Edge cloud computing service computing power perception scheduling method and device, medium and equipment

Abstract

This application relates to a computing power-aware scheduling method for edge cloud computing services, applied to edge devices. The method includes: a virtual computing power agent deployed on the edge device acquiring in real-time values ​​of multiple key factors affecting the device's service capabilities; calculating the current service's computing power consumption and the system's available computing power based on the values ​​of these key factors; determining and evaluating service capabilities based on the current service's computing power consumption and the system's available computing power; and sending the evaluated service capabilities to a scheduling system, so that the scheduling system can schedule and allocate services according to the evaluated service capabilities. This method can rationally allocate services to available devices, maximizing resource utilization efficiency.

Description

Technical Field

[0001] This application relates to the field of edge computing, and in particular to a method, apparatus, medium and device for edge cloud computing service computing power awareness scheduling. Background Technology

[0002] In related technologies, Content Delivery Networks (CDNs) and edge cloud systems, as distributed cloud platforms, are dedicated to providing fast and reliable content delivery and edge computing services. To effectively manage and allocate server resources located around the world, mainstream scheduling schemes typically consider bandwidth utilization optimization: by monitoring the remaining bandwidth of servers and allocating tasks accordingly, traffic can be routed to servers with higher remaining bandwidth, achieving more efficient bandwidth utilization and providing faster and more stable content delivery.

[0003] However, server-based methods for scheduling remaining bandwidth have the following problems:

[0004] Resource utilization: Scheduling based on server's remaining bandwidth can lead to unbalanced resource utilization. Some servers may have more remaining bandwidth, and allocating traffic to these servers based on their remaining bandwidth can cause system server overload. This could result in some servers being under low load while others are under high load.

[0005] Service Priority: Different services may have different priorities and importance. Server-based remaining bandwidth scheduling may not fully consider service priorities, resulting in some important services being restricted or affected.

[0006] Increased scheduling complexity: As business scenarios become more complex, managing and optimizing server-based remaining bandwidth scheduling can become more intricate and challenging. It requires comprehensive consideration of the needs and constraints of multiple services to configure and optimize appropriate scheduling strategies.

[0007] To address these issues, a new resource scheduling method is needed to improve scheduling performance and efficiency, better balance server resource utilization, consider business priorities, and provide higher-level scheduling strategy configuration and optimization. Summary of the Invention

[0008] To overcome the problems existing in related technologies, this application provides a method, apparatus, medium and equipment for edge cloud computing service computing power awareness scheduling.

[0009] According to a first aspect of this application, an edge cloud computing service computing power-aware scheduling method is provided, applied to edge devices, comprising:

[0010] Virtual computing agents deployed on edge devices acquire real-time values ​​of multiple key factors affecting the device's service capabilities;

[0011] The computing power consumption and available computing power of the system for the current business are calculated based on the values ​​of the aforementioned key factors.

[0012] The service capability is assessed based on the computing power consumption of the current business and the available computing power of the system.

[0013] The evaluation service capabilities are sent to the scheduling system so that the scheduling system can schedule and allocate services based on the evaluation service capabilities.

[0014] In some embodiments of this application, based on the aforementioned scheme, the values ​​of multiple key factors affecting the service capabilities of the device include CPU utilization, disk I / O utilization, memory utilization, load quantity, response time, and bandwidth usage. The calculation of the current service's computing power consumption and the system's available computing power based on the values ​​of these multiple key factors includes:

[0015] Set a weighting coefficient for each key factor;

[0016] Set the maximum computing power of the device;

[0017] The sum of the products of the magnitude and weight coefficient of each key factor obtained before the equipment processes the business is determined as the initial computing power consumption;

[0018] The sum of the products of the value of each key factor and its weight coefficient when the device processes the business is determined as the current computing power consumption;

[0019] The difference between the current computing power consumption and the initial computing power consumption is determined as the computing power consumption of the current service.

[0020] The difference between the device's maximum computing power and its current computing power consumption is determined as the system's available computing power.

[0021] In some embodiments of this application, based on the foregoing scheme, setting weight coefficients for each key factor further includes:

[0022] When the value of a key factor is less than a preset threshold, the weighting coefficient is the first weighting coefficient; when the value of a key factor is greater than or equal to the preset threshold, the weighting coefficient is the second weighting coefficient; wherein the second weighting coefficient is greater than the first weighting coefficient.

[0023] In some embodiments of this application, based on the foregoing scheme, determining the evaluation service capability based on the current business's computing power consumption and the system's available computing power includes:

[0024] When the available computing power of the system is greater than the first preset threshold, it is determined that the evaluation service capacity is greater than the computing power consumption of the current business.

[0025] When the available computing power of the system is less than the first preset threshold, it is determined that the evaluation service capability is less than the computing power consumption of the current business, until the available computing power of the system is equal to the first preset threshold.

[0026] In some embodiments of this application, based on the foregoing scheme, determining the service capability assessment based on the current business's computing power consumption and the system's available computing power further includes:

[0027] When the available computing power of the system is greater than the second preset threshold, it is determined that the evaluation service capacity is greater than the computing power consumption of the current business.

[0028] Gradually increase the evaluation service capability until the available computing power of the system is less than or equal to a third preset threshold.

[0029] In subsequent weeks, if the computing power consumption of the current service is less than the evaluation service capacity of the previous week, the evaluation service capacity is reduced; if the computing power consumption of the current service is greater than the evaluation service capacity of the previous week, the evaluation service capacity is increased; until the computing power consumption of the current service equals the evaluation service capacity, the evaluation service capacity at this time is determined to be the optimal evaluation service capacity.

