Resource-aware scheduling method, system, device and medium for edge cloud environment

Abstract

The application discloses a kind of resource perception scheduling methods, systems, equipment and medium for edge cloud environment.The method comprises: in response to task scheduling request, based on the preset load index, the load index value corresponding to each item of candidate end node is collected, the load index includes CPU occupancy, memory occupancy, node task proportion and temporary storage margin;The load index value is standardized, and the objective weight corresponding to each load index is generated based on the standardized load index value;Based on the objective weight corresponding to each load index, the score of each candidate end node is calculated, and the optimal scheduling node is selected based on the score of each candidate end node, and task scheduling is executed.Compared with the prior art, the application can make task resource scheduling more comprehensive and fine.At the same time, the accuracy of task resource scheduling is improved, and resource use tilt is prevented.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of network communication, and particularly relates to a resource-aware scheduling method, system, device and medium for an edge cloud environment. BACKGROUND

[0002] With the rapid development of 5G, Internet of Things (IoT), industrial Internet and other technologies, edge computing, as a new computing paradigm, is gradually becoming an important way to solve distributed computing and data processing. However, the complexity and heterogeneity of the edge computing environment bring a series of challenges to resource management and task scheduling. In the traditional cloud computing environment, resource management is mainly concentrated in large data centers, while in the edge computing environment, the number of devices is large, the computing power is not the same, and the network bandwidth is limited, making the task scheduling extremely complex. Edge devices usually include smartphones, sensors, smart home devices, etc., which exist in a decentralized manner at the edge of the network, with limited resources, limited energy, and limited computing power, which puts higher requirements on the real-time performance and energy efficiency of tasks.

[0003] The default scheduling algorithm of the existing edge cloud environment can only process CPU and memory resources in extreme cases, which cannot meet the needs of different applications for multiple resources, resulting in insufficient comprehensive and fine resource scheduling. At the same time, the static weight allocation method cannot adapt to the dynamic changes of resource utilization in the edge cloud environment, which is easy to cause excessive tilt of resource use.

[0004] Therefore, in view of the above technical problems, it is necessary to provide a new task scheduling strategy for use in the edge cloud environment.

[0005] The information disclosed in this BACKGROUND section is only intended to increase an understanding of the general context of the present application and is not to be taken in any way as an acknowledgment or any form of suggestion that this information forms prior art that is already known in the art. SUMMARY

[0006] The present application aims to provide a resource-aware scheduling method, system, device and medium for an edge cloud environment, which can improve the flexibility and adaptability of task resource scheduling, prevent resource overtilt, and meet more needs.

[0007] To achieve the above-mentioned purpose, the technical scheme provided by an embodiment of the present application is as follows:

[0008] In a first aspect, the present application provides a resource-aware scheduling method for an edge cloud environment, comprising:

[0009] In response to the task scheduling request, based on the preset load index, the load index values corresponding to the candidate end nodes are collected, the load index includes CPU occupancy rate, memory occupancy rate, node task proportion and temporary storage capacity;

[0010] The load index values are standardized, and the objective weight corresponding to each load index is generated based on the standardized load index values;

[0011] Based on the objective weight corresponding to each load index, the score of each candidate end node is calculated, and the optimal scheduling node is selected to perform task scheduling.

[0012] In one or more embodiments of the application, the method further comprises:

[0013] Obtaining the demand load index value of the task to be scheduled;

[0014] If the end node load index value meets the demand load index value, the end node is listed as a candidate end node;

[0015] If the end node load index value does not meet the demand load index value, the end node is filtered so that the end node that does not meet the demand load index value does not participate in this task scheduling.

[0016] In one or more embodiments of the application, the standardization formula of the load index value is:

[0017]

[0018] Wherein, x nj is the load index value of the jth load index corresponding to the nth candidate end node, max(x j ) is the maximum value of the jth load index in each candidate end node, and min(x j ) is the minimum value of the jth load index in each candidate end node.

[0019] In one or more embodiments of the application, the formula for generating the objective weight corresponding to each load data based on the standardized load data is:

[0020]

[0021] Wherein, x nj is the load index value of the jth load index corresponding to the nth candidate end node; is the average value of the jth load index of each candidate end node; N is the total number of candidate end nodes; x′ ni represents the standardized load index value of the ith index corresponding to the nth candidate end node; denotes the average value of the i-th index after normalization, p is the number of indexes constituting the load data, W j denotes the weight value corresponding to the j-th load index, r ij denotes the correlation coefficient of the i-th load index and the j-th load index.

