Computing power resource sensing system, computing power resource sensing method, computing power resource sensing equipment and medium

Abstract

The embodiment of the invention provides a computing power resource sensing system and method, equipment and a medium, and relates to the field of cloud native containers. The computing power resource sensing system comprises a sensing controller and a multi-cloud manager, the sensing controller comprises a computing power sensing module and a data synchronization module, the computing power sensing module monitors whether each computing power resource pool changes or not, and when it is monitored that the first computing power resource pool changes, the multi-cloud manager sends the first computing power resource pool to the multi-cloud manager. And calling the multi-cloud manager to query manufacturer specification data in the first computing power resource pool and send the manufacturer specification data to the data synchronization module, recording the manufacturer specification data in a manufacturer specification table by the data synchronization module, abstracting the manufacturer specification data into a node format, and recording the node format in a node table. Through the mode, real-time perception of the computing power resources is realized, and the computing power resources are converted into a unified representation form, so that scheduling and distribution of the computing power resources are facilitated.

Description

Technical Field

[0001] The present application relates to the field of cloud native container technology, and more specifically, to a computing resource perception system, method, device and medium. Background Art

[0002] With the rise of cloud computing and edge computing, computing nodes are widely distributed and large in scale, the heterogeneity of computing resources is increasing, and the dynamic nature of computing resources is becoming more and more obvious. In order to meet these challenges, computing power network technology has emerged. It uses distributed computing technology to achieve parallel processing of tasks, relies on high-performance network communication to achieve collaboration between nodes, and senses and identifies available computing resources in real time, and can automatically adapt to changes in computing resources and scale expansion. For computing power networks, how to achieve unified management and measurement of computing power, unified orchestration and scheduling, and unified cross-domain scheduling are key technical issues for computing power networks. Traditional computing models usually statically assign computing tasks to predefined nodes and cannot dynamically adapt to changes and imbalances in computing resources. Based on computing power resource perception technology, computing power networks can sense and identify available computing resources in the network in real time, thereby optimizing task scheduling and resource management. The importance and necessity of computing power resource perception technology in modern distributed computing environments cannot be ignored.

[0003] At present, there is no unified standard for the implementation of computing power resource perception. Computing power resource perception technology faces some challenges and difficulties in practice, including the following aspects:

[0004] 1. Data heterogeneity and consistency: The heterogeneity of computing nodes is a major challenge for computing resource perception. Different types of nodes may have different hardware configurations, operating systems, and performance characteristics. Therefore, computing resource perception technology needs to process data from various heterogeneous nodes and convert them into unified indicators and representations for resource comparison and selection.

[0005] 2. Node dynamics and changes: In a distributed computing environment, the dynamics and changes of computing nodes are common. Nodes can join or leave the network at any time, and their computing power and availability may also change. Computing resource perception technology needs to be able to perceive and adapt to these changes in a timely manner, and dynamically update and adjust resource perception results.

[0006] 3. Scalability and extensibility: Computing resource perception technology needs to have good scalability and extensibility to adapt to the needs of large-scale distributed computing environments. As the number of nodes increases, the amount of data and computational complexity will also increase. Therefore, computing resource perception technology needs to design and implement efficient algorithms and architectures to support large-scale resource perception and management.

[0007] Therefore, a computing resource perception method that can transform and unify heterogeneous node data, dynamically perceive node changes, and meet node scalability is urgently needed. Summary of the invention

[0008] Based on the above technical problems, the embodiments of the present application provide a computing resource perception system, method, device and medium, which aim to perceive computing resources in real time and convert them into a unified form of expression to facilitate scheduling and allocation.

[0009] A first aspect of an embodiment of the present application provides a computing resource perception system, the system comprising:

[0010] A perception controller and a multi-cloud manager, wherein the perception controller includes a computing power perception module and a data synchronization module;

[0011] The computing power perception module monitors whether each computing power resource pool changes. When a change is detected in the first computing power resource pool, the multi-cloud manager is called to query the manufacturer specification data under the first computing power resource pool and send it to the data synchronization module, where the manufacturer specification data includes: specification information of computing power resources;

[0012] The data synchronization module records the manufacturer specification data in a manufacturer specification table, and abstracts the manufacturer specification data into a node format and records it in a node table.

[0013] Optionally, the system further comprises:

[0014] The computing power perception module periodically calls the multi-cloud manager to obtain the vendor regional water level data under each computing power resource pool, and sends it to the data synchronization module, wherein the vendor regional water level data includes: name, associated resource pool identification code, total processor amount, available processor amount, total memory amount, and available memory amount;

[0015] The data synchronization module records the manufacturer area water level data in the resource pool water level table, and abstracts the manufacturer area water level data into a node format and records it in the node table.

