Collaborative optimization method for real-time cloud service scheduling and multi-server system configuration

Abstract

The invention provides a collaborative optimization method for real-time cloud service scheduling and multi-server system configuration, and the method comprises the steps: determining the task type of a to-be-executed task through system hardware request information, carrying out the dynamic adjustment of the current CPU weight, the current memory weight and the current disk weight, and carrying out the real-time cloud service scheduling and multi-server system configuration. The method comprises the steps of determining a target CPU weight, a target memory weight and a target disk weight, determining a performance score of each cloud server according to a CPU utilization rate, a memory occupancy rate, a disk IOPS, the target CPU weight, the target memory weight and the target disk weight, and adjusting the weights in real time based on different task types to dynamically evaluate the performance scores of the cloud servers. The target cloud server can be determined in a targeted mode, dynamic requirements of different task types are met, the to-be-executed task is scheduled to the target cloud server to be executed, the scheduling effect is improved, and normal execution of the task is guaranteed.

Description

Technical Field

[0001] The present application relates to the server field, and in particular to a collaborative optimization method for real-time cloud service scheduling and multi-server system configuration. Background Art

[0002] A cloud server cluster is a system that connects multiple cloud servers via a network, creating a shared resource and collaborative work environment. By combining the computing, storage, and network resources of multiple servers, it provides high-performance, high-availability, and scalable services. Currently, when executing tasks, collaborative scheduling between different cloud servers is based on the default Kubernetes scheduler, which uses preset static weights to allocate resources for task scheduling. This system cannot adapt to the dynamic demands of different task types, cannot achieve effective scheduling, and can easily affect task execution. Summary of the Invention

[0003] The present application provides a collaborative optimization method for real-time cloud service scheduling and multi-server system configuration to solve at least one problem existing in the related art. The technical solution is as follows:

[0004] The present invention provides a method for collaborative optimization of real-time cloud service scheduling and multi-server system configuration, including:

[0005] Obtain system hardware configuration information of each cloud server in real time, including CPU utilization, memory usage, and disk IOPS;

[0006] Obtaining the task to be executed and the system hardware request information corresponding to the task to be executed, and determining the task type of the task to be executed based on the system hardware request information;

[0007] According to the task type, the current CPU weight, current memory weight, and current disk weight are dynamically adjusted to determine the target CPU weight, target memory weight, and target disk weight;

[0008] The performance score of each cloud server is determined based on CPU utilization, memory occupancy, disk IOPS, target CPU weight, target memory weight, and target disk weight. The target cloud server is determined based on the performance score, and the tasks to be executed are scheduled to the target cloud server for execution.

[0009] In one embodiment, determining the task type of the task to be executed based on the system hardware request information includes:

[0010] Determine the CPU demand based on the number of CPU cores requested in the system hardware request information;

[0011] Determine the memory requirement based on the memory size requested in the system hardware request information;

[0012] Determine the IOPS requirement based on the disk IOPS requested in the system hardware request information;

[0013] Determine the task type to be executed based on the CPU demand, memory demand, and IOPS demand.

[0014] In one embodiment, the CPU demand is determined based on a ratio of the requested number of CPU cores to the baseline number of cores;

[0015] Determine the memory requirement based on the ratio of the requested memory size to the baseline memory size;

[0016] Determine the IOPS requirement based on the ratio of the requested disk IOPS to the benchmark disk IOPS.

[0017] In one embodiment, determining the task type of the task to be executed based on the CPU demand, the memory demand, and the IOPS demand includes:

[0018] When the CPU demand is greater than the first threshold, the memory demand is less than the second threshold, and the IOPS demand is less than the second threshold, determining that the task type to be executed is computationally intensive;

[0019] When the CPU demand is less than the third threshold and the IOPS demand is greater than the third threshold, determining that the task type of the task to be executed is IO intensive;

[0020] When the CPU demand is less than the third threshold, the memory demand is greater than the first threshold, and the IOPS demand is less than the second threshold, determining that the task type to be executed is memory intensive;

[0021] Otherwise, the task type of the task to be executed is determined to be other types;

[0022] The first threshold is greater than the third threshold, and the third threshold is greater than the second threshold.