[0022] In one or more embodiments of the present application, the method further comprises:

[0023] Deploying a monitoring component for monitoring and collecting load index values corresponding to the alternative end node at the edge node.

[0024] In one or more embodiments of the present application, the score of the current node is calculated based on the objective weight corresponding to each load data of the alternative end node, comprising:

[0025] Multiplying each load index value of the alternative end node by the weight value corresponding to the load index, and taking the sum of the product of each load index value and the weight value corresponding to the load index as the current score of the alternative end node.

[0026] In a second aspect, the present application provides a resource-aware scheduling system for edge cloud environment, comprising:

[0027] The acquisition module is configured to, in response to a task scheduling request, acquire each load index value corresponding to the alternative end node based on the preset load index, wherein the load index includes CPU occupancy, memory occupancy, node task proportion, and temporary storage capacity.

[0028] The processing module is configured to normalize the load index value, and generate an objective weight corresponding to each load index based on the normalized load index value.

[0029] The scheduling module is configured to calculate the score of each alternative end node based on the objective weight corresponding to each load index, and select the optimal scheduling node to perform task scheduling.

[0030] In one or more embodiments of the present application, the system is further configured to:

[0031] Obtain the demand load index value of the task to be scheduled.

[0032] If the end node load index value meets the demand load index value, the end node is listed as an alternative end node.

[0033] If the end node load index value does not meet the demand load index value, the end node is filtered so that the end node that does not meet the demand load index value does not participate in this task scheduling.

[0034] In a third aspect, the present application provides a computer device, comprising a memory and a processor, which are connected in communication with each other, and the memory stores computer instructions, and the processor executes the computer instructions to perform the resource-aware scheduling method for edge cloud environment.

[0035] In a fourth aspect, the present application provides a computer readable storage medium, which stores computer instructions for making a computer execute the resource-aware scheduling method for edge cloud environment.

[0036] Compared with the prior art, the resource-aware scheduling method for edge cloud environment provided by the present application can cover various computing indicators in the edge cloud environment, and has high flexibility in the selection of computing indicators. Different indicators have different processing strategies based on their action mechanisms, so as to meet the needs of different applications for various resources, and make the task resource scheduling more comprehensive and fine. At the same time, the present application adopts the strategy of collecting the resource utilization of the node performance in the application deployment process in real time, and generates the objective weight of the task scheduling in this round in real time. The objective weight dynamically changes based on the real-time data of the node indicators, and more accurately reflects the dynamic changes of the resources in the edge cloud environment, which is suitable for various cloud environment implementation scenarios, can improve the accuracy of task resource scheduling, and prevent resource use tilt. BRIEF DESCRIPTION OF DRAWINGS

[0037] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments described in the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0038] Figure 1 is a schematic diagram of an edge cloud environment system in an embodiment of the present application;

[0039] Figure 2 is a flowchart of the resource-aware scheduling method for edge cloud environment in an embodiment of the present application;

[0040] Figure 3 is a structural block diagram of the resource-aware scheduling system for edge cloud environment in an embodiment of the present application;

[0041] Figure 4 is a structural block diagram of an electronic device in an embodiment of the present application. DETAILED DESCRIPTION

[0042] In order to make the person skilled in the art better understand the technical solutions in the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor should belong to the scope of protection of the present application.

[0043] Unless otherwise clearly indicated, throughout the description and the claims, the term "comprise" or variations such as "comprises" or "comprising" will be understood to imply the inclusion of a stated element or group of elements but not the exclusion of any other element or group of elements.

[0044] In one technical solution, the method for the edge device node to perform task resource scheduling specifically includes: obtaining CPU occupancy and memory occupancy corresponding to the end node, and calculating the score of the candidate node by using fixed uniform weight based on the default scheduling algorithm under the edge cloud environment. Then, the task resource is scheduled based on the score.

[0045] For example, the preferred algorithms in the default scheduler are roughly as follows:

[0046] Minimum request priority score:

[0047]

[0048] Resource balance allocation score:

[0049]

[0050] Wherein, Score is the score value of the node, S cpu is the total CPU time, N cpu is the idle time of the CPU, S mem is the total memory size, N mem is the used memory size.