[0030] In some embodiments of this application, based on the aforementioned scheme, when the available computing power of the system is less than a fourth preset threshold, a request to evict the current service is sent to the scheduling system.

[0031] According to another aspect of this application, a computing power-aware scheduling method for edge cloud computing services is provided, applied to a scheduling system, comprising:

[0032] The system receives evaluation service capabilities of edge devices from virtual computing agents deployed on edge devices, and schedules the evaluation of available devices and their service capabilities within the system.

[0033] Service scheduling is performed based on the available devices and their service capabilities.

[0034] In some embodiments of this application, based on the aforementioned scheme, the edge cloud computing service computing power awareness scheduling method further includes: receiving a request from an edge device to evict the current service, re-determining available devices, and scheduling the service in the edge device to other available devices.

[0035] According to another aspect of this application, an edge cloud computing service computing power awareness scheduling device is provided, applied to edge devices, comprising:

[0036] The value acquisition module is used to acquire the values ​​of multiple key factors that affect the service capabilities of the equipment in real time.

[0037] The calculation module is used to calculate the computing power consumption of the current business and the available computing power of the system based on the values ​​of the multiple key factors.

[0038] The evaluation module is used to determine the evaluation service capability based on the computing power consumption of the current service and the available computing power of the system;

[0039] The reporting module is used to send the evaluation service capabilities to the scheduling system, so that the scheduling system can schedule and allocate services according to the evaluation service capabilities.

[0040] In some embodiments of this application, based on the foregoing scheme, the values ​​of the multiple key factors affecting the device's service capabilities include CPU utilization, disk I / O utilization, memory utilization, load quantity, response time, and bandwidth usage. The calculation module is further used for:

[0041] Set a weighting coefficient for each key factor;

[0042] Set the maximum computing power of the device;

[0043] The sum of the products of the magnitude and weight coefficient of each key factor obtained before the equipment processes the business is determined as the initial computing power consumption;

[0044] The sum of the products of the value of each key factor and its weight coefficient when the device processes the business is determined as the current computing power consumption;

[0045] The difference between the current computing power consumption and the initial computing power consumption is determined as the computing power consumption of the current service.

[0046] The difference between the device's maximum computing power and its current computing power consumption is determined as the system's available computing power.

[0047] In some embodiments of this application, based on the foregoing scheme, the evaluation module is further configured to:

[0048] When the available computing power of the system is greater than a first preset threshold, it is determined that the evaluation service capability is greater than the computing power consumption of the current business.

[0049] When the available computing power of the system is less than a first preset threshold, it is determined that the evaluation service capability is less than the computing power consumption of the current business, until the available computing power of the system equals the first preset threshold.

[0050] In some embodiments of this application, based on the foregoing scheme, the evaluation module is further configured to:

[0051] When the available computing power of the system is greater than the second preset threshold, it is determined that the evaluation service capacity is greater than the computing power consumption of the current business.

[0052] Gradually increase the evaluation service capability until the available computing power of the system is less than or equal to a third preset threshold.

[0053] In subsequent weeks, if the computing power consumption of the current service is less than the evaluation service capacity of the previous week, the evaluation service capacity is reduced; if the computing power consumption of the current service is greater than the evaluation service capacity of the previous week, the evaluation service capacity is increased; until the computing power consumption of the current service equals the evaluation service capacity, the evaluation service capacity at this time is determined to be the optimal evaluation service capacity.

[0054] In some embodiments of this application, based on the foregoing scheme, the reporting module is further configured to:

[0055] When the available computing power of the system is less than the fourth preset threshold, a request to evict the current service is sent to the scheduling system.

[0056] According to another aspect of this application, an edge cloud computing service computing power awareness scheduling device is provided, applied to a scheduling system, comprising:

[0057] The receiving module is used to receive the evaluation service capabilities of the edge devices sent by the virtual computing power agents deployed on the edge devices, and to schedule the available devices and service capabilities of the available devices in the evaluation system.

[0058] The scheduling module is used to schedule services based on the available devices and their service capabilities.

[0059] In some embodiments of this application, based on the foregoing scheme, the scheduling module is further configured to:

[0060] Upon receiving a request from an edge device to evict the current service, a new available device is determined, and the service in the edge device is rescheduled to another available device.

[0061] According to another aspect of this application, a computer-readable storage medium is provided, on which a computer program is stored, wherein when the computer program is executed, the steps of an edge cloud computing service computing power awareness scheduling method are implemented.

[0062] According to another aspect of this application, a computer device is provided, including a processor, a memory, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of an edge cloud computing service computing power awareness scheduling method.

[0063] This application deploys virtual computing power agents on edge devices. These agents acquire values ​​of multiple key factors affecting the device's service capabilities. Based on these values, they determine the current service's computing power consumption, the system's available computing power, and further assess the service capability. This assessed service capability is then sent to the scheduling system, allowing the system to allocate services based on each device's assessed service capability. A higher reported assessed service capability indicates that the device can provide more services, maximizing resource utilization efficiency by rationally allocating services to available devices. The virtual computing power agents calculate the current service's computing power consumption and the system's available computing power in real time. Therefore, devices can adjust their assessed service capabilities promptly based on actual conditions. The scheduling system can dynamically schedule and allocate services based on real-time system load information and service computing power requirements. Assessing service capabilities based on the system's available computing power ensures that services do not exceed the performance limits of a single machine, contributing to service stability and reliability and preventing performance degradation or failures caused by single-machine overload. The scheduling system enables more balanced, flexible, and efficient resource allocation, improving the CDN platform's service quality and user satisfaction.

[0064] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0065] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0066] Figure 1 This is a flowchart illustrating an edge cloud computing service computing power-aware scheduling method according to an exemplary embodiment.