[0023] In one embodiment, dynamically adjusting the current CPU weight, the current memory weight, and the current disk weight according to the task type to determine the target CPU weight, the target memory weight, and the target disk weight includes:

[0024] Obtain historical task execution data corresponding to compute-intensive, IO-intensive, and memory-intensive tasks respectively;

[0025] Analyze historical task execution data to determine the computation quality value corresponding to computation-intensive tasks, the IO quality value corresponding to IO-intensive tasks, and the memory quality value corresponding to memory-intensive tasks.

[0026] Based on the task type of the task to be executed, the target CPU weight after adjustment is determined according to the computing quality value and the current CPU weight. The target disk weight after adjustment is determined according to the IO quality value and the current disk weight. The target memory weight after adjustment is determined according to the memory quality value and the current memory weight.

[0027] In one embodiment, analyzing the historical task execution data separately to determine the computing quality value corresponding to the computing intensive task, the IO quality value corresponding to the IO intensive task, and the memory quality value corresponding to the memory intensive task includes:

[0028] Analyze historical task execution data to determine the SLA time limit, the average completion time of the first task corresponding to compute-intensive tasks, the proportion of idle CPU resources, the average completion time of the second task corresponding to IO-intensive tasks, the proportion of idle disk resources, and the average completion time of the third task corresponding to memory-intensive tasks, and the proportion of idle memory resources.

[0029] Determine a first ratio, a second ratio, and a third ratio of the average completion time of the first task, the average completion time of the second task, and the average completion time of the third task to the time upper limit of the SLA respectively;

[0030] Respectively determining differences between the preset value and the first ratio, the second ratio, the third ratio, the CPU idle resource ratio, the disk idle resource ratio, and the memory idle resource ratio to obtain a first difference, a second difference, a third difference, a fourth difference, a fifth difference, and a sixth difference;

[0031] Based on the sum of the first difference and the fourth difference, the computing quality value corresponding to the computing intensive type is determined; based on the sum of the second difference and the fifth difference, the IO quality value corresponding to the IO intensive type is determined; based on the sum of the third difference and the sixth difference, the memory quality value corresponding to the memory intensive type is determined.

[0032] In one embodiment, determining the adjusted target CPU weight based on the task type to be executed, the calculation quality value, and the current CPU weight, determining the adjusted target disk weight based on the IO quality value and the current disk weight, and determining the adjusted target memory weight based on the memory quality value and the current memory weight includes:

[0033] When the task type of the task to be executed is compute-intensive, the first amplitude is determined according to the product of the computation quality value and the preset learning rate, the adjusted target CPU weight is determined according to the sum of the current CPU weight and the first amplitude, and the current disk weight and the current memory weight are respectively reduced by half of the first amplitude to obtain the adjusted target disk weight and the adjusted target memory weight;

[0034] When the task type of the task to be executed is IO intensive, the second amplitude is determined based on the product of the IO quality value and the preset learning rate. The adjusted target disk weight is determined based on the sum of the current disk weight and the second amplitude. The current CPU weight is reduced by the second amplitude to obtain the adjusted target CPU weight, and the current memory weight is used as the adjusted target memory weight.

[0035] When the task type of the task to be executed is memory intensive, the third amplitude is determined based on the product of the memory quality value and the preset learning rate. The adjusted target memory weight is determined based on the sum of the current memory weight and the third amplitude. The current disk weight is reduced by the third amplitude to obtain the adjusted target disk weight. The current CPU weight is used as the adjusted target CPU weight.

