pool, it can also query idle tokens from higher-level regional resource pools step by step. That is to say, when allocating resource instances for the task request, the resource instances are scheduled in the order of distance from the client from near to far. On the one hand, the probability of allocating idle tokens to the task request is increased, and the waiting time of the task request is reduced. On the other hand, the task request does not need to be sent to the resource management center for allocation, and the idle token can be directly obtained by the edge node from the resource management center, thereby improving the response speed. Moreover, the resource instance is allocated nearby, so as to achieve the rapid dispatch of tasks nearby and improve resource utilization. Moreover, by managing the token of the resource instance through the token bucket of the resource instance, the task request for accessing the resource instance can be limited, thereby achieving the purpose of protecting the resource instance.

[0087] After receiving the idle token, the edge node sends the current task request to the processing node where the resource instance indicated by the token is located based on the address information of the resource instance included in the token. After receiving the task request, the processing node assigns the task request to the resource instance for processing, and after obtaining the processing result, returns the processing result to the client, or returns the processing result to the edge node, which forwards the processing result to the client.

[0088] After the resource instance processes the task request, it also sends the task completion notification corresponding to the task request to the resource management center. The resource management center receives the task completion notification sent by the resource instance indicated by the token, which is used to indicate that the resource instance has completed the processing of the task request; a new token is added to the token bucket of the resource instance in the regional resource pool to which the token belongs.

[0089] After the resource instance indicated by the token allocated to the task request completes the processing of the task request, it will notify the resource management center. The resource management center will add a new free token to the token bucket of the resource instance in the regional resource pool to which the token belongs. In this way, after the resource instance completes the processing of the task request, the token can be released in time and a new token can be added to the corresponding token bucket, so that the token management of the resource instance can be cyclical and sustainable.

[0090] In the above embodiments, the edge node goes to the resource management center to obtain an idle token after receiving the task request. In other embodiments of the present application, the edge node may also obtain a certain number of idle tokens from the resource management center in advance, and after receiving the task request, it may first use the locally pre-stored tokens, and then go to the resource management center to obtain tokens if there are no locally pre-stored tokens.

[0091] Specifically, the edge node determines a second preset number of services with the highest request frequency based on historical task data; sends a token acquisition request to the resource management center, the token acquisition request including the service identifiers of the second preset number of services; receives the first preset number of tokens corresponding to each service in the second preset number of services returned by the resource management center; and stores the mapping relationship between the service identifiers of the second preset number of services and the corresponding first preset number of tokens.

[0092] Among them, the historical task data may include the task requests received by the edge node within the past preset time period, and the preset time period may be the past week, the past month, the past three months, the past six months, etc. The task request carries the service identifier of the service to be requested. The edge node may first determine the service identifiers that appear in all task requests received within the past preset time period, and then count the number of task requests carrying each service identifier in all task requests received within the past preset time period. Sort the service identifiers in descending order according to the number of task requests counted, and determine the second preset number of service identifiers in front as the service identifiers of the second preset number of services with the highest request frequency.

[0093] The edge node sends a token acquisition request to the resource management center, and the token acquisition request carries the acquired second preset number of service identifiers. The resource management center receives the token acquisition request sent by the edge node, and based on the service identifier included in the token acquisition request, determines the token bucket of the resource instance that provides the service indicated by these service identifiers from the regional resource pool corresponding to the geographical area where the edge node is located; obtains the first preset number of tokens from the token bucket corresponding to each determined service; and sends the first preset number of tokens corresponding to each service identifier to the edge node. The edge node stores the correspondence between each service identifier and the corresponding first preset number of tokens.

[0094] The first preset number may be, but is not limited to, 5, 8, or 10. The second preset number may be, but is not limited to, 3, 5, or 8.

[0095] Based on historical task data, the edge node analyzes one or more services that are most frequently requested by users, and obtains a certain number of tokens of resource instances that provide these services from the resource management center in advance. In this way, when the current edge node receives a task request requesting these services, it can directly allocate idle tokens to the task request from the edge node without having to request tokens from the resource management center. This can further shorten the time to allocate tokens for task requests and improve the response speed of task requests. Moreover, the tokens of services that are requested more frequently by users are obtained in advance. The tokens of these services are more likely to match the user's task requests, which can relieve the pressure on the resource management center to allocate tokens when there are more task requests requesting these services.