[0067] Figure 2 This is a schematic diagram illustrating the determination of optimal assessment service capabilities according to an exemplary embodiment.

[0068] Figure 3 This is a flowchart illustrating an edge cloud computing service computing power-aware scheduling method according to an exemplary embodiment.

[0069] Figure 4 This is a block diagram of an edge cloud computing service computing power awareness scheduling device according to an exemplary embodiment.

[0070] Figure 5 This is a block diagram of an edge cloud computing service computing power awareness scheduling device according to an exemplary embodiment.

[0071] Figure 6 This is a block diagram illustrating a computer device for computing power-aware scheduling of edge cloud computing services, according to an exemplary embodiment. Detailed Implementation

[0072] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of this application can be arbitrarily combined with each other.

[0073] In related technologies, Content Delivery Networks (CDNs) and edge cloud systems, as distributed cloud platforms, are dedicated to providing fast and reliable content delivery and edge computing services. To effectively manage and allocate server resources located globally, mainstream scheduling schemes typically consider bandwidth utilization optimization: by monitoring the remaining bandwidth of servers and allocating tasks accordingly, traffic is scheduled to servers with higher remaining bandwidth. However, server-based remaining bandwidth scheduling methods suffer from problems such as uneven resource utilization, inability to schedule traffic according to business priorities, and increased scheduling complexity.

[0074] To address the problems existing in the prior art, this application provides a computing power-aware scheduling method for edge cloud computing services.

[0075] Figure 1 This is a flowchart illustrating an edge cloud computing service computing power-aware scheduling method according to an exemplary embodiment. (Reference) Figure 1 The edge cloud computing service computing power awareness scheduling method is applied to edge devices and includes at least steps S11-S14.

[0076] Step S11: The virtual computing agent deployed on the edge device acquires the values ​​of multiple key factors affecting the device's service capabilities in real time.

[0077] To implement the edge cloud computing power service-aware scheduling method provided in this application, a virtual computing power agent needs to be deployed on the edge device. This agent is responsible for monitoring various key factors of the local machine, such as CPU, I / O, bandwidth, and consumed system resources. The virtual computing power agent can obtain relevant system and resource data in real time through integration with the device's operating system or other components.

[0078] Based on the characteristics of edge cloud computing services, identify the key factors affecting device service capabilities in advance. These key factors may include device memory usage, disk I / O capacity, CPU utilization, and available bandwidth.

[0079] After identifying the key factors affecting the equipment's service capabilities, the virtual computing agent obtains the values ​​of these factors in real time, such as CPU utilization, disk I / O utilization, memory utilization, load quantity, response time, and bandwidth usage.

[0080] Step S12: Calculate the computing power consumption of the current business and the available computing power of the system based on the values ​​of multiple key factors.

[0081] Based on several key factors affecting the service capabilities of the equipment, a weighted average algorithm is designed to calculate the computing power consumption of the equipment's current business and the available computing power of the system.

[0082] Step S13: Determine the evaluation service capability based on the current computing power consumption of the business and the available computing power of the system.

[0083] The assessment service capability is determined based on the current computing power consumption of the business and the available computing power of the system. The assessment service capability is the evaluation value of the computing power that the system can still provide for the business under the current conditions.

[0084] Step S14: Send the evaluation service capabilities to the scheduling system so that the scheduling system can schedule and allocate services based on the evaluation service capabilities.

[0085] The scheduling system determines how much traffic to allocate to an edge device based on the assessed service capabilities reported by the edge device, so that the edge device can provide services within its capacity and prevent device overload.

[0086] Since the assessment of service capabilities is based on multiple key factors that affect the service capabilities of devices, it comprehensively considers multiple different influencing factors. Compared with traditional scheduling methods that only schedule based on remaining bandwidth, it can achieve optimal utilization of edge resources and balanced allocation of services, thereby improving the overall performance and efficiency of the system.

[0087] In an exemplary embodiment, the values ​​of several key factors affecting the device's service capabilities include CPU utilization, disk I / O utilization, memory utilization, load quantity, response time, and bandwidth usage. Step S12, calculating the computing power consumption of the current service and the available computing power of the system based on the values ​​of these key factors, includes:

[0088] Set a weighting coefficient for each key factor;

[0089] Set the maximum computing power of the device;

[0090] The sum of the products of the magnitude and weight coefficient of each key factor obtained before the equipment processes the business is determined as the initial computing power consumption;

[0091] The sum of the products of the value of each key factor and its weight coefficient when the device processes the business is determined as the current computing power consumption;

[0092] The difference between the current computing power consumption and the initial computing power consumption is determined as the computing power consumption of the current service.

[0093] The difference between the device's maximum computing power and its current computing power consumption is determined as the system's available computing power.

[0094] In this embodiment, the values ​​of several key factors affecting the device's service capabilities include CPU utilization, disk I / O utilization, memory utilization, workload, response time, and bandwidth usage. These key factors include not only the utilization of the edge device's system resources (CPU, disk I / O, memory) but also service-related factors (workload, response time, and bandwidth usage).

[0095] Assign weight coefficients to each key factor. Because different key factors have varying impacts on service capabilities, the weight coefficients for different key factors can be determined based on actual needs, past experience, historical data, and other relevant factors.

[0096] For example, set the weight coefficient for CPU utilization as a, the weight coefficient for disk I / O utilization as b, the weight coefficient for memory utilization as c, the weight coefficient for the number of loads as d, the weight coefficient for response time as e, and the weight coefficient for bandwidth usage as f.

[0097] Set the maximum computing power of the device. The maximum computing power of the device can be set based on experience or the hardware configuration of different devices. For ease of understanding, in this embodiment, the maximum computing power of the device is uniformly set to 100.