[0036] When the task type of the task to be executed is other types, the current CPU weight, the current disk weight, and the current memory weight are used as the adjusted target CPU weight, the adjusted target disk weight, and the adjusted target memory weight, respectively.

[0037] In one embodiment, determining the performance score of each cloud server based on CPU utilization, memory occupancy, disk IOPS, target CPU weight, target memory weight, and target disk weight includes:

[0038] Determine the CPU availability score, memory availability score, and disk availability score based on CPU utilization, memory usage, and disk IOPS respectively;

[0039] The performance score of each cloud server is obtained by taking a weighted sum of the CPU availability score, target CPU weight, memory availability score, target memory weight, disk availability score, and target disk weight.

[0040] In one embodiment, determining the CPU availability score value, the memory availability score value, and the disk availability score value based on the CPU utilization, the memory occupancy, and the disk IOPS respectively includes:

[0041] Determine a fourth ratio of the CPU utilization to the CPU overload threshold, a fifth ratio of the memory occupancy to the memory overload threshold, and a sixth ratio of the disk IOPS to the disk overload IOPS value;

[0042] The CPU availability score value, the memory availability score value, and the disk availability score value are determined according to the differences between the preset value and the fourth ratio, the fifth ratio, and the sixth ratio, respectively.

[0043] In one embodiment, determining a target cloud server based on the performance score and scheduling the task to be executed to the target cloud server for execution includes:

[0044] The cloud server with the highest performance score is selected as the target cloud server, and the tasks to be executed are scheduled to the target cloud server.

[0045] The beneficial effects of the above technical solution include at least:

[0046] By obtaining the system hardware configuration information of each cloud server in real time, the system hardware configuration information includes CPU utilization, memory occupancy and disk IOPS, obtaining the tasks to be executed and the system hardware request information corresponding to the tasks to be executed, and determining the task type of the tasks to be executed based on the system hardware request information. According to the task type, the current CPU weight, current memory weight and current disk weight are dynamically adjusted to determine the target CPU weight, target memory weight and target disk weight. According to the CPU utilization, memory occupancy, disk IOPS, target CPU weight, target memory weight and target disk weight, the performance score of each cloud server is determined. The weight is adjusted in real time based on different task types to dynamically evaluate the performance score of the cloud server, which is conducive to targeted determination of the target cloud server, adapting to the dynamic needs of different task types, and scheduling the tasks to be executed to the target cloud server for execution, which is conducive to improving the scheduling effect and ensuring the normal execution of the task.

[0047] The above summary is for illustrative purposes only and is not intended to be limiting in any way. In addition to the illustrative aspects, embodiments and features described above, further aspects, embodiments and features of the present application will be readily apparent by reference to the accompanying drawings and the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In the accompanying drawings, unless otherwise specified, the same reference numerals throughout the multiple drawings represent the same or similar components or elements. These drawings are not necessarily drawn to scale. It should be understood that these drawings only depict some embodiments disclosed in this application and should not be construed as limiting the scope of this application.

[0049] Figure 1 This is a flowchart of the steps of a collaborative optimization method for real-time cloud service scheduling and multi-server system configuration according to an embodiment of the present application. DETAILED DESCRIPTION

[0050] Hereinafter, only certain exemplary embodiments are briefly described. As will be appreciated by those skilled in the art, the described embodiments may be modified in various ways without departing from the spirit or scope of the present application. Therefore, the drawings and description are to be regarded as illustrative in nature and not restrictive.

[0051] Reference Figure 1, a flowchart of a collaborative optimization method for real-time cloud service scheduling and multi-server system configuration according to an embodiment of the present application is shown. The collaborative optimization method for real-time cloud service scheduling and multi-server system configuration may include at least steps S100-S400:

[0052] S100: Obtain system hardware configuration information of each cloud server in real time.

[0053] Optionally, the system hardware configuration information includes CPU utilization, memory occupancy, and disk IOPS (Input / Output Operations Per Second).