[0016] Optionally, the system further comprises:

[0017] The computing power sensing module monitors whether each sales computing power resource changes. When a change is detected in a first sales computing power resource, the computing power sensing module queries the sales computing power resource information of the first sales computing power resource and sends it to the data synchronization module. The sales computing power resource information includes: name, identification code, specification identification code, associated resource pool identification code, selling price, and sold-out information.

[0018] The data synchronization module abstracts the sales computing power resource information into a node format and records it in the node table.

[0019] Optionally, the system further comprises:

[0020] The computing power perception module periodically calls the multi-cloud manager to obtain the sold-out information of each sales computing power resource, and synchronously updates the sold-out information to the sales computing power resource table and sends it to the data synchronization module;

[0021] The data synchronization module abstracts the sold-out information of the sales computing power resources into a node format and records it in the node table.

[0022] Optionally, the system further comprises:

[0023] A scheduler, the scheduler comprising a cache module;

[0024] The scheduler monitors data changes in the node table and the unit table, and records the data changes in the cache module;

[0025] The scheduler schedules the corresponding computing resources in the node table according to the computing deployment requirements in the unit table.

[0026] Optionally, the system further comprises:

[0027] Acquire the computing power deployment requirement of the user through the perception controller, and send the computing power deployment requirement to the data synchronization module;

[0028] The data synchronization module abstracts the computing power deployment requirements into a unit format and records it in the unit table.

[0029] A second aspect of an embodiment of the present application provides a computing resource perception method, the method comprising:

[0030] Monitor each computing resource pool for changes, and when a change is detected in the first computing resource pool, query the manufacturer specification data under the first computing resource pool;

[0031] The manufacturer specification data is recorded in a manufacturer specification table, and the manufacturer specification data is abstracted into a node format and recorded in a node table.

[0032] Optionally, the method further comprises:

[0033] Obtain the regional water level data of each computing resource pool;

[0034] The manufacturer area water level data is recorded in a resource pool water level table, and the manufacturer area water level data is abstracted into a node format and recorded in the node table.

[0035] Optionally, the method further comprises:

[0036] Monitor whether each sales computing power resource changes, and when a change is detected in a first sales computing power resource, query the sales computing power resource information of the first sales computing power resource and send it to the data synchronization module;

[0037] The sales computing power resource information is abstracted into a node format and recorded in the node table.

[0038] Optionally, the method further comprises:

[0039] Periodically calling the multi-cloud manager to obtain the sold-out information of each sales computing power resource, and synchronously updating the sold-out information to the sales computing power resource table and sending it to the data synchronization module;

[0040] The sold-out information of the sales computing power resources is abstracted into a node format and recorded in the node table.

[0041] Optionally, the method further comprises:

[0042] Monitoring data changes in the node table and the unit table, and recording the data changes in the cache module;

[0043] According to the computing power deployment requirements in the unit table, the corresponding computing power resources in the node table are scheduled.

[0044] Optionally, the method further comprises:

[0045] Obtain the computing power deployment requirements of the user, and send the computing power deployment requirements to the data synchronization module;

[0046] The computing power deployment requirements are abstracted into a unit format and recorded in the unit table.

[0047] A third aspect of an embodiment of the present application provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed by the processor, a computing power resource perception method as in the second aspect of an embodiment of the present application is implemented.

[0048] A fourth aspect of an embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the computing power resource perception method of the second aspect of the embodiment of the present application is implemented.

[0049] The fifth aspect of the embodiments of the present application provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the computing power resource perception method of the second aspect of the embodiments of the present application.

[0050] Through the computing power resource perception system proposed in the embodiment of the present application, the computing power perception module in the perception controller is used to monitor the changes occurring in each resource pool. When a change occurs, the multi-cloud manager is called to obtain the manufacturer specification data under the first computing power resource pool and sent to the data synchronization module in the perception controller. The data synchronization module records the data in the manufacturer specification table and abstracts it into a node format and records it in the node table. In the same way, the computing power perception module can also periodically call the multi-cloud manager to obtain the manufacturer's regional water level data under each computing power resource pool, and also record and abstract it. The computing power perception module can also monitor changes in sales computing power resources, and abstract its related computing power resource information into a node format and record it in the node table. After receiving the user's computing power deployment requirements, the perception controller abstracts the computing power deployment requirements into a unit format and records it in the unit table. Then the scheduler schedules computing power resources by using the contents in the unit table and the node table.