[0096] After pre-storing the first preset number of tokens of one or more services in the edge node in the above manner, when the edge node receives a task request again, the edge node first parses the service identifier of the requested service from the task request, and then queries the mapping relationship between the pre-stored service identifier and the token to see whether the service identifier exists in the mapping relationship. If so, it is determined that the mapping relationship contains a token corresponding to the service requested by the task request. In this case, a token corresponding to the service requested by the task request is obtained from the mapping relationship, and the task request is sent to the resource instance indicated by the obtained token. In the case where the token corresponding to the requested service does not exist in the mapping relationship, the edge node performs the operation of step 102 again, obtains the regional identification information of the regional resource pool to which the resource instance of the service indicated by the service identifier belongs from the resource partition configuration information, and then obtains the corresponding idle token from the resource management center based on the regional identification information. The specific process of obtaining the token has been described in detail in the previous embodiment and will not be repeated here.

[0097] By pre-storing the mapping relationship between the service ID and the token of the service on the edge node, when receiving a task request for these services, the token can be obtained locally first, and then obtained from the resource management center when there is no free token locally. This can minimize the time spent on allocating tokens for task requests, so that the user's task request can be processed without waiting, making the processing of task requests more efficient.

[0098] In some embodiments of the present application, as the first preset number of tokens corresponding to a service identifier stored on the edge node is consumed, when the tokens corresponding to the service identifier are reduced to a certain number, the edge node can obtain the token corresponding to the service identifier from the resource management center to restore the number of tokens corresponding to the service identifier stored locally on the edge node to the first preset number. The above reduction to a certain number can be reduced to zero, or reduced to half of the original number, etc. The edge node can replenish the locally consumed tokens in a timely manner in the above manner.

[0099] In an embodiment of the present application, the edge node can locate the regional resource pool to which the resource instance of the requested service belongs based on the service identifier carried by the task request and the locally stored resource partition configuration information, and directly request the resource management center to query the idle token in the regional resource pool, effectively shortening the time spent on obtaining the token, shortening the response time, and improving processing efficiency. Moreover, the resource management center queries the idle tokens from the low level to the high level according to the resource pools of each level corresponding to the requested service, that is, the resource instances are scheduled in the order from near to far according to the distance from the client, which increases the probability of allocating idle tokens to the task request, reduces the waiting time of the task request, improves the response speed, achieves the effect of quickly dispatching tasks nearby, and improves resource utilization. Moreover, by managing the token of the resource instance through the token bucket of the resource instance, it is possible to limit the flow of task requests to access the resource instance, thereby achieving the effect of protecting the resource instance.

[0100] Some other embodiments of the present application provide a resource scheduling method, which is applied to a resource management center. Figure 2 , the method specifically comprises the following steps:

[0101] Step 201: The resource management center receives the regional identification information sent by the edge node, where the regional identification information is used to identify the regional resource pool to which the resource instance providing the service requested by the task request belongs; the regional resource pool includes a token bucket of resource instances providing services in the corresponding geographical area.

[0102] Step 202: The resource management center obtains an idle token from the token bucket of the resource instance in the regional resource pool based on the regional identification information.

[0103] Step 203: The resource management center sends the token to the edge node, so that the edge node sends the task request to the resource instance indicated by the token for processing.

[0104] The specific operation details of the resource management center are the same as those of the resource management center in the previous embodiment and will not be repeated here.

[0105] The resource scheduling method provided in the above-mentioned embodiment of the present application is based on the same inventive concept as the resource scheduling method provided in the previous embodiment, and has the same beneficial effects as the method adopted, run or implemented by the application program stored therein.

[0106] Some other embodiments of the present application also provide a resource scheduling method, see Figure 3 The method is applied to the edge node, and the method specifically comprises the following steps:

[0107] Step 301: The edge node receives a task request from a client, where the task request includes a service identifier of the requested service.