[0054] S200: Obtain a task to be executed and system hardware request information corresponding to the task to be executed, and determine a task type of the task to be executed according to the system hardware request information.

[0055] S300: Dynamically adjust the current CPU weight, the current memory weight, and the current disk weight according to the task type to determine the target CPU weight, the target memory weight, and the target disk weight.

[0056] S400. Determine the performance score of each cloud server based on CPU utilization, memory occupancy, disk IOPS, target CPU weight, target memory weight, and target disk weight, and determine the target cloud server based on the performance score, and schedule the task to be executed to the target cloud server for execution.

[0057] The method of the embodiment of the present application can be executed by the main cloud server, which performs cloud service scheduling to achieve the management of the remaining multiple cloud servers. For example, when a task is to be executed, the system hardware configuration information of the cloud server is obtained, the system hardware configuration information is processed, and task scheduling is performed to achieve overall collaborative optimization.

[0058] The technical solution of the embodiment of the present application obtains the system hardware configuration information of each cloud server in real time, the system hardware configuration information includes CPU utilization, memory occupancy and disk IOPS, obtains the task to be executed and the system hardware request information corresponding to the task to be executed, and determines the task type of the task to be executed based on the system hardware request information. According to the task type, the current CPU weight, the current memory weight and the current disk weight are dynamically adjusted to determine the target CPU weight, the target memory weight and the target disk weight. According to the CPU utilization, memory occupancy, disk IOPS, the target CPU weight, the target memory weight and the target disk weight, the performance score of each cloud server is determined. The weight is adjusted in real time based on different task types to dynamically evaluate the performance score of the cloud server, which is conducive to targeted determination of the target cloud server, adapting to the dynamic needs of different task types, and scheduling the task to be executed to the target cloud server for execution, which is conducive to improving the scheduling effect and ensuring the normal execution of the task.

[0059] In one embodiment, the administrator can configure the tasks to be executed in the main cloud server. The main cloud server obtains the tasks to be executed and the system hardware request information corresponding to the tasks to be executed, such as but not limited to the requested number of CPU cores, the requested memory size, and the requested disk IOPS.

[0060] In one embodiment, step S200 determines the task type of the task to be executed according to the system hardware request information, including steps S210-S240:

[0061] S210 : Determine the CPU demand according to the number of CPU cores requested in the system hardware request information.

[0062] Specifically, the CPU demand is determined according to the ratio of the requested number of CPU cores to the baseline number of cores.

[0063] S220: Determine the memory requirement according to the memory size requested in the system hardware request information.

[0064] Specifically, the memory demand is determined according to the ratio of the requested memory size to the benchmark memory size.

[0065] S230: Determine the IOPS requirement according to the disk IOPS requested in the system hardware request information.

[0066] Specifically, the IOPS requirement is determined according to the ratio of the requested disk IOPS to the benchmark disk IOPS.

[0067] It should be noted that the benchmark number of cores (eg, 2 cores), benchmark memory size (eg, 2 GB), and benchmark disk IOPS (eg, 1000 IOPS) can be obtained based on historical data analysis and are not specifically limited.

[0068] S240: Determine the task type of the task to be executed according to the CPU demand, the memory demand, and the IOPS demand.

[0069] In one embodiment, including S2401-S2404:

[0070] S2401: When the CPU demand is greater than a first threshold, the memory demand is less than a second threshold, and the IOPS demand is less than a second threshold, determine that the task type of the task to be executed is computationally intensive.

[0071] Optionally, when the CPU demand is greater than the first threshold, the memory demand is less than the second threshold, and the IOPS demand is less than the second threshold, it means that the current task to be executed has high CPU consumption, and the demand for memory and IOPS is low. It can be determined that the task type of the task to be executed is compute-intensive, such as AI model training, data prediction tasks, etc.

[0072] S2402: When the CPU demand is less than the third threshold and the IOPS demand is greater than the third threshold, determine that the task type of the task to be executed is IO intensive.