[0051] In this application, by abstracting various information related to computing power resources into node format and computing power deployment requirements into unit format, different computing power resources are unified into the same presentation format, and resource allocation, comparison and selection can be facilitated; at the same time, through the real-time monitoring of computing power resource pool changes by the perception controller, as well as the timed monitoring of resource pool water level data, sales computing power resource sold-out information, etc., real-time perception of computing power resource changes and dynamic updates and adjustments based on changes are achieved; and, in this application, computing power resource specification information in each computing power resource pool is obtained, and the unified abstract planning is made into node format, which can adapt to the requirements of large-scale distributed computing environment and has good scalability and extensibility. Therefore, the computing power resource perception system proposed in this application meets the needs of perceiving and managing large-scale computing power resources, and greatly improves the work efficiency of users in scheduling computing power resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments of the present application 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 drawings can be obtained based on these drawings without paying any creative labor.

[0053] Figure 1 This is a schematic diagram of a computing resource perception system proposed in an embodiment of the present application;

[0054] Figure 2 It is a flowchart of the steps of a computing resource perception method proposed in one embodiment of the present application;

[0055] Figure 3 It is a schematic diagram of an electronic device shown in an embodiment of the present application. DETAILED DESCRIPTION

[0056] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0057] In the drawings, the size of the constituent elements, the thickness of the layer or the area may be exaggerated for the sake of clarity. Therefore, any implementation of the present disclosure is not necessarily limited to the size shown in the drawings, and the shapes and sizes of the components in the drawings do not reflect the true proportions. In addition, the drawings schematically show ideal examples, and any implementation of the present disclosure is not limited to the shapes or values ​​shown in the drawings.

[0058] In the related technologies, the huge computing resource network has put forward higher requirements for the perception and scheduling of computing resources. Since the computing resource network relies on high-performance network communication to achieve mutual cooperation between nodes, the unified management measurement, unified orchestration and scheduling, and unified cross-domain scheduling of computing resources in the computing resource network are key technical issues for computing resource network management. Therefore, computing resource perception technology is of great significance to the computing network platform. Computing resource perception technology needs to have the ability to perceive and identify available computing resources in the network in real time, and provide basic data for decision-making such as task scheduling, resource allocation and load balancing by monitoring and collecting information such as the computing power, performance indicators and availability of nodes. Computing resource perception technology enables the computing network platform to intelligently adapt to the changes and imbalances of computing resources and achieve more efficient distributed computing and resource management. Secondly, computing resource perception technology can automatically perceive the addition and departure of new nodes, dynamically adapt to the increase and decrease of computing resources, thereby achieving the expansion of system scale and flexible use of resources, so that the computing network can cope with the ever-changing computing environment and needs, maintain efficient computing power and adapt to the growing workload.

[0059] Therefore, based on the application needs of computing power resource perception in related technologies, this application proposes a computing power resource perception system, including:

[0060] A perception controller and a multi-cloud manager, wherein the perception controller includes a computing power perception module and a data synchronization module;

[0061] The computing power perception module monitors whether each computing power resource pool changes. When a change is detected in the first computing power resource pool, the multi-cloud manager is called to query the manufacturer specification data under the first computing power resource pool and send it to the data synchronization module, where the manufacturer specification data includes: specification information of computing power resources;

[0062] The data synchronization module records the manufacturer specification data in a manufacturer specification table, and abstracts the manufacturer specification data into a node format and records it in a node table.

[0063] In the embodiment of the present application, in order to meet the demand for computing power resource perception technology in the relevant technology, a computing power resource perception system is proposed, which includes a perception controller and a multi-cloud manager. As the core component of the computing power perception system, the perception controller undertakes most of the work in the computing power perception process, which is specifically divided into a computing power perception module and a data synchronization module.

[0064] First, the computing power perception module will monitor whether there are changes in each computing power resource pool, such as the addition, change, deletion and other specific operations of the computing power resource pool. The computing power resource pool is a collection of computing power resources. The information of each computing power resource pool will be recorded in the computing power resource pool table. The computing power resource pool table is stored in a database. In an optional embodiment, the database can be a database such as MySQL. Among them, the computing power resource pool table contains the relevant information of all computing power resource pools managed by the computing power resource perception system. If a new computing power resource pool needs to be added to the system, the computing power resource pool related information needs to be added to the computing power resource pool table. Specifically, the computing power resource pool information contained in the computing power resource pool table may include name (id), identification code (uuid), cloud account name (cloud account id), resource pool configuration (resourcepool config) and other information.