[0051] Based on the above formula, the scheme quantifies the node state by CPU usage and memory usage, and then allocates task resources. The setting of the fixed weight value is only to control the value range of the final node score value, and does not reflect the actual resource situation of the current node. This static weight allocation method cannot adapt to the dynamic changes of resource utilization in the edge cloud environment, and is easy to cause excessive tilt of resource use. At the same time, only from the CPU occupancy rate and the memory occupancy rate, the current edge cloud computing environment cannot accurately reflect the state of each node. The algorithm itself does not consider the variability and conflict of resource indicators, and direct allocation cannot meet the needs of different applications for multiple resources, and resource scheduling is not comprehensive and fine.

[0052] The inventors of the present application found the main shortcomings of the above-mentioned technology, and based on the shortcomings of the technology, a new technical implementation idea is proposed: the load indicators used for calculation in the node are multi-dimensional, and the weights of the resource indicators are dynamically calculated through the multi-dimensional indicators. The variability and conflict of resource indicators are fully considered, and an objective weight corresponding to each resource indicator can be allocated according to the actual running situation of the cluster, and the weight is recalculated each time the scheduling is performed, and then the node score is calculated to select the best scheduling node, which can better adapt to the complex demand environment. It meets the problems of poor real-time performance and insufficient resource perception in actual implementation scenarios.

[0053] As Figure 1 The edge cloud environment system schematic diagram is shown, the structure of the present application mainly relates to edge cloud computing environment, its core component includes: edge node: distributed in the network edge, close to the data source computing entity. Responsible for collecting the resource and state information of the edge node, and the scheduling decision of the task; end device: distributed in the network most edge, the computing entity that generates data. Running some basic tasks; cloud data center: provide additional computing resources, process large-scale or non-real-time high requirements tasks. These components jointly constitute a multi-level computing environment, and the present application focuses on the resource allocation and management between the edge node and the end device, that is, the task scheduling process from the edge node to the end node.

[0054] Please refer to Figure 2 The flowchart of the resource perception scheduling method for the edge cloud environment in an embodiment of the present application is shown. The resource perception scheduling method for the edge cloud environment specifically includes the following steps:

[0055] S201: In response to a task scheduling request, based on a preset load indicator, collecting each load indicator value corresponding to the candidate end node;

[0056] It should be noted that the load index in the present application refers to the factors affecting the processing capacity of the end node to the task resource in the task resource scheduling of the end node, which can include but is not limited to CPU occupancy rate, memory occupancy rate, node task proportion, and temporary storage margin, etc. The present application embodiment does not limit the selection of the load index and the number of load indexes.

[0057] The CPU occupancy rate is a measure of the amount of time the CPU spends executing various tasks and processes on the computer, which is calculated as the ratio of the time required for the CPU to execute and process various tasks and programs to the total CPU time in the present application embodiment. The memory occupancy rate value is the ratio of the used memory to the total memory, the used memory refers to the sum of the physical memory currently allocated for use by the application program, and the total memory refers to the total amount of physical memory available to the system. The task proportion refers to the ratio of the number of tasks running on each node to the number of all tasks running in the cluster. This index is used to measure the workload of a single node in the cluster. By calculating the proportion of the number of tasks running on each node to the total number of tasks, the load of the node can be evaluated, thereby helping the scheduling algorithm to more reasonably allocate new tasks and avoid overloading some nodes while other nodes have idle resources. The temporary storage margin refers to the local storage used by the host node to store temporary data generated during task execution, which directly affects the running environment and performance of the task. If the temporary storage margin resource of a node is excessively consumed, it may cause the node to be unable to run new Pods or affect the performance of existing Pods.

[0058] In the present application embodiment, the way of collecting the corresponding load data of the end node is not limited, in an exemplary embodiment, a monitoring component for monitoring and collecting the load index values corresponding to the alternative end nodes can be deployed on the edge node.

[0059] For example, the load index values corresponding to the alternative end nodes can be collected by deploying a Metrics component on the edge node, and then monitoring the data collected by the Proxy agent in the end node, and then monitoring the resource usage of the end nodes in the cluster. The Metrics provides a powerful toolkit for measuring the behavior of key components in a production environment, and some behaviors in the service can be monitored and counted. Proxy is an important network proxy component, and Proxy can route requests to appropriate nodes according to the decision of the scheduler, so Proxy agent is deployed on each end node. For another example, prometheus can also be used to monitor the memory usage, CPU usage and other load indexes of the end nodes.