[0073] Optionally, when the CPU demand is less than the third threshold and the IOPS demand is greater than the third threshold, it means that the current task to be executed has low CPU consumption and high IOPS demand, and the task type of the task to be executed is determined to be IO intensive, such as database query.

[0074] S2403: When the CPU demand is less than the third threshold, the memory demand is greater than the first threshold, and the IOPS demand is less than the second threshold, determine that the task type of the task to be executed is memory intensive.

[0075] Optionally, when the CPU demand is less than the third threshold, the memory demand is greater than the first threshold, and the IOPS demand is less than the second threshold, it means that the current task to be executed has low CPU consumption, low IOPS demand, and high memory requirement, and the task type of the task to be executed is determined to be memory intensive, such as a cache service.

[0076] S2404: Otherwise, determine that the task type of the task to be executed is other types;

[0077] It should be noted that the first threshold is greater than the third threshold, and the third threshold is greater than the second threshold; if the above three situations are not met, it means that it may be other types of tasks or composite tasks, and at this time the task type of the to-be-executed task is determined as other types.

[0078] In an implementation, the step S300 includes steps S310-S330:

[0079] S310, respectively acquiring historical task execution data corresponding to the compute-intensive, IO-intensive and memory-intensive.

[0080] S320, respectively analyzing the historical task execution data to determine the compute quality value corresponding to the compute-intensive, the IO quality value corresponding to the IO-intensive, and the memory quality value corresponding to the memory-intensive. Specifically:

[0081] 1. respectively analyzing the historical task execution data to determine the upper limit of time T of SLA (Service Level Agreement, service level agreement) in time SLA , the first task average completion time T1 corresponding to the compute-intensive (that is, calculating the average time according to the completion time of all compute-intensive tasks), the CPU idle resource proportion R1 (that is, calculating the average idle rate based on the CPU idle rate of all compute-intensive tasks), the second task average completion time T2 corresponding to the IO-intensive (that is, calculating the average time according to the completion time of all IO-intensive tasks), the disk idle resource proportion R2 (that is, calculating the average idle rate based on the disk idle rate of all IO-intensive tasks), the third task average completion time T3 corresponding to the memory-intensive (that is, calculating the average time according to the completion time of all memory-intensive tasks), and the memory idle resource proportion R3 (that is, calculating the average idle rate based on the memory idle rate of all memory-intensive tasks).

[0082] 2. respectively determining the first ratio of the first task average completion time, the second task average completion time, the third task average completion time and the upper limit of time of SLA the second ratio and the third ratio

[0083] 3. respectively determining the difference between the preset value and the first ratio, the second ratio, the third ratio, the CPU idle resource proportion, the disk idle resource proportion and the memory idle resource proportion, to obtain the first difference, the second difference, the third difference, the fourth difference, the fifth difference and the sixth difference.

[0084] Exemplarily, the preset value is 1, and the difference between the preset value 1 and the first ratio, the second ratio, the third ratio, the CPU idle resource proportion, the disk idle resource proportion and the memory idle resource proportion is determined to obtain the first difference Second difference The third difference a fourth difference 1-R1, a fifth difference 1-R2, and a sixth difference 1-R3.

[0085] 4. Determine the computation quality value corresponding to the computation-intensive type based on the sum of the first difference and the fourth difference, determine the IO quality value corresponding to the IO-intensive type based on the sum of the second difference and the fifth difference, and determine the memory quality value corresponding to the memory-intensive type based on the sum of the third difference and the sixth difference. Specifically:

[0086] The calculation quality value Q1 corresponding to the calculation intensive type is:

[0087] IO quality value Q2 corresponding to IO intensive:

[0088] Memory quality value Q3 corresponding to memory intensive:

[0089] S330. Based on the task type of the task to be executed, determine the adjusted target CPU weight according to the calculation quality value and the current CPU weight, determine the adjusted target disk weight according to the IO quality value and the current disk weight, and determine the adjusted target memory weight according to the memory quality value and the current memory weight.