[0098] The sum of the products of the magnitude and weight coefficient of each key factor obtained before the device processes a task is determined as the initial computing power consumption. Before processing a task, the edge device needs to be started, along with related system software and application software, all of which consume a portion of the edge device's computing resources. The computing power consumed by the edge device due to system software and application software before processing the task is considered as the initial computing power consumption.

[0099] For example, computing power consumption = CPU utilization × a + disk I / O utilization × b + memory utilization × c + number of loads × d + response time × e + bandwidth usage × f.

[0100] When calculating the initial computing power consumption, it is assumed that CPU utilization = 0.2, disk I / O utilization = 0.1, memory utilization = 0.2, load quantity = 0, response time = 0, bandwidth usage = 0, and weight coefficients a = b = c = d = e = f = 10. The calculated initial computing power consumption at this time is 5. In this embodiment, for ease of calculation, the weight coefficients are uniformly set to 10. In practical applications, the weight coefficients can be set according to the actual situation. The weight coefficients of different key factors can be the same or different, and this application does not impose any restrictions.

[0101] After the device receives the edge computing task, the CPU utilization is 0.5, disk I / O utilization is 0.6, memory utilization is 0.8, the number of loads is 5, the response time is 2, the bandwidth usage is 0.1, and a = b = c = d = e = f = 10. The current computing power consumption is calculated to be 10 × 0.5 + 10 × 0.6 + 10 × 0.8 + 10 × 5 + 10 × 2 + 10 × 0.1 = 90.

[0102] The current computing power consumption is the difference between the current computing power consumption and the initial computing power consumption, i.e., 90-5=85.

[0103] The available computing power of the system is the difference between the maximum computing power of the device and the current computing power consumption, i.e., 100-90=10.

[0104] From the above calculations, it's clear that if the weights of each key factor remain constant, and if the system resources corresponding to a certain key factor are exhausted (e.g., memory utilization reaches 100%), while the utilization, quantity, or duration of other key factors are normal, the calculated current computing power consumption is not significant, and the system's available computing power remains relatively high. In other words, based on the system's available computing power, the device can still provide more services. However, in reality, the device's memory resources are exhausted, and it can no longer provide services. Of course, the same problem exists for other key factors. Once the value of a certain key factor reaches a threshold, this key factor becomes the primary factor restricting the device's service capacity.

[0105] To ensure that the calculated available computing power of the system reasonably reflects the actual situation of the device, in an exemplary embodiment, setting weight coefficients for each key factor further includes:

[0106] When the value of a key factor is less than a preset threshold, the weighting coefficient is the first weighting coefficient; when the value of a key factor is greater than or equal to the preset threshold, the weighting coefficient is the second weighting coefficient; the second weighting coefficient is greater than the first weighting coefficient. When the value of a key factor is less than the preset threshold, the smaller first weighting coefficient is used; when the value of a key factor is greater than or equal to the preset threshold, the larger second weighting coefficient is used. This fully reflects the actual impact of each key factor. For example, taking bandwidth usage as an example, when bandwidth usage is normal and available bandwidth is sufficient, the first weighting coefficient is 10; when bandwidth usage is greater than or equal to 95%, the second weighting coefficient is 100. Even if the device receives 0 edge computing tasks at this time, and the system's available computing power is calculated to be 0, the scheduling system will no longer allocate traffic to that device to prevent its bandwidth from being exhausted or exceeding its available bandwidth, thus increasing costs. For example, when the memory utilization rate is less than 90%, the first weighting coefficient is 10; when the memory utilization rate is greater than or equal to 90%, the second weighting coefficient is 110. Once the memory utilization rate of the device reaches 90% during the service process, the computing power consumption of the current business is calculated to be greater than 100, and the available computing power of the system is less than 0. At this time, it can be considered that the available computing power of the system of the device is 0, and it can no longer provide services to prevent the device from being overloaded.

[0107] By setting a first weight coefficient and a second weight coefficient for each key factor, the contribution of the value of the key factor to the current computing power consumption is small when the resources corresponding to the key factor are sufficient, and large when the resources corresponding to the key factor are scarce. This results in the available computing power of the computing system being 0, so that the scheduling system no longer schedules traffic to that device. On the one hand, this can prevent device overload, and on the other hand, it can also prevent a large number of requests from accumulating on a certain device and affecting the response speed.

[0108] In an exemplary embodiment, step S13, determining the service capability assessment based on the current service's computing power consumption and the system's available computing power includes:

[0109] When the available computing power of the system is greater than the first preset threshold, it is determined that the evaluation service capacity is greater than the computing power consumption of the current business.

[0110] When the available computing power of the system is less than the first preset threshold, it is determined that the evaluation service capability is less than the computing power consumption of the current business, until the available computing power of the system is equal to the first preset threshold.

[0111] For example, assuming the first threshold is 30, in the first cycle, the device receives 6 computing tasks, the device's current computing power consumption is 30, and the system's available computing power is 65. The system's available computing power is greater than the first threshold, so the assessed service capability is determined to be greater than the current computing power consumption, assumed to be 60. Based on the assessed service capability reported by the device, the scheduling system determines to schedule 12 computing tasks to the device in the next cycle. In the second cycle, the device's current computing power consumption is 60, and the system's available computing power is 35, both greater than the first threshold. The device determines that the assessed service capability is greater than the current computing power consumption in this cycle, and sets the assessed service capability to 65. In the third cycle, based on the device's reported assessed service capability of 65, the scheduling system schedules 14 computing tasks to the device. In the third cycle, the device's current computing power consumption is 65, and the system's available computing power is 30, equal to the first preset threshold. Thereafter, the device can consistently determine the assessed service capability to be 65, allowing the scheduling system to continuously schedule 12 computing tasks to the device in subsequent cycles, enabling the device to continuously provide high-quality computing services.