[0108] Step 302: The edge node obtains the regional identification information of the regional resource pool to which the resource instance providing the service indicated by the service identifier belongs from the resource partition configuration information; the resource partition configuration information includes the correspondence between the instance information of the resource instance in the geographical area to which the edge node belongs and the regional identification information of the regional resource pool corresponding to the geographical area.

[0109] Step 303: The edge node sends the region identification information to the resource management center, so that the resource management center allocates a token to the task request based on the region identification information.

[0110] Step 304: The edge node receives the token from the resource management center and sends a task request to the resource instance indicated by the token.

[0111] The specific operation details of the edge node are the same as those of the edge node in the previous embodiment, and will not be repeated here.

[0112] The resource scheduling method provided in the above-mentioned embodiment of the present application is based on the same inventive concept as the resource scheduling method provided in the previous embodiment, and has the same beneficial effects as the method adopted, run or implemented by the application program stored therein.

[0113] Some other embodiments of the present application also provide a resource scheduling system, see Figure 4 ,The system includes : edge nodes and resource management center. Figure 4 Three edge nodes are schematically drawn in FIG. In practical applications, any number of edge nodes may be included.

[0114] The edge node is used to receive a task request from a client, where the task request includes a service identifier of a requested service; obtain, from resource partition configuration information, the regional identification information of a regional resource pool to which a resource instance providing the service indicated by the service identifier belongs; the regional identification information is used to identify the regional resource pool to which a resource instance providing the service requested by the task request belongs; the resource partition configuration information includes a correspondence between instance information of a resource instance in a geographic region to which the edge node belongs and regional identification information of a regional resource pool corresponding to the geographic region; send the regional identification information to a resource management center, receive a token allocated by the resource management center based on the regional identification information, and send a task request to the resource instance indicated by the token;

[0115] The resource management center is used to receive the regional identification information sent by the edge node; based on the regional identification information, obtain the idle token from the regional resource pool; and send the token to the edge node.

[0116] The resource scheduling system provided in the above-mentioned embodiments of the present application and the resource scheduling method provided in the previous embodiments are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the application programs stored therein.

[0117] The embodiment of the present application also provides a resource scheduling device, which is used to perform the operations of the resource management center in the resource scheduling method provided in any of the above embodiments. Figure 5 , the device comprises:

[0118] The receiving module 401 is used to receive the regional identification information sent by the edge node, where the regional identification information is used to identify the regional resource pool to which the resource instance providing the service requested by the task request belongs; the regional resource pool includes a token bucket of resource instances providing services in the corresponding geographical area;

[0119] The token acquisition module 402 is used to acquire an idle token from the regional resource pool based on the regional identification information;

[0120] The sending module 403 is used to send the token to the edge node, so that the edge node sends the task request to the resource instance indicated by the token for processing.

[0121] The token acquisition module 402 is used to traverse the token bucket corresponding to the resource instance in the multi-level regional resource pool to which the resource instance belongs, based on the regional identification information, in order from low to high levels of the regional resource pool, to see if there are any free tokens; when traversing to find free tokens, stop the traversal operation and obtain the traversed free tokens; wherein, in any two adjacent levels of regional resource pools in the multi-level regional resource pools, the coverage of the higher-level regional resource pool is greater than the coverage of the lower-level regional resource pool.

[0122] Before receiving the regional identification information sent by the edge node, the device also includes: a regional resource pool management module, which is used to perform a registration operation of the resource instance, create a corresponding token bucket for the resource instance, and each token stored in the token bucket includes the address information of the resource instance; determine the multi-level regional resource pool to which the resource instance belongs according to the geographical area where the resource instance is located; and associate the instance information of the resource instance with the token bucket of the resource instance to the regional resource pools at each level to which the resource instance belongs.

[0123] After associating the instance information of the resource instance and the token bucket of the resource instance to the regional resource pools at all levels to which the resource instance belongs, the regional resource pool management module is also used to determine the instance information of the resource instance and the regional identification information of the regional resource pools at all levels as the resource partition configuration information corresponding to the resource instance; determine the target geographical area corresponding to the regional resource pool with the smallest coverage in the regional resource pools at all levels; and send the resource partition configuration information to the edge nodes within the target geographical area.