[0065] After monitoring the changes in the computing power resource pool, the computing power perception module will call the multi-cloud manager to perform relevant queries. The multi-cloud manager is a special manager proposed in this application to obtain information from each computing power resource pool. After a computing power resource pool changes, the multi-cloud manager is called to obtain relevant information in the computing power resource pool, specifically the manufacturer specification data under the computing power resource pool. We call this computing power resource pool the first computing power resource pool, where the manufacturer specification data contains the specification information of the computing power resources under the computing power resource pool. These specification information can be information such as the type and capacity of computing power resources that reflect their specific computing capabilities. By obtaining the specification information of computing power resources of different computing power resources, different computing power resources under different resource pools can be included in a unified comparison standard for comparison and classification. This unified management of the specification information of computing power resources can not only manage computing power resources more clearly and controllably, but also provide clearer and more reliable comparison standards when users need to use computing power resources, which is convenient for users to select computing power resources. The vendor specification data of the computing resource pool obtained by the multi-cloud manager may include: name (id), identification code (uuid), associated resource pool identification code (pool uuid), specification information (cloudprovider config), plug-in code (plugin code) and other information.

[0066] The computing power perception module obtains the manufacturer specification data under the first computing power resource pool by calling the multi-cloud manager query, and it is also necessary to synchronize this information to the data synchronization module that belongs to the perception controller. As a module for data recording and data conversion in the perception controller, the data synchronization module will record these data in the corresponding manufacturer specification table after receiving the manufacturer specification data synchronized by the computing power perception module. The manufacturer specification table is also stored in the system database and is used to record the manufacturer specification data of each manufacturer under the computing power resource pool. This is not only a record of data, but also a backup of the data. At the same time, the data synchronization module will abstract the obtained manufacturer specification data into a node format, extract the relevant information of each computing power resource in the manufacturer specification data, which can be specifically the name (id), the associated resource pool identification code (pool uuid), specification information (spec) and other information, and record the abstracted results in the node table. The node table is stored in the system database and can not only be accessed and modified by the perception controller, but also serve as the basis for subsequent computing power resource scheduling.

[0067] In this way, after sensing the changes in the computing resource pool, the computing resource perception system finally abstracts the specific computing resources in the computing resource pool into a node format that is convenient for management and scheduling, and stores them in the node table for management and scheduling, thereby achieving unified planning of computing resources and solving the problem of heterogeneity of computing nodes in the computing network.

[0068] In a specific embodiment, the present application can be implemented through the open source container orchestration platform k8s (Kubernetes, Helmsman), adopting the Informer mechanism in k8s and the List-Watch method, in which the computing power perception module monitors the computing power resource pool in the computing power resource pool table. In response to changes in the computing power resource pool, the manufacturer specification data of the computing power resource pool is obtained through the multi-cloud manager and synchronized to the data synchronization module. The data synchronization module abstracts the data into a node format that can be recognized by the k8s platform and records it in the node table.

[0069] In combination with the above embodiments, in one implementation, the present application further provides a fault handling system, which further includes:

[0070] The computing power perception module periodically calls the multi-cloud manager to obtain the vendor regional water level data under each computing power resource pool, and sends it to the data synchronization module, wherein the vendor regional water level data includes: name, associated resource pool identification code, total processor amount, available processor amount, total memory amount, and available memory amount;

[0071] The data synchronization module records the manufacturer area water level data in the resource pool water level table, and abstracts the manufacturer area water level data into a node format and records it in the node table.

[0072] In the embodiment of the present application, since the computing power resources in the system are not static, while the computing power resource pool described above may change, the computing power resources already existing in the computing power resource pool in the system will also change. The ultimate goal of establishing a computing power network is to allocate and schedule computing power resources according to different user requirements.

[0073] Therefore, on the basis that computing power resources in the system may undergo various changes, the computing power resource perception system also needs to understand and control the changes in computing power resources. Specifically, the computing power perception module in the perception controller will periodically call the multi-cloud manager to obtain the vendor regional water level data under each computing power resource pool in the system. The vendor regional water level data reflects the utilization rate and capacity of the computing power resource pool, specifically including the current usage status and the remaining available resources. The computing power resources in each computing power resource pool are collectively regarded as a reservoir, in which the amount of water that has been contained and the amount of water that can be contained are the meaning of the vendor regional water level data. Specifically, the vendor regional water level data may include the following information: name (id), associated resource pool identification code (pool uuid), total processor amount (cpu total), available processor amount (cpuavailable), total memory amount (mem total), available memory amount (mem available), etc. Among them, the associated resource pool identification code is related to which computing resource pool it is associated with, the total number of processors represents the total number of processors available in the manufacturer's area, and the maximum computing power demand that can be carried, and the available number of processors is the total number of processors that can be selected by users in the manufacturer's area, excluding occupied computing resources and unavailable computing resources. Similarly, the total amount of memory represents the total amount of memory available in the manufacturer's area, and the available amount of memory is the total amount of memory that can be selected by users in the manufacturer's area, excluding occupied computing resources and unavailable computing resources.