[0060] It should be noted that for a specific to-be-scheduled task, not all end nodes can meet the task demand of the to-be-scheduled task. For example, the to-be-scheduled task requires the remaining memory of the corresponding processing end node to be 1G, and the remaining memory of node A is only 500M. If node A is not processed, node A will still participate in the data collection and scoring calculation, causing waste of computing resources. If node A is elected, it may even cause the task to be unable to be processed, data loss, and other adverse situations. Therefore, the end nodes that do not meet the task demand should be filtered before the election of the scheduling node, so that they do not participate in the election of the scheduling node of the task, and only the nodes that meet the task resource demand are considered for task scheduling.

[0061] In a specific embodiment of the present application, the resource-aware scheduling for the edge cloud environment further comprises: obtaining a demand load index value of a to-be-scheduled task; if the end node load index value meets the demand load index value, the end node is listed as a candidate end node; if the end node load index value does not meet the demand load index value, the end node is filtered, so that the end node that does not meet the demand load index value does not participate in this task scheduling.

[0062] For example, for a certain to-be-scheduled task P j , the demand vector is D j , and the set of nodes N is filtered, and the node n i , the local resource vector is F j , when D j <F j , the node n i is added to the schedulable sequence. Wherein, less than here means that each element in D i is less than or equal to the corresponding element in the vector F j .

[0063] S202: standardizing the load index value, and generating an objective weight corresponding to each load index based on the standardized load index value;

[0064] It should be noted that because different indexes have different effects on the processing task resource capacity of the node, for different load indexes, the embodiments of the present application adopt a classification standardization strategy. Specifically, the load indexes used for end node scoring are divided into positive indexes and reverse indexes. The greater the index value of the positive index, the higher the corresponding end node score, which is more conducive to the processing of related task resources by the node. The greater the index value of the reverse index, the lower the corresponding end node score, which is less conducive to the processing of related task resources by the node.

[0065] In an exemplary embodiment of the present application, the standardization formula for the load data is:

[0066]

[0067] wherein x nj is the load index value of the jth load index corresponding to the nth alternative end node, max(x j ) is the maximum value of the jth load index in each alternative end node, and min(x j ) is the minimum value of the jth load index in each alternative end node.

[0068] Based on the above formula, after the user analyzes and selects the load index based on the actual implementation scenario, first, the index is classified based on the nature of the index. Further, the load index values of different types on the alternative end node are standardized by corresponding methods. Such processing method can maximize the original characteristics of the load data, and the standardized result is also the most accurate.

[0069] The present application also provides another group of embodiments. After selecting the load index and classifying, the inverse index in the selected load index can be converted into a positive index based on the preset index normalization method, such as the inverse transformation method, and then all the indexes can be brought into the same group of standardization formulas. The purpose of data standardization required by the present application can also be achieved. It can be understood that the positive index can also be uniformly converted into the inverse index, and the present application will not repeat the description of this embodiment.

[0070] Further, the formula for generating the objective weight corresponding to each load data based on the standardized load index value is:

[0071]

[0072] wherein x nj is the load index value of the jth load index corresponding to the nth alternative end node; is the average value of the jth load index of each alternative end node; N is the total number of alternative end nodes; x′ ni represents the standardized load index value corresponding to the ith index of the nth alternative end node; represents the average value of the standardized index, p is the number of indexes constituting the load data, W j is the weight value corresponding to the jth load index, r ij is the correlation coefficient of the ith load index and the jth load index.

[0073] Standard deviation S jThe dispersion of the load index j in each end node can be measured. The greater the value of the standard deviation, the greater the difference in the values of the load index of each end node, the more information it can reflect, and the stronger the evaluation strength of the load index itself, so more weight should be allocated to the load index.

[0074] r ij The correlation coefficient between the load index i and j is represented, and the correlation coefficient is used to represent the correlation between the load indexes. The stronger the correlation with other load indexes, the less conflictive the load index is with other load indexes, the more information it reflects, and the more repetitive the evaluation content it can embody, which to some extent weakens the evaluation strength of the load index and reduces the weight allocated to the index.