[0124] After associating the instance information of the resource instance and the token bucket of the resource instance to the regional resource pools at all levels to which the resource instance belongs, the regional resource pool management module is also used to detect the failure event of the resource instance, delete the instance information and token bucket of the resource instance from the multi-level regional resource pools to which the resource instance belongs; send a configuration update instruction to the edge node in the target geographic area, the configuration update instruction is used to instruct the edge node to delete the configuration information corresponding to the resource instance from the resource partition configuration information.

[0125] After sending the token to the edge node, the device also includes: a token recovery module, which is used to receive a task completion notification sent by the resource instance indicated by the token, and the task completion notification is used to indicate that the resource instance has completed the processing of the task request; and add a new token to the token bucket of the resource instance in the regional resource pool to which the token belongs.

[0126] The device also includes: a token pre-acquisition module, which is used to receive a token acquisition request sent by an edge node, the token acquisition request including a service identifier; based on the service identifier, determine a token bucket that provides a resource instance of the service indicated by the service identifier from a regional resource pool corresponding to the geographical area where the edge node is located; obtain a first preset number of tokens from the determined token bucket; and send the first preset number of tokens to the edge node.

[0127] The resource scheduling device provided in the above-mentioned embodiment of the present application and the resource scheduling method provided in the embodiment of the present application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the application programs stored therein.

[0128] The embodiment of the present application also provides a resource scheduling device, which is used to perform the operation of the edge node in the resource scheduling method provided in any of the above embodiments. Figure 6 , the device comprises:

[0129] The receiving module 501 is used to receive a task request from a client, where the task request includes a service identifier of a requested service;

[0130] An acquisition module 502 is used to acquire, from the resource partition configuration information, the regional identification information of the regional resource pool to which the resource instance providing the service indicated by the service identifier belongs; the resource partition configuration information includes the correspondence between the instance information of the resource instance in the geographical area to which the edge node belongs and the regional identification information of the regional resource pool corresponding to the geographical area;

[0131] A sending module 503, used to send the region identification information to the resource management center, so that the resource management center allocates a token to the task request based on the region identification information;

[0132] The receiving module 501 is also used to receive a token from a resource management center;

[0133] The sending module 503 is further configured to send a task request to the resource instance indicated by the token.

[0134] Before obtaining the regional identification information of the regional resource pool to which the resource instance providing the service indicated by the service identifier belongs from the resource partition configuration information, the device also includes: a storage module for receiving the resource partition configuration information sent by the resource management center and storing the resource partition configuration information.

[0135] The device also includes: an update module, which is used to receive a configuration update instruction sent by a resource management center. The configuration update instruction is sent by the resource management center when it detects a failure event of a resource instance. The configuration update instruction is used to instruct the edge node to delete the configuration information corresponding to the resource instance from the resource partition configuration information; in response to the configuration update instruction, the resource partition configuration information is updated.

[0136] The device also includes: a token pre-acquisition module, which is used to determine a second preset number of services with the highest request frequency based on historical task data; send a token acquisition request to a resource management center, the token acquisition request including the service identifiers of the second preset number of services; receive the first preset number of tokens corresponding to each service in the second preset number of services returned by the resource management center; and store the mapping relationship between the service identifiers of the second preset number of services and the corresponding first preset number of tokens.

[0137] After storing the mapping relationship between the service identifiers of the second preset number of services and the corresponding first preset number of tokens, the device also includes: a token allocation module, which is used to receive a task request, and if there is a token corresponding to the service requested by the task request in the mapping relationship, obtain a token corresponding to the service requested by the task request from the mapping relationship, and send the task request to the resource instance indicated by the obtained token; if there is no token corresponding to the requested service in the mapping relationship, execute to obtain the regional identification information of the regional resource pool to which the resource instance of the service indicated by the service identifier belongs from the resource partition configuration information.