[0074] After obtaining the manufacturer regional water level data in the above-mentioned resource pools, it is also necessary to convert these data into a format that the system can recognize and schedule. Similarly, it is necessary to synchronize the manufacturer regional water level data obtained in the computing power perception module through the data synchronization module. The data synchronization module records the manufacturer regional water level data. Specifically, the data synchronization module records the manufacturer regional water level data in the resource pool water level table for recording and backup. The resource pool water level table is also stored in the database. At the same time, the data synchronization module abstracts the manufacturer regional water level data into a node format and records it in the node table. Specifically, it is reflected in the status information (status) in the node table for subsequent computing power resource scheduling.

[0075] In combination with the above embodiments, in one implementation, the present application further provides a fault handling system, which further includes:

[0076] The computing power sensing module monitors whether each sales computing power resource changes. When a change is detected in a first sales computing power resource, the computing power sensing module queries the sales computing power resource information of the first sales computing power resource and sends it to the data synchronization module. The sales computing power resource information includes: name, identification code, specification identification code, associated resource pool identification code, selling price, and sold-out information.

[0077] The data synchronization module abstracts the sales computing power resource information into a node format and records it in the node table.

[0078] In the embodiment of the present application, in the process of unified management and scheduling of computing resources in each computing resource pool, since each computing resource belongs to a different manufacturer, due to various reasons, not all computing resources are often provided to users for selection and scheduling. Therefore, it is also necessary to divide the computing resources available for user selection into sales computing resources. Similarly, the sales computing resources also need to be sensed and supervised by a perception controller.

[0079] Specifically, the relevant information of each sales computing power resource is stored in the sales computing power resource table, which is stored in the database. The computing power perception module monitors the changes of the sales computing power resources in the sales computing power resource table, and the specific changes can be addition, change, deletion, etc. When the first sales computing power resource is monitored to change, the computing power resource perception module will query the sales computing power resource information of the first sales computing power resource in the sales resource table, wherein the sales computing power resource information includes various relevant information attached when the computing power resource is used as a sales computing power resource. The sales computing power resources are divided and managed through these sales computing power resource information, which is not only a unified management of the computing power resources in each computing power resource pool, but also convenient for users to select computing power resources. Among them, the sales computing power resource information can specifically include: name (id), identification code (uuid), specification identification code (spec uuid), associated resource pool identification code (pool uuid), price (price), sold out information (sold out), etc., which reflect the various attributes of the computing power resource as a sales computing power resource. Among them, the specification identification code is associated with the computing power specification information of the computing power resources for sale, and the associated resource pool identification code is associated with which computing power resource pool the computing power resources for sale belong to. In addition, the selling price is also a factor that needs to be considered when sensing computing power resources. Since the computing power resource network integrates a large number of computing power resources from different computing power resource pools, when users choose computing power resources, there will inevitably be a variety of computing power resources that meet the user's usage needs. Since different manufacturers operate computing power resources in different computing power resource pools, under the same usage conditions, the conditions of each computing power resource are also different. This means that when users choose computing power resources, they need to consider not only whether the computing power resources can meet the usage needs and whether there is a computing power resource overflow, but also the cost-effectiveness of the computing power resources. If the usage needs can be met, whether there are options with lower prices. The sold-out information reflects whether the sales computing power resources are occupied. If they are occupied, they will be displayed as sold out in the sold-out information.

[0080] After obtaining the sales computing power resource information of the first sales computing power resource, the computing power perception module synchronizes the sales computing power resource information to the data synchronization module. After receiving the sales computing power resource information, the data synchronization module abstracts the sales computing power resource information into a node format and records it in the node table, which can be specifically reflected in the product identification code (product uuid), specification information (spec), status information (status) and label (lable) in the node table for use in subsequent computing power resource scheduling.

[0081] In combination with the above embodiments, in one implementation, the present application further provides a fault handling system, which further includes:

[0082] The computing power perception module periodically calls the multi-cloud manager to obtain the sold-out information of each sales computing power resource, and synchronously updates the sold-out information to the sales computing power resource table and sends it to the data synchronization module;

[0083] The data synchronization module abstracts the sold-out information of the sales computing power resources into a node format and records it in the node table.

[0084] In the embodiment of the present application, after sensing the sales computing power resources and abstracting the sales computing power resource information of the sales computing power resources into a node format and recording them in the node table, it is also necessary to monitor the sold-out information of the sales computing power resources. As mentioned above, as a computing power resource available for users to choose, whether the sales computing power resources are sold out is a critical factor. Since the computing power resources in different computing power resource pools can not only be managed by the computing power resource perception system proposed in this application, but can also be accessed and called through other platforms, therefore, it is not possible to judge whether the sales computing power resources are idle or occupied based solely on whether they have been sold, and it is necessary to monitor whether they are sold out.