[0112] For example, suppose that in the fourth cycle, the device's current computing power consumption is 75, and the system's available computing power is 20, which is less than the first preset threshold. Therefore, the assessed service capacity is determined to be less than 75, and is assumed to be 60. In the fifth cycle, based on the device's reported assessed service capacity of 65, the scheduling system schedules 14 computing tasks to the device. The device's current computing power consumption is 65, and the system's available computing power is 30.

[0113] By setting a reasonable first preset threshold for the system's available computing power, the device can always provide services based on sufficient available computing power, enabling it to quickly complete computing tasks, ensure timely service, and improve user experience.

[0114] In the above embodiments, a first preset threshold for the system's available computing power is set, enabling the server to provide services near the first preset threshold. However, since different devices have different processing capabilities, and different services consume different amounts of computing power, the method of setting a first preset threshold for the system's available computing power has the following problems: if the first preset threshold is set too high, the device's maximum performance will not be utilized, resulting in some computing resources being idle. If the first preset threshold is set too low, the device may always be working under overload.

[0115] In one exemplary embodiment, determining the service capability assessment based on the computing power consumption of the current service and the available computing power of the system further includes:

[0116] When the available computing power of the system is greater than the second preset threshold, it is determined that the evaluation service capacity is greater than the computing power consumption of the current business.

[0117] Gradually increase the evaluation service capacity until the system's available computing power is less than or equal to the third preset threshold;

[0118] In subsequent cycles, if the computing power consumption of the current service is less than the evaluation service capacity of the previous cycle, the evaluation service capacity is reduced; if the computing power consumption of the current service is greater than or equal to the evaluation service capacity of the previous cycle, the evaluation service capacity is increased; until the computing power consumption of the current service equals the evaluation service capacity, the evaluation service capacity at this time is determined to be the optimal evaluation service capacity.

[0119] For example, Figure 2 This is a schematic diagram illustrating the determination of optimal evaluation service capabilities according to an exemplary embodiment. Figure 2 In the graph, the horizontal axis represents the number of scheduling cycles, and the vertical axis represents the computing power consumption of the current service. The square brackets in the graph contain three values, from left to right: the computing power consumption of the current service, the assessed service capacity, and the system's available computing power. For example, the point corresponding to cycle 1 in the graph represents the initial stage, where the edge device has just started up and is waiting to receive edge computing tasks. The initial computing power consumption is 5, the current service's computing power consumption is 0, and the system's available computing power is 95.

[0120] Assume the second preset threshold is 50 and the third preset threshold is 1. At this point, the system's available computing power is greater than the second preset threshold of 50, and the assessed service capacity is set to 30 to allow the scheduling system to quickly allocate computing tasks to edge devices. In subsequent scheduling cycles, the assessed service capacity is gradually increased until the system's available computing power is less than or equal to the third preset threshold of 1 (cycle 1-cycle 8).

[0121] In subsequent cycles, if the computing power consumption of the current business is less than the assessed service capacity of the previous cycle, the assessed service capacity is reduced. For example, in cycle 9, the current business's computing power consumption is 89, and the assessed service capacity of cycle 8 is 95. Since the current business's computing power consumption is less than the assessed service capacity of the previous cycle, the assessed service capacity is reduced to 85. In cycle 10, the current business's computing power consumption is 79, and the assessed service capacity of the previous cycle is 85, so the assessed service capacity for the next cycle is determined to be 60.

[0122] If the computing power consumption of the current business is greater than the assessment service capacity of the previous period, the assessment service capacity is increased. In period 11, the computing power consumption of the current business is 66, which is greater than the assessment service capacity of the previous period is 60. The assessment service capacity of the next period is determined to be 73.

[0123] Similarly, in cycle 12, the computing power consumption of the previous service (75) is greater than the evaluation service capacity (70) of the previous cycle, so the evaluation service capacity is increased to 74. In cycle 13, the computing power consumption of the previous service (69) is less than the evaluation service capacity (74) of the previous cycle, so the evaluation capacity is decreased to 72. Finally, in cycle 14, the computing power consumption of the previous service (72) is equal to the evaluation computing power (72) of the previous cycle. Therefore, the evaluation service capacity of 72 at this point is determined to be the optimal evaluation service capacity, and this optimal evaluation service capacity will be maintained at 72 in subsequent cycles.

[0124] The optimal assessment service capability is the maximum computing power that the equipment can provide. This allows the subsequent scheduling system to schedule tasks based on the optimal assessment service capability, enabling the equipment to perform at its maximum capacity while ensuring that tasks can be completed in a timely manner.

[0125] In one exemplary embodiment, when the available computing power of the system is less than a fourth preset threshold, a request to evict the current service is sent to the scheduling system.

[0126] If a critical factor of an edge device suddenly reaches a preset threshold, such as response time, the second weighting coefficient of response time is used when calculating the device's available computing power. The result is that the edge device's available computing power is 0, and its evaluated computing power is also 0. The edge device then sends a request to the scheduling system to evict the current service. At this point, the scheduling system re-determines available devices and reschedules the service from the edge device to other available devices. This prevents users from waiting for extended periods and ensures that services are processed promptly.

[0127] Figure 3 This is a flowchart illustrating an edge cloud computing service computing power-aware scheduling method according to an exemplary embodiment. (Reference) Figure 3 The edge computing service computing power-aware scheduling method is applied to the scheduling system, including:

[0128] Step S31: Receive the evaluation service capabilities of the edge device sent by the virtual computing agent deployed on the edge device, and schedule the system to evaluate the available devices and service capabilities of the available devices in the system.