[0138] The resource scheduling device provided in the above-mentioned embodiment of the present application and the resource scheduling method provided in the embodiment of the present application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the application programs stored therein.

[0139] The present application also provides an electronic device to execute the resource scheduling method. Figure 7 , which shows a schematic diagram of an electronic device provided by some embodiments of the present application. Figure 7As shown, the electronic device 10 includes: a processor 1000, a memory 1001, a bus 1002 and a communication interface 1003, and the processor 1000, the communication interface 1003 and the memory 1001 are connected via the bus 1002; the memory 1001 stores a computer program that can be run on the processor 1000, and the processor 1000 executes the resource scheduling method provided in any of the aforementioned embodiments of the present application when running the computer program.

[0140] The memory 1001 may include a high-speed random access memory (RAM), and may also include a non-volatile memory, such as at least one disk memory. The communication connection between the system network element and at least one other network element is realized through at least one communication interface 1003 (which may be wired or wireless), and the Internet, wide area network, local area network, metropolitan area network, etc. may be used.

[0141] The bus 1002 may be an ISA bus, a PCI bus, or an EISA bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. The memory 1001 is used to store programs, and the processor 1000 executes the programs after receiving execution instructions. The resource scheduling method disclosed in any implementation of the aforementioned embodiment of the present application may be applied to the processor 1000, or implemented by the processor 1000.

[0142] The processor 1000 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit or software instructions in the processor 1000. The above processor 1000 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a readily available programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiments of the present application can be directly embodied as a hardware decoding processor to be executed, or the hardware and software modules in the decoding processor can be combined and executed. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 1001, and the processor 1000 reads the information in the memory 1001 and completes the steps of the above method in combination with its hardware.

[0143] The electronic device provided in the embodiment of the present application and the resource scheduling method provided in the embodiment of the present application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented therein.

[0144] The present application also provides a computer-readable storage medium corresponding to the resource scheduling method provided in the above embodiment. Figure 8 The computer-readable storage medium shown is a CD 30 on which a computer program (ie, a program product) is stored. When the computer program is run by a processor, the resource scheduling method provided by any of the aforementioned implementations will be executed.

[0145] It should be noted that examples of the computer-readable storage medium may also include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other optical or magnetic storage media, which are not listed here one by one.

[0146] The computer-readable storage medium provided in the above-mentioned embodiments of the present application and the resource scheduling method provided in the embodiments of the present application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the application programs stored therein.

[0147] An embodiment of the present application also provides a computer program product corresponding to the resource scheduling method provided in the aforementioned embodiment. The computer program product includes a computer program, and the computer program is executed by a processor to implement the resource scheduling method provided in the aforementioned embodiment.

[0148] The computer program product provided by the above-mentioned embodiments of the present application and the resource scheduling method provided by the embodiments of the present application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the application programs stored therein.

[0149] It should be noted that:

[0150] In the description provided herein, a large number of specific details are described. However, it is understood that the embodiments of the present application can be practiced without these specific details. In some instances, well-known structures and technologies are not shown in detail so as not to obscure the understanding of this description.

[0151] Similarly, it should be understood that in order to streamline the present application and help understand one or more of the various inventive aspects, in the above description of the exemplary embodiments of the present application, the various features of the present application are sometimes grouped together into a single embodiment, figure, or description thereof. However, the disclosed method should not be interpreted as reflecting the following schematic diagram: the claimed application requires more features than the features explicitly stated in each claim. More specifically, as reflected in the claims below, the inventive aspects are less than all the features of the single embodiment disclosed above. Therefore, the claims following the specific embodiment are hereby expressly incorporated into the specific embodiment, wherein each claim itself serves as a separate embodiment of the present application.

[0152] In addition, those skilled in the art will appreciate that, although some embodiments described herein include certain features included in other embodiments but not other features, the combination of features of different embodiments is meant to be within the scope of the present application and form different embodiments. For example, in the claims below, any one of the claimed embodiments may be used in any combination.