[0085] In this application, the multi-cloud manager is called periodically through the computing power perception module, and the sold-out information of each sales computing power resource is obtained through the multi-cloud manager. Through this timed query method, it is determined whether each sales computing power resource is in a sold-out state or an unsold-out state, which facilitates the computing power resource perception system to uniformly manage and control the sales computing power resources. If a user selects a sales computing power resource and finds that the sales computing power resource is sold out, it will not only take up the user's time in selecting the sales computing power resource, but may also affect the user's original computing power resource application plan. Therefore, it is necessary to obtain the sold-out information of the sales computing power resources periodically.

[0086] After the computing power perception module obtains the sold-out information of the sales computing power resources, it also needs to send the sold-out information to the data synchronization module synchronously. In addition, when the sold-out information changes, for example, if a sales computing power resource was not sold out originally, but it is found that the sales computing power resource has been sold out after obtaining the sold-out information regularly, it is necessary to modify the sold-out information of the sales computing power resource in the sales computing power resource table in time. At the same time, the data synchronization module also needs to abstract the sold-out information into a node format and record it in the node table, specifically, it can be reflected in the label of the node table.

[0087] In combination with the above embodiments, in one implementation, the present application further provides a fault handling system, which further includes:

[0088] A scheduler, the scheduler comprising a cache module;

[0089] The scheduler monitors data changes in the node table and the unit table, and records the data changes in the cache module;

[0090] The scheduler schedules the corresponding computing resources in the node table according to the computing deployment requirements in the unit table.

[0091] In the embodiment of the present application, in addition to the perception controller and the multi-cloud manager, the computing power resource perception system also includes a scheduler, which is used to perform the computing power resource scheduling part. The scheduler is the core of the computing power resource scheduling, and the computing power resources are scheduled by matching the user's needs with the computing power resources in the system. Specifically, the scheduler monitors the data in the node table and the unit table, and makes scheduling actions according to the monitoring results. Among them, the node table is as described above, which records the name (id) of each computing power resource, the associated resource pool identification code (pool uuid), the product identification code (product uuid), the specification information (spec), the status information (status) and the label (lable) and other information. This information represents the various related information of the computing power resources that need to be considered when scheduling computing power resources, and provides optional solutions to the user's computing power deployment needs based on the computing power resource information of each computing power resource.

[0092] Here, the user's computing power deployment requirements are reflected in the unit table. The unit table stores the user's computing power deployment requirements. The scheduler will respond to data changes in the unit table and schedule the available computing power resources in the node table according to the user's computing power deployment requirements. It will provide users with computing power resource options that meet the computing power deployment requirements for users to choose from, and allocate the computing power resources selected by the user to the user for use, thereby realizing the scheduling of computing power resources.

[0093] At the same time, the scheduler also includes a cache module (Scheduler Cache), which caches some key information of the scheduler during the scheduling process. After the scheduler monitors the data changes in the node table and the unit table, it will synchronize the monitored data to the cache module. Among them, the key information cached by the cache module includes but is not limited to the status of computing resources, available computing resources, allocated computing resources, sold out computing resources, recent scheduling results, completed user demand scheduling, computing resources information scheduled to users, etc. In other words, all key information of each link in the computing resource scheduling process can be cached in the cache module in the scheduler. In this way, not only the scheduling performance and efficiency can be improved, but also the actions performed in the scheduling can be recorded.

[0094] In combination with the above embodiments, in one implementation, the present application further provides a fault handling system, which further includes:

[0095] Acquire the computing power deployment requirement of the user through the perception controller, and send the computing power deployment requirement to the data synchronization module;

[0096] The data synchronization module abstracts the computing power deployment requirements into a unit format and records it in the unit table.

[0097] In an embodiment of the present application, the process of scheduling computing power resources also includes obtaining the user's computing power deployment requirements. As described above, the scheduler provides the user with available computing power resources in the node table based on the user's computing power deployment requirements in the unit table during the process of scheduling computing power resources, and among them, it is necessary to first obtain the user's computing power deployment requirements.

[0098] In this application, the computing power deployment requirements of the user are obtained through the perception controller, which may include the type of computing power resources that the user expects to select, the expected price, the time required, etc. After obtaining the computing power deployment requirements of the user, the perception controller will synchronize the computing power deployment requirements to the data synchronization module, and the data synchronization module will also abstract the computing power deployment requirements, abstract the computing power deployment requirements into a unit format, and record them in the unit table. Among them, the computing power deployment requirements recorded in the unit table may include: name (id), user name (user id) and node name (node ​​id), etc. By abstracting the user's needs into a unit format, the scheduler can identify them to perform the scheduling of computing power resources. So far, the closed loop of the data flow of computing power resources in each computing power resource pool and the computing power resource perception system control flow is completed through the perception controller, multi-cloud manager and scheduler, realizing the real-time perception and synchronization of computing power resources, unifying the representation of computing power resources, thereby achieving the purpose of optimizing computing power resource scheduling and computing power resource management.