[0075] Finally, the correlation coefficient of each end node load index and the standard deviation corresponding to each load index value can be combined to calculate the weight value corresponding to each load index in this round of task scheduling. Since the size of the weight value depends on the real-time collection of the load index value at the beginning of this round of task scheduling, the real-time performance of the weight value is better. It can better reflect the dynamic performance of each end node in the current edge-end environment.

[0076] For example, after pre-screening in the cluster, there are three remaining candidate end nodes, namely node 1, node 2, and node 3. The user selects 4 load indexes based on the corresponding implementation environment, and can calculate the standard deviation of each index: S1, S2, S3, and S4 through the above formula. After data collection and standardization of the collected data, a 3x4 matrix is formed, where the rows represent different index collection amounts of the same node, and the columns represent the index values corresponding to the same index of different nodes:

[0077]

[0078] Further calculation based on the correlation coefficient formula can obtain the correlation coefficient matrix:

[0079]

[0080] It can be understood that when i is equal to j, the two indexes participating in the correlation coefficient calculation are actually the same index, so for the case where i is equal to j in the correlation coefficient, it can not be calculated, and the item is directly set to 1.

[0081] Further, the weight value corresponding to each index can be calculated based on the above extreme parameters. For example, the calculation of the weight W1 corresponding to the index x1:

[0082]

[0083] S203: based on the objective weight corresponding to each load index, calculate the score of each candidate end node, and select the optimal scheduling node based on the score of each candidate end node to perform task scheduling.

[0084] In an exemplary embodiment of the present application, based on the objective weight corresponding to each load data of the candidate node, the current score of the node is calculated, including: multiplying each load index value of the candidate end node with the weight value corresponding to the load index, and taking the sum of the product of each load index value and the weight value corresponding to the load index as the current score of the candidate end node.

[0085] The corresponding score formula can be expressed as:

[0086]

[0087] wherein Score n represents the score value corresponding to the nth node in this round of scheduling, W i represents the calculated weight value of the corresponding index i, x ni represents the load index value of the nth node for the load index i.

[0088] For example, in the above embodiment, three remaining candidate end nodes, namely node 1, node 2 and node 3, are used in the cluster after pre-screening. The user selects 4 load indexes based on the corresponding implementation environment. It is calculated that the weight values corresponding to each load index are W1, W2, W3 and W4; in the data collection stage, the load index values of node 1 for the 4 indexes are x 11 , x 12 , x 13 and x 14 . Then, taking node 1 as an example, its corresponding score should be:

[0089]

[0090] It should be noted that in the embodiment of the present application, in order to ensure the real-time consideration of the performance index resource utilization of the end node in the application deployment process, avoid uneven load of the cluster, and affect the system performance. Therefore, before each round of task scheduling, a new round of collection of load index values and calculation of load index weight values is needed, and then the optimal scheduling node for the current task to be forwarded in this round is obtained.

[0091] In order to further demonstrate the significant effect of the resource-aware scheduling method for edge cloud environment provided by the present application, the following is a set of control experiments:

[0092] Two clusters are set in the experiment, one cluster uses the original scheduler, and the other cluster uses the resource-aware scheduling method for the edge cloud environment provided by the application. The clusters are deployed on Aliyun servers, and the basic system environment is centos7. For each cluster size, the number of slave nodes is 4 except for the master node.

[0093] In order to realize the periodic creation of tasks, the task based on time interval repeated scheduling is created by selecting CronJob (periodic task), and the time of scheduling process, the space occupation of scheduler and the scheduling result are recorded. In order to measure whether the resource utilization of the cluster is balanced, the experimental evaluation index adopts the variance of the utilization rate of a certain resource on each node of the cluster to show, and the calculation process is as follows:

[0094]

[0095] U ij The utilization rate of resource j on node i is represented by n, and the total number of slave nodes is represented by n, The average utilization rate of resource j is represented by n.

[0096]

[0097] S j The variance of the utilization rate of resource j in the cluster can reflect whether the utilization of resource j is balanced.

[0098] At the same time, the average value of the variance is defined for the overall situation of the resource, to reflect the resource imbalance degree of the whole cluster:

[0099]

[0100] S avg The average value of the variance of the utilization rate of all resources in the cluster is represented by p, which represents the number of indicators. This measure comprehensively considers the fluctuation of each resource in the cluster, and provides a global evaluation of resource balance. The calculated resource imbalance value reflects the overall situation of the whole cluster in terms of resource utilization.