[0129] Step S32: Perform service scheduling based on available devices and their service capabilities.

[0130] The scheduling system receives assessments of the service capabilities of edge devices from virtual computing agents deployed on those devices. If the assessed service capability is less than a preset value, the edge device is considered unavailable. If the assessed service capability is greater than or equal to the preset value, the edge device is considered available. The higher the assessed service capability of an available edge device, the stronger its service capability. Based on the assessments of the service capabilities reported by each edge device, the scheduling system evaluates the available devices in the system and their service capabilities.

[0131] The scheduling system can determine the number of tasks for different services corresponding to the assessment service capabilities of edge devices, as well as the computing power required for different services, based on historical data. Therefore, the scheduling system can compare the computing power required for the service with the assessment service capabilities of the edge devices, based on available devices and their service capabilities. If the assessment service capabilities of a device are sufficient to meet the service requirements, the service is directly assigned to that device; if the assessment service capabilities of a device are insufficient to meet the service requirements, the service needs to be assigned to other devices with sufficient assessment service capabilities.

[0132] Because the virtual computing power agents of edge devices report and evaluate service capabilities in real time, the scheduling system performs scheduling based on the latest evaluated service capabilities of the edge devices in each cycle. This allows for dynamic monitoring of changes in system load information and business computing power demands during business operations, enabling timely scheduling and allocation. Under normal circumstances, it ensures that business tasks are scheduled to devices within their available system computing power range, guaranteeing timely task processing and response from edge devices, preventing device overload, and improving scheduling efficiency.

[0133] In an exemplary embodiment, the edge computing service computing power-aware scheduling method further includes: receiving a request from an edge device to evict the current service, re-determining available devices, and scheduling the service in the edge device to other available devices.

[0134] If a critical factor of an edge device reaches a preset threshold, such as bandwidth usage reaching a preset threshold and no remaining bandwidth resources are available, the edge device's available computing power and assessed computing power will both be 0. The edge device will then send a request to the scheduling system to evict the current service. At this point, the scheduling system will re-determine available devices and reschedule the service from the edge device to other available devices, ensuring that the service can be processed in a timely manner.

[0135] To better understand the edge cloud computing service computing power awareness scheduling method provided in this application, specific embodiments are described below.

[0136] Deploy virtual computing agents on edge devices in cloud computing systems to collect values ​​of multiple key factors affecting device service capabilities, including CPU utilization, disk I / O utilization, memory utilization, load quantity, response time, and bandwidth usage.

[0137] Set a first weighting coefficient and a second weighting coefficient for each key factor.

[0138] After the device is started, the virtual computing agent obtains the values ​​of several key factors in real time.

[0139] Using the formula: Computing power consumption = CPU utilization × a + Disk I / O utilization × b + Memory utilization × c + Number of loads × d + Response time × e + Bandwidth usage × f, determine the initial computing power consumption, current computing power consumption, current service computing power consumption, and available system computing power.

[0140] If the values ​​of each key factor are within the normal range, i.e., less than the preset threshold, the virtual computing power proxy edge calculates the current computing power consumption, the current business computing power consumption, and the system's available computing power based on the first weight coefficient of each key factor. In each cycle, the edge device determines its assessed service capabilities and sends this assessment to the scheduling system. The scheduling system then schedules tasks to the edge device based on its assessed service capabilities.

[0141] After multiple cycles, the edge devices determine their optimal service capabilities and send them to the scheduling system. The scheduling system then uses these optimal capabilities to schedule tasks for the edge devices, enabling them to provide edge services with the maximum computing power they can offer and maximizing their efficiency.

[0142] If, during a given period, the value of any one of the key factors reaches a preset threshold, the weighting coefficient corresponding to that key factor is adjusted to a second weighting coefficient. The virtual computing agent can then promptly calculate that the available computing power of the edge device is less than a fourth preset threshold and send a request to the scheduling system to evict the current service. This indicates that the edge device can no longer provide edge services, and the scheduling system can then reassign available devices to reschedule the services from that edge device to other available devices. This prevents computing service interruptions and ensures timely responses to users.

[0143] According to the foregoing embodiments, the edge cloud computing service computing power-aware scheduling method provided in this application can better utilize the computing resources of edge devices, rationally allocate services to available devices, and maximize resource utilization. The virtual computing power agent of the edge device can determine and evaluate service capabilities in each scheduling cycle and send them to the scheduling system. The scheduling system can make real-time resource adjustments based on actual conditions to adapt to constantly changing business needs and system load. Scheduling based on the remaining computing power of the service ensures that the service does not exceed the performance limits of a single machine, guaranteeing the stability and reliability of the service and avoiding performance degradation or failures caused by single-machine overload. This improves the overall performance and efficiency of the system, enabling services to receive better support and response.

[0144] Figure 4 This is a block diagram illustrating an edge cloud computing service computing power-aware scheduling device according to an exemplary embodiment. (Reference) Figure 4 The edge computing service computing power perception and scheduling device is applied to edge devices and includes: a value acquisition module 401, a calculation module 402, an evaluation module 403, and a reporting module 404.

[0145] The value acquisition module 401 is configured to acquire values ​​of multiple key factors that affect the service capabilities of the device in real time.

[0146] The calculation module 402 is configured to calculate the computing power consumption of the current business and the available computing power of the system based on the values ​​of multiple key factors.

[0147] The assessment module 403 is configured to determine the assessment service capability based on the computing power consumption of the current business and the available computing power of the system.