[0099] In a specific embodiment, as described above, the present application can be implemented through the open source container orchestration platform k8s (Kubernetes, helmsman), using the Informer mechanism in k8s, using the List-Watch method, the scheduler monitors the data changes in the node table (node ​​table) and the pod table (unit table), and performs the scheduling of computing resources in response to the data changes in the node table and the pod table. In this way, in the embodiment of the present application, on the basis of the k8s platform, the original computing resource perception process is converted into the node (node) format and pod (unit) format that the k8s platform can recognize, and the scheduling of computing resources, that is, the process of a user's computing power deployment demand seeking computing power resources to achieve, is converted into the process of selecting the optimal solution node for the pod on the k8s platform. Therefore, the scheduler pre-selection, optimal scheduling mechanism and the extension mechanism of the Scheduling Framework on the k8s platform can also be reused to achieve a more unified computing resource perception and division, and the scheduling of computing resources can be performed more efficiently and reasonably.

[0100] In summary, in an optional embodiment, the computing power resource perception system proposed in this application is as follows: Figure 1 As shown, Figure 1This is a schematic diagram of a computing power resource perception system proposed in an embodiment of the present application. As shown in the figure, the computing power perception module in the perception controller monitors the computing power resource pool information in the computing power resource pool table. In response to changes in the first computing power resource pool, it calls the multi-cloud manager to query the manufacturer specification data under the first computing power resource pool, and sends the query results to the data synchronization module. The data synchronization module records the manufacturer specification data in the manufacturer specification table, and abstracts it into a node format and records it in the node table; and the computing power perception module in the perception controller will periodically call the multi-cloud manager to obtain the manufacturer area water level data under each computing power resource pool, and synchronously send it to the data synchronization module. The data synchronization module records the manufacturer area water level data in the resource pool water level table, and abstracts it into a node format and records it in the node table; the computing power perception module in the perception controller will also monitor Listen to the changes of sales computing power resources in the sales computing power resource table, and synchronously send the sales computing power resource information of the first sales computing power resource whose Dharma body has changed to the data synchronization module. The data synchronization module abstracts it into a node format and records it in the node table; the computing power perception module in the perception controller will also call the multi-cloud manager at regular intervals to query the sold-out information of each sales computing power resource. Similarly, it will be synchronously sent to the data synchronization module. The data synchronization module abstracts it into a node format and records it in the node table; the perception controller will also obtain the user's computing power deployment requirements, and abstract it into a unit format, and record it in the unit table. The scheduler listens to the data changes in the node table and the unit table, and records it in the cache module. The scheduler will also schedule the computing power resources in the node table according to the computing power deployment requirements in the unit table.

[0101] The present application also proposes a computing resource perception method, such as Figure 2 As shown, Figure 2 It is a flowchart of a computing power resource perception method provided in an embodiment of the present application. The method is applied to the computing power resource perception system in the embodiment of the present application, including steps S201 to S202:

[0102] Step S201: monitoring each computing resource pool for changes, and when a change is detected in the first computing resource pool, querying the manufacturer specification data under the first computing resource pool;

[0103] Step S202: Record the manufacturer specification data in a manufacturer specification table, and abstract the manufacturer specification data into a node format and record it in a node table.

[0104] In combination with the above embodiments, in one implementation, the present application further provides a computing resource perception method, including:

[0105] Obtain the regional water level data of each computing resource pool;

[0106] The manufacturer area water level data is recorded in a resource pool water level table, and the manufacturer area water level data is abstracted into a node format and recorded in the node table.

[0107] In combination with the above embodiments, in one implementation, the present application further provides a computing resource perception method, including:

[0108] Monitor whether each sales computing power resource changes, and when a change is detected in a first sales computing power resource, query the sales computing power resource information of the first sales computing power resource and send it to the data synchronization module;

[0109] The sales computing power resource information is abstracted into a node format and recorded in the node table.

[0110] In combination with the above embodiments, in one implementation, the present application further provides a computing resource perception method, including:

[0111] Periodically calling the multi-cloud manager to obtain the sold-out information of each sales computing power resource, and synchronously updating the sold-out information to the sales computing power resource table and sending it to the data synchronization module;

[0112] The sold-out information of the sales computing power resources is abstracted into a node format and recorded in the node table.

[0113] In combination with the above embodiments, in one implementation, the present application further provides a computing resource perception method, including:

[0114] Monitoring data changes in the node table and the unit table, and recording the data changes in the cache module;

[0115] According to the computing power deployment requirements in the unit table, the corresponding computing power resources in the node table are scheduled.