[0153] The above is only a preferred specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A resource scheduling method, characterized in that: Applied to a resource management center, the method comprises: Receiving regional identification information sent by an edge node, the regional identification information is used to identify a regional resource pool to which a resource instance providing a service requested by the task request belongs; the regional resource pool includes a token bucket of resource instances providing the service in a corresponding geographical area; Based on the area identification information, obtaining an idle token from a token bucket of a resource instance in a resource pool of the corresponding area; The token is sent to the edge node, so that the edge node sends the task request to the resource instance indicated by the token for processing.

2. The method according to claim 1, characterized in that The acquiring, based on the area identification information, an idle token from a token bucket of a resource instance in a resource pool of the area to which the resource instance belongs, comprises: Based on the regional identification information, in order from low to high levels of the regional resource pools, traverse level by level to see whether there are free tokens in the token bucket corresponding to the resource instance in the multi-level regional resource pools to which the resource instance belongs; When an idle token is found during the traversal, the traversal operation is stopped, and the idle token found during the traversal is acquired; Among them, in any two adjacent levels of regional resource pools in the multi-level regional resource pools, the coverage range of the higher-level regional resource pool is greater than the coverage range of the lower-level regional resource pool.

3. The method according to claim 1 or 2, characterized in that: Before receiving the area identification information sent by the edge node, the method further includes: Execute a registration operation of the resource instance, create a corresponding token bucket for the resource instance, each token stored in the token bucket includes address information of the resource instance; Determine, according to the geographical area where the resource instance is located, a multi-level regional resource pool to which the resource instance belongs; The instance information of the resource instance and the token bucket of the resource instance are correspondingly associated with the regional resource pools at all levels to which the resource instance belongs.

4. The method according to claim 3, characterized in that After the instance information of the resource instance and the token bucket of the resource instance are correspondingly associated with the regional resource pools at all levels to which the resource instance belongs, the method further includes: Determine the resource partition configuration information corresponding to the resource instance by using the instance information of the resource instance and the regional identification information of the regional resource pools at each level; Determine the target geographical area corresponding to the regional resource pool with the smallest coverage among the regional resource pools at all levels; The resource partition configuration information is sent to edge nodes within the target geographic area.

5. The method according to claim 4, characterized in that After the instance information of the resource instance and the token bucket of the resource instance are correspondingly associated with the regional resource pools at all levels to which the resource instance belongs, the method further includes: Detecting a failure event of the resource instance, deleting instance information and a token bucket of the resource instance from the multi-level regional resource pool to which the resource instance belongs; A configuration update instruction is sent to an edge node in the target geographical area, wherein the configuration update instruction is used to instruct the edge node to delete the configuration information corresponding to the resource instance from the resource partition configuration information.

6. The method according to claim 1 or 2, characterized in that: After sending the token to the edge node, the method further includes: Receive a task completion notification sent by the resource instance indicated by the token, wherein the task completion notification is used to indicate that the resource instance has completed processing of the task request; A new token is added to the token bucket of the resource instance in the regional resource pool to which the token belongs.

7. The method according to claim 1 or 2, characterized in that: The method further comprises: Receiving a token acquisition request sent by the edge node, wherein the token acquisition request includes a service identifier; Based on the service identifier, determining a token bucket of a resource instance that provides the service indicated by the service identifier from a regional resource pool corresponding to the geographical area where the edge node is located; Obtaining a first preset number of tokens from the determined token bucket; The first preset number of tokens is sent to the edge node.

8. A resource scheduling method, characterized in that: Applied to an edge node, the method comprises: Receiving a task request from a client, wherein the task request includes a service identifier of a requested service; Acquire, from the resource partition configuration information, the regional identification information of the regional resource pool to which the resource instance providing the service indicated by the service identifier belongs; the resource partition configuration information includes the correspondence between the instance information of the resource instance in the geographical area to which the edge node belongs and the regional identification information of the regional resource pool corresponding to the geographical area; Sending the region identification information to a resource management center so that the resource management center allocates a token to the task request based on the region identification information; The token is received from the resource management center, and the task request is sent to the resource instance indicated by the token.