[0148] The reporting module 404 is configured to send an evaluation of service capabilities to the scheduling system so that the scheduling system can schedule and allocate services based on the evaluation of service capabilities.

[0149] In one exemplary embodiment, the computing module 402 is further configured to:

[0150] Set a weighting coefficient for each key factor.

[0151] Set the maximum computing power of the device.

[0152] The sum of the products of the magnitude and weight coefficient of each key factor obtained before the equipment processes the business is determined as the initial computing power consumption.

[0153] The sum of the products of the value of each key factor and its weight coefficient when the device processes the business is determined as the current computing power consumption.

[0154] The difference between the current computing power consumption and the initial computing power consumption is determined as the computing power consumption of the current service.

[0155] The difference between the device's maximum computing power and its current computing power consumption is determined as the system's available computing power.

[0156] In one exemplary embodiment, the evaluation module 403 is further configured to:

[0157] When the available computing power of the system is greater than the first preset threshold, it is determined that the evaluation service capacity is greater than the computing power consumption of the current business.

[0158] When the available computing power of the system is less than or equal to the first preset threshold, it is determined that the evaluation service capability is less than or equal to the computing power consumption of the current business.

[0159] In one exemplary embodiment, the evaluation module 403 is further configured to:

[0160] When the available computing power of the system is greater than the second preset threshold, it is determined that the evaluation service capacity is greater than the computing power consumption of the current business.

[0161] Gradually increase the evaluation service capability until the available computing power of the system is less than or equal to a third preset threshold.

[0162] In subsequent weeks, if the computing power consumption of the current service is less than the evaluation service capacity of the previous week, the evaluation service capacity is reduced; if the computing power consumption of the current service is greater than the evaluation service capacity of the previous week, the evaluation service capacity is increased; until the computing power consumption of the current service equals the evaluation service capacity, the evaluation service capacity at this time is determined to be the optimal evaluation service capacity.

[0163] In one exemplary embodiment, the reporting module 404 is further configured to:

[0164] When the available computing power of the system is less than the fourth preset threshold, a request to evict the current service is sent to the scheduling system.

[0165] Figure 5 This is a block diagram illustrating an edge cloud computing service computing power-aware scheduling device according to an exemplary embodiment. (Reference) Figure 5 The edge computing computing power perception and scheduling device is applied to the scheduling system and includes: a receiving module 501 and a scheduling module 502.

[0166] The receiving module 501 is configured to receive the evaluation service capabilities of the edge devices sent by the virtual computing power agent deployed on the edge devices, and to schedule the available devices and service capabilities of the available devices in the system.

[0167] The scheduling module 502 is configured to perform service scheduling based on available devices and the service capabilities of available devices.

[0168] In one exemplary embodiment, the scheduling module 502 is further configured to:

[0169] Upon receiving a request from an edge device to evict the current service, the system re-identifies available devices and reschedules the service from the edge device to other available devices.

[0170] Figure 6 This is a block diagram illustrating a computer device 600 for power-aware scheduling of edge cloud computing services, according to an exemplary embodiment. For example, the computer device 600 may be provided as a server. (Refer to...) Figure 6 The computer device 600 includes a processor 601, the number of which can be set to one or more as needed. The computer device 600 also includes a memory 602 for storing instructions executable by the processor 601, such as application programs. The number of memories can be set to one or more as needed. The stored application programs can be one or more. The processor 601 is configured to execute instructions to perform the aforementioned edge cloud computing service computing power-aware scheduling method.

[0171] Those skilled in the art will understand that embodiments of this application can be provided as methods, apparatus (devices), or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. Computer storage media include volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data), including but not limited to RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible by a computer. Furthermore, it is well known to those skilled in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and can include any information delivery medium.

[0172] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (devices), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations 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, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0173] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes

[0174] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0175] In this application, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that an article or device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such an article or device. Without further limitation, an element defined by the phrase "comprising..." does not exclude the presence of other identical elements in the article or device that includes said element.

[0176] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0177] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if these modifications and variations fall within the scope of the claims of this application and their equivalents, the intent of this application also includes these modifications and variations.

Claims

1. A computing power-aware scheduling method for edge cloud computing services, characterized in that, Applied to edge devices, including: Virtual computing agents deployed on edge devices acquire real-time values ​​of multiple key factors affecting the device's service capabilities; The computing power consumption and available computing power of the system for the current business are calculated based on the values ​​of the aforementioned key factors. The service capability is assessed based on the computing power consumption of the current business and the available computing power of the system. The evaluation service capabilities are sent to the scheduling system so that the scheduling system can schedule and allocate services based on the evaluation service capabilities.

2. The edge cloud computing service computing power-aware scheduling method as described in claim 1, characterized in that, The values ​​of the multiple key factors affecting the service capacity of the equipment include CPU utilization, disk I / O utilization, memory utilization, load quantity, response time, and bandwidth usage. The calculation of the computing power consumption of the current service and the available computing power of the system based on the values ​​of these multiple key factors includes: Set a weighting coefficient for each key factor; Set the maximum computing power of the device; The sum of the products of the magnitude and weight coefficient of each key factor obtained before the equipment processes the business is determined as the initial computing power consumption; The sum of the products of the value of each key factor and its weight coefficient when the device processes the business is determined as the current computing power consumption; The difference between the current computing power consumption and the initial computing power consumption is determined as the computing power consumption of the current service. The difference between the device's maximum computing power and its current computing power consumption is determined as the system's available computing power.

3. The edge cloud computing service computing power-aware scheduling method as described in claim 2, characterized in that, The method of setting weight coefficients for each key factor also includes: When the value of the key factor is less than a preset threshold, the weighting coefficient is a first weighting coefficient; when the value of the key factor is greater than or equal to the preset threshold, the weighting coefficient is a second weighting coefficient; wherein the second weighting coefficient is greater than the first weighting coefficient.