[0116] In combination with the above embodiments, in one implementation, the present application further provides a computing resource perception method, including:

[0117] Obtain the computing power deployment requirements of the user, and send the computing power deployment requirements to the data synchronization module;

[0118] The computing power deployment requirements are abstracted into a unit format and recorded in the unit table.

[0119] Based on the same design concept, another embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps in the computing power resource perception method described in any of the above embodiments of the present application are implemented.

[0120] Based on the same design concept, another embodiment of the present application provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps in the computing power resource perception method described in any of the above embodiments of the present application.

[0121] Based on the same design concept, another embodiment of the present application provides an electronic device, such as Figure 3 shown. Figure 3 This is a schematic diagram of an electronic device shown in an embodiment of the present application. The electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes, the steps in the computing power resource perception method described in any of the above embodiments of the present application are implemented.

[0122] As for the method embodiment, since it is basically similar to the system embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the system embodiment.

[0123] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0124] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, devices, or computer program products. Therefore, the embodiments of the present application may adopt the form of complete hardware embodiments, complete software embodiments, or embodiments in combination with software and hardware. Moreover, the embodiments of the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0125] The embodiments of the present application are described with reference to the flowcharts and / or block diagrams of the methods, terminal devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0126] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0127] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0128] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present application.

[0129] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0130] The above is a detailed introduction to a computing resource perception system, method, device and medium provided by the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for general technical personnel in this field, according to the idea of ​​the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A computing resource perception system, characterized in that: include: A perception controller and a multi-cloud manager, wherein the perception controller includes a computing power perception module and a data synchronization module; The computing power perception module monitors whether each computing power resource pool changes. When a change is detected in the first computing power resource pool, the multi-cloud manager is called to query the manufacturer specification data under the first computing power resource pool and send it to the data synchronization module, where the manufacturer specification data includes: specification information of computing power resources; The data synchronization module records the manufacturer specification data in a manufacturer specification table, and abstracts the manufacturer specification data into a node format and records it in a node table.

2. The computing resource perception system according to claim 1, characterized in that: The computing power perception module periodically calls the multi-cloud manager to obtain the vendor regional water level data under each computing power resource pool, and sends it to the data synchronization module, wherein the vendor regional water level data includes: name, associated resource pool identification code, total processor amount, available processor amount, total memory amount, and available memory amount; The data synchronization module records the manufacturer area water level data in the resource pool water level table, and abstracts the manufacturer area water level data into a node format and records it in the node table.

3. The computing resource perception system according to claim 1, characterized in that: The computing power sensing module monitors whether each sales computing power resource changes. When a change is detected in a first sales computing power resource, the computing power sensing module queries the sales computing power resource information of the first sales computing power resource and sends it to the data synchronization module. The sales computing power resource information includes: name, identification code, specification identification code, associated resource pool identification code, selling price, and sold-out information. The data synchronization module abstracts the sales computing power resource information into a node format and records it in the node table.

4. The computing resource perception system according to claim 3, characterized in that: The computing power perception module periodically calls the multi-cloud manager to obtain the sold-out information of each sales computing power resource, and synchronously updates the sold-out information to the sales computing power resource table and sends it to the data synchronization module; The data synchronization module abstracts the sold-out information of the sales computing power resources into a node format and records it in the node table.

5. The computing resource perception system according to claim 1, characterized in that: Also included: a scheduler, the scheduler including a cache module; The scheduler monitors data changes in the node table and the unit table, and records the data changes in the cache module; The scheduler schedules the corresponding computing resources in the node table according to the computing deployment requirements in the unit table.

6. The computing resource perception system according to claim 5, characterized in that: Acquire the computing power deployment requirement of the user through the perception controller, and send the computing power deployment requirement to the data synchronization module; The data synchronization module abstracts the computing power deployment requirements into a unit format and records it in the unit table.

7. A computing resource perception method, characterized in that: Applied to the computing resource perception system according to any one of claims 1 to 6, the method comprising: Monitor each computing resource pool for changes, and when a change is detected in the first computing resource pool, query the manufacturer specification data under the first computing resource pool; The manufacturer specification data is recorded in a manufacturer specification table, and the manufacturer specification data is abstracted into a node format and recorded in a node table.

8. The computing resource perception method according to claim 7, characterized in that: Also includes: Obtain the regional water level data of each computing resource pool; The manufacturer area water level data is recorded in a resource pool water level table, and the manufacturer area water level data is abstracted into a node format and recorded in the node table.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the computer program is executed by the processor, the computing power resource perception method as described in claim 7 or 8 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the computing power resource perception method as described in claim 7 or 8 is implemented.

Images