[0101] Through experimental verification in a real edge cloud environment, the scheduling strategy of the application improves the standard deviation of the comprehensive load by 57% compared with the default scheduling algorithm in the case of large cluster load, which shows better resource scheduling effect

[0102] Please refer to Figure 3 Based on the same inventive concept as the aforementioned resource-aware scheduling method for the edge cloud environment, an embodiment of the application provides a resource-aware scheduling system 300 for the edge cloud environment, which includes an acquisition module 301, a processing module 302 and a scheduling module 303.

[0103] Specifically, the collection module 301 is configured to, in response to a task scheduling request, collect values of various load indicators corresponding to the candidate end nodes based on preset load indicators, the load indicators including CPU occupancy, memory occupancy, node task proportion, and temporary storage capacity; the processing module 302 is configured to standardize the values of the load indicators, and generate objective weights corresponding to the various load indicators based on the standardized values of the load indicators; and the scheduling module 303 is configured to calculate scores of the candidate end nodes based on the objective weights corresponding to the various load indicators, and select an optimal scheduling node based on the scores of the candidate end nodes to perform task scheduling.

[0104] Please refer to Figure 4 The electronic device 400 includes at least one processor 401, a memory 402 (e.g., a non-transitory memory), a storage 403, and a communication interface 404, and the at least one processor 401, the memory 402, the storage 403, and the communication interface 404 are connected together via a bus 405. The at least one processor 401 is configured to invoke at least one program instruction stored or encoded in the memory 402, so as to enable the at least one processor 401 to perform various operations and functions of the resource-aware scheduling method for an edge cloud environment described in various embodiments of the present specification.

[0105] In the embodiments of the present specification, the electronic device 400 can include, but is not limited to, a personal computer, a server computer, a workstation, a desktop computer, a laptop computer, a notebook computer, a mobile electronic device, a smart phone, a tablet computer, a cellular phone, a personal digital assistant (PDA), a handheld device, a messaging device, a wearable electronic device, a consumer electronic device, and the like.

[0106] The embodiments of the present application also provide a computer readable medium, and the computer readable medium carries computer execution instructions. When the computer execution instructions are executed by a processor, the computer execution instructions can be used to implement various operations and functions of the resource-aware scheduling method for an edge cloud environment described in various embodiments of the present specification.

[0107] The computer readable medium in the present application can be a computer readable signal medium or a computer readable storage medium or any combination of the two. The computer readable storage medium may, for example, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or apparatus, or any combination of the above. More specific examples of the computer readable storage medium can include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, the computer readable storage medium can be any tangible medium that contains or stores a program that can be used by or in connection with an instruction execution system, apparatus or device.

[0108] In the present application, the computer readable signal medium can include a data signal propagated in a baseband or as a carrier wave in a propagated data signal that can bear computer readable program code. Such propagated data signal can take a variety of forms, including but not limited to electro-magnetic, optical or any suitable combination thereof. The computer readable signal medium can also be any computer readable medium that is not a storage medium and that can communicate, propagate or transport program for use by or in connection with an instruction execution system, apparatus or device. Program code contained in the computer readable medium can be transmitted using any suitable medium, including but not limited to wireless, wire line, optical fiber cable, RF, etc., or any suitable combination of the above.

[0109] Those skilled in the art should clearly understand that the embodiments of the present application can be provided as a method, a system or a computer program product. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Moreover, the present application can take 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.) containing computer usable program code.

[0110] The present application is described with reference to the flowcharts and / or block diagrams of the methods, apparatuses, systems, and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams 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 apparatus to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing apparatus create a means for implementing the functions specified in the flowcharts and / or block diagrams. These computer program instructions can also be stored in a computer readable medium that can guide the computer or other programmable data processing apparatus to work in a specific manner, so that the computer readable medium contains the instructions for implementing the functions specified in the flowcharts and / or block diagrams.Figure 1 one or more processes and / or means for Figure 1 one or more blocks or means for

[0111] The foregoing description of specific exemplary embodiments of the application has been presented for the purposes of illustration and description. It is not intended to be a limitation on the broad features of the application. Obviously, many modifications and variations will be apparent to those skilled in the art in light of this teaching. The description has been presented for the purpose of illustration and example, not limitation. The scope of the application is defined by the appended claims and their equivalents.