4. The edge cloud computing service computing power-aware scheduling method as described in claim 1, characterized in that, The evaluation of service capabilities based on the current computing power consumption and available computing power of the system includes: When the available computing power of the system is greater than a first preset threshold, it is determined that the evaluation service capability is greater than the computing power consumption of the current business. When the available computing power of the system is less than a first preset threshold, it is determined that the evaluation service capability is less than the computing power consumption of the current business, until the available computing power of the system equals the first preset threshold.

5. The edge cloud computing service computing power-aware scheduling method as described in claim 1, characterized in that, The assessment of service capabilities based on the current computing power consumption and available computing power of the system also includes: When the available computing power of the system is greater than the second preset threshold, it is determined that the evaluation service capacity is greater than the computing power consumption of the current business. Gradually increase the evaluation service capability until the available computing power of the system is less than or equal to a third preset threshold. In subsequent weeks, if the computing power consumption of the current service is less than the evaluation service capacity of the previous week, the evaluation service capacity is reduced; if the computing power consumption of the current service is greater than the evaluation service capacity of the previous week, the evaluation service capacity is increased; until the computing power consumption of the current service equals the evaluation service capacity, the evaluation service capacity at this time is determined to be the optimal evaluation service capacity.

6. The edge cloud computing service computing power-aware scheduling method as described in any one of claims 4 or 5, characterized in that, When the available computing power of the system is less than the fourth preset threshold, a request to evict the current service is sent to the scheduling system.

7. A computing power-aware scheduling method for edge cloud computing services, characterized in that, Applied to scheduling systems, including: The system receives evaluation service capabilities of edge devices from virtual computing agents deployed on edge devices, and schedules the evaluation of available devices and their service capabilities within the system. Service scheduling is performed based on the available devices and their service capabilities.

8. The edge cloud computing service computing power-aware scheduling method as described in claim 7, characterized in that, Also includes: Upon receiving a request from an edge device to evict the current service, a new available device is determined, and the service in the edge device is rescheduled to another available device.

9. An edge computing service computing power awareness and scheduling device, characterized in that, Applied to edge devices, including: The value acquisition module is used to acquire the values ​​of multiple key factors that affect the service capabilities of the equipment in real time. The calculation module is used to calculate the computing power consumption of the current business and the available computing power of the system based on the values ​​of the multiple key factors. The evaluation module is used to determine the evaluation service capability based on the computing power consumption of the current service and the available computing power of the system; The reporting module is used to send the evaluation service capabilities to the scheduling system, so that the scheduling system can schedule and allocate services according to the evaluation service capabilities.

10. The edge cloud computing service computing power awareness and scheduling device as described in claim 9, characterized in that, The values ​​of the key factors affecting the device's service capabilities include CPU utilization, disk I / O utilization, memory utilization, load quantity, response time, and bandwidth usage. The calculation module is also used for: Set a weighting coefficient for each key factor; Set the maximum computing power of the device; The sum of the products of the magnitude and weight coefficient of each key factor obtained before the equipment processes the business is determined as the initial computing power consumption; The sum of the products of the value of each key factor and its weight coefficient when the device processes the business is determined as the current computing power consumption; The difference between the current computing power consumption and the initial computing power consumption is determined as the computing power consumption of the current service. The difference between the device's maximum computing power and its current computing power consumption is determined as the system's available computing power.

11. The edge cloud computing service computing power awareness and scheduling device as described in claim 9, characterized in that, The evaluation module is also used for: When the available computing power of the system is greater than a first preset threshold, it is determined that the evaluation service capability is greater than the computing power consumption of the current business. When the available computing power of the system is less than a first preset threshold, it is determined that the evaluation service capability is less than the computing power consumption of the current business, until the available computing power of the system equals the first preset threshold.

12. The edge cloud computing service computing power awareness and scheduling device as described in claim 9, characterized in that, The evaluation module is also used for: When the available computing power of the system is greater than the second preset threshold, it is determined that the evaluation service capacity is greater than the computing power consumption of the current business. Gradually increase the evaluation service capability until the available computing power of the system is less than or equal to a third preset threshold. In subsequent weeks, if the computing power consumption of the current service is less than the evaluation service capacity of the previous week, the evaluation service capacity is reduced; if the computing power consumption of the current service is greater than the evaluation service capacity of the previous week, the evaluation service capacity is increased; until the computing power consumption of the current service equals the evaluation service capacity, the evaluation service capacity at this time is determined to be the optimal evaluation service capacity.

13. The edge cloud computing service computing power awareness and scheduling device as described in any one of claims 11 or 12, characterized in that, The reporting module is also used for: When the available computing power of the system is less than the fourth preset threshold, a request to evict the current service is sent to the scheduling system.

14. An edge computing service computing power awareness and scheduling device, characterized in that, Applied to scheduling systems, including: The receiving module is used to receive the evaluation service capabilities of the edge devices sent by the virtual computing power agents deployed on the edge devices, and to schedule the available devices and service capabilities of the available devices in the evaluation system. The scheduling module is used to schedule services based on the available devices and their service capabilities.

15. The edge cloud computing service computing power awareness and scheduling device as described in claim 14, characterized in that, The scheduling module is also used for: Upon receiving a request from an edge device to evict the current service, a new available device is determined, and the service in the edge device is rescheduled to another available device.

16. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed, it implements the steps of the method as described in any one of claims 1-8.

17. A computer device comprising a processor, a memory, and a computer program stored in the memory, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1-8.

Images