9. The method according to claim 8, characterized in that Before acquiring the regional identification information of the regional resource pool to which the resource instance providing the service indicated by the service identification belongs from the resource partition configuration information, the method further includes: The resource partition configuration information sent by the resource management center is received, and the resource partition configuration information is stored.

10. The method according to claim 8 or 9, characterized in that: The method further comprises: receiving a configuration update instruction sent by the resource management center, wherein the configuration update instruction is sent by the resource management center when the resource management center detects a failure event of the resource instance, and the configuration update instruction is used to instruct the edge node to delete the configuration information corresponding to the resource instance from the resource partition configuration information; In response to the configuration update instruction, the resource partition configuration information is updated.

11. The method according to claim 8 or 9, characterized in that: The method further comprises: Determine, based on the historical task data, a second preset number of services with the highest request frequencies; Sending a token acquisition request to the resource management center, wherein the token acquisition request includes the service identifiers of the second preset number of services; receiving a first preset number of tokens corresponding to each service in the second preset number of services returned by the resource management center; The mapping relationship between the service identifiers of the second preset number of services and the corresponding first preset number of tokens is stored.

12. The method according to claim 11, characterized in that After storing the mapping relationship between the service identifiers of the second preset number of services and the corresponding first preset number of tokens, the method further includes: Upon receiving the task request, if there is a token corresponding to the service requested by the task request in the mapping relationship, obtaining a token corresponding to the service requested by the task request from the mapping relationship, and sending the task request to the resource instance indicated by the obtained token; In the case that the token corresponding to the requested service does not exist in the mapping relationship, the step of obtaining, from the resource partition configuration information, the regional identification information of the regional resource pool to which the resource instance providing the service indicated by the service identification belongs.

13. A resource scheduling system, characterized in that: The system includes: an edge node and a resource management center; The edge node is used to receive a task request from a client, wherein the task request includes a service identifier of a requested service; obtain, from resource partition configuration information, the regional identification information of a regional resource pool to which a resource instance providing the service indicated by the service identifier belongs; the regional identification information is used to identify the regional resource pool to which the resource instance providing the service requested by the task request belongs; the resource partition configuration information includes a correspondence between instance information of a resource instance in a geographic region to which the edge node belongs and regional identification information of a regional resource pool corresponding to the geographic region; send the regional identification information to the resource management center, receive a token allocated by the resource management center based on the regional identification information, and send the task request to the resource instance indicated by the token; The resource management center is used to receive the area identification information sent by the edge node; based on the area identification information, obtain an idle token from the regional resource pool to which it belongs; and send the token to the edge node.

14. A resource scheduling device, characterized in that: Applied to a resource management center, the device comprises: A receiving module, configured to receive regional identification information sent by an edge node, wherein the regional identification information is used to identify a regional resource pool to which a resource instance providing the service requested by the task request belongs; the regional resource pool includes a token bucket of resource instances providing the service in a corresponding geographical area; A token acquisition module, used to acquire an idle token from the regional resource pool to which the token belongs based on the regional identification information; The sending module is used to send the token to the edge node, so that the edge node sends the task request to the resource instance indicated by the token for processing.

15. A resource scheduling device, characterized in that: Applied to an edge node, the device comprises: A receiving module, used for receiving a task request from a client, wherein the task request includes a service identifier of a requested service; an acquisition module, configured to acquire, from resource partition configuration information, the regional identification information of the regional resource pool to which the resource instance providing the service indicated by the service identifier belongs; the resource partition configuration information includes a correspondence between instance information of the resource instance in the geographical area to which the edge node belongs and regional identification information of the regional resource pool corresponding to the geographical area; A sending module, used for sending the area identification information to a resource management center, so that the resource management center allocates a token to the task request based on the area identification information; The receiving module is further used to receive the token from the resource management center; The sending module is further used to send the task request to the resource instance indicated by the token.

16. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: The processor runs the computer program to implement the method according to any one of claims 1-7 or 8-12.

17. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the method according to any one of claims 1-7 or 8-12.

18. A computer program product, comprising a computer program, characterized in that The computer program is executed by a processor to implement the method according to any one of claims 1 to 7 or 8 to 12.

Images