[0112] It will be apparent to those skilled in the art that the present application is not limited to the specific details of the foregoing exemplary embodiments and that various changes and modifications can be made thereto without departing from the spirit or scope of the application. Thus, the scope of the application should be determined by the appended claims and their equivalents, not by the above description which is given to enable others skilled in the art to practice the application. Any reference signs in the claims should not be construed as limiting the scope of the claims.

[0113] Furthermore, it is to be understood that the application can be carried out by specifically different embodiments and that specific embodiments of the application can be combined together or combined with the application in other ways. It is therefore clear that only some of the embodiments of the application actually comprise one single independent technical solution. The application is therefore not limited to the embodiments described in the specification, but rather the embodiments described in the specification are intended merely to illustrate the application. The person skilled in the art will be able to devise other embodiments of the application without departing from the scope of the application.

Claims

1. A resource-aware scheduling method for edge cloud environments, characterized in that: include: In response to a task scheduling request, the system collects various load indicator values ​​corresponding to candidate end nodes based on preset load indicators, including CPU occupancy, memory occupancy, node task ratio, and temporary storage margin; The load index value is standardized, and the objective weight corresponding to each load index is generated based on the standardized load index value; the formula for generating the objective weight corresponding to each load data based on the standardized load data is: in, For the The candidate end node corresponds to the The load index value of each load index; For each candidate end node The average value of load indicators; is the total number of candidate end nodes; Indicates the Alternative end node Each indicator corresponds to the normalized load indicator value; Indicates the The average value after standardization of the indicators, is the number of indicators that make up the load data, For the The weight value corresponding to each load indicator, For the The load index and The correlation coefficient of the load indicators; Based on the objective weights corresponding to the various load indicators, the scores of the candidate end nodes are calculated, and the optimal scheduling node is selected to perform task scheduling; Based on the objective weights corresponding to the load data of the candidate end node, the current score of the node is calculated, including: multiplying the load index values ​​of the candidate end node by the weight values ​​corresponding to the load index, and taking the sum of the products of the load index values ​​and the weight values ​​corresponding to the load index as the current score of the candidate end node.

2. The resource-aware scheduling method for edge cloud environments according to claim 1, characterized in that: The method further comprises: Obtain the demand load index value of the task to be scheduled; If the load index value of the end node meets the required load index value, the end node is listed as a candidate end node; If the load index value of the end node does not meet the required load index value, the end node is filtered so that the end node that does not meet the required load index value does not participate in the current task scheduling.

3. The resource-aware scheduling method for edge cloud environments according to claim 1, characterized in that: The formula for normalizing the load index value is: in, For the The candidate end node corresponds to the The load index value of the load index, For each candidate end node The maximum value of the load index, For each candidate end node The minimum value of the load indicator.

4. The resource-aware scheduling method for edge cloud environments according to claim 1, characterized in that: The method further comprises: A monitoring component is deployed on the edge node to monitor and collect the load indicator values ​​corresponding to the candidate end nodes.

5. A resource-aware scheduling system for an edge cloud environment, applying the resource-aware scheduling method for an edge cloud environment as described in any one of claims 1 to 4, characterized in that: include: A collection module is used to collect various load indicator values ​​corresponding to the candidate end node in response to the task scheduling request based on the preset load indicators, wherein the load indicators include CPU occupancy, memory occupancy, node task ratio, and temporary storage margin; a processing module, configured to normalize the load index value and generate an objective weight corresponding to each load index based on the normalized load index value; The scheduling module is used to calculate the scores of the candidate end nodes based on the objective weights corresponding to the various load indicators, and select the optimal scheduling node to perform task scheduling.

6. The resource-aware scheduling system for edge cloud environments according to claim 5, characterized in that: The system is also used to: Obtain the demand load index value of the task to be scheduled; If the load index value of the end node meets the required load index value, the end node is listed as a candidate end node; If the load index value of the end node does not meet the required load index value, the end node is filtered so that the end node that does not meet the required load index value does not participate in the current task scheduling.

7. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the resource-aware scheduling method for an edge cloud environment according to any one of claims 1 to 4 by executing the computer instructions.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the resource-aware scheduling method for edge cloud environments according to any one of claims 1 to 4.

Images