[0090] First, when the task type to be executed is computationally intensive, the first amplitude ηQ1 is determined according to the product of the computational quality value and the preset learning rate η, and the first amplitude ηQ1 is determined according to the current CPU weight α now The sum of the first amplitude ηQ1 and the adjusted target CPU weight α is determined new =α now +ηQ1. Then, the current disk weight β now , current memory weight γ now Reduce the first amplitude by half to obtain the adjusted target disk weight β new and the adjusted target memory weight γ new , ensuring that the sum of all weights after adjustment is still 1.

[0091] Secondly, when the task type to be executed is IO intensive, the second amplitude ηQ2 is determined according to the product of the IO quality value and the preset learning rate, and the second amplitude ηQ2 is determined according to the current disk weight β now The sum of the second magnitude determines the adjusted target disk weight β new =β now +ηQ2, and the current CPU weight α now Reduce the second amplitude to obtain the adjusted target CPU weight α new , the current memory weight γnow as the adjusted target memory weight γ new .

[0092] Then, when the task type to be executed is memory-intensive, the third amplitude ηQ3 is determined according to the product of the memory quality value and the preset learning rate, and the third amplitude ηQ3 is determined according to the current memory weight γ now The sum of the third magnitude determines the adjusted target memory weight γ new＝ γ now+ ηQ3, and the current disk weight β now Reduce the third amplitude to get the adjusted target disk weight β new , set the current CPU weight α now As the adjusted target CPU weight α new .

[0093] Finally, when the task type to be executed is other types, the current CPU weight, current disk weight, and current memory weight are used as the adjusted target CPU weight α new , adjusted target disk weight β new and the adjusted target memory weight γ new , that is, the weights are not changed.

[0094] In one embodiment, step S400 determines the performance score of each cloud server based on CPU utilization, memory occupancy, disk IOPS, target CPU weight, target memory weight, and target disk weight, including steps S410-S420:

[0095] S410 : Determine a CPU availability score value, a memory availability score value, and a disk availability score value according to the CPU utilization rate, the memory occupancy rate, and the disk IOPS, respectively.

[0096] First, determine the CPU utilization C B and CPU overload threshold C MAX (i.e., if the threshold is exceeded, it is considered overloaded) the fourth ratio, memory usage and M B Memory overload threshold M MAX The fifth ratio (i.e., exceeding this threshold is considered overloaded), disk IOPSD B and disk overload IOPS value D MAX (ie, if the threshold is exceeded, it is considered to be overloaded).

[0097] Secondly, according to the preset value (exemplarily 1) and the fourth ratio Fifth Ratio Sixth ratio The difference between the two values ​​determines the CPU available score value. Memory available score value and the disk availability score

[0098] S420: Perform a weighted summation based on the CPU availability score, target CPU weight, memory availability score, target memory weight, disk availability score, and target disk weight to obtain a performance score for each cloud server. Specifically, the performance score formula for each cloud server is:

[0099]

[0100] Finally, the cloud server with the highest performance score is used as the target cloud server. The master cloud server dispatches the tasks to be executed to the target cloud server, so that multiple cloud servers can achieve collaborative optimization and ensure that each cloud server executes the most suitable tasks to be executed, which is conducive to improving the task execution effect and efficiency.

[0101] In the description of this specification, the reference terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples" mean that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials, or characteristics described may be combined in any appropriate manner in any one or more embodiments or examples. In addition, those skilled in the art may combine and integrate different embodiments or examples described in this specification, as well as features of different embodiments or examples, unless they are mutually inconsistent.

[0102] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of technical features indicated. Thus, features specified as "first" or "second" may explicitly or implicitly include at least one of such features. In the description of this application, "plurality" means two or more, unless otherwise specifically defined.

[0103] Any process or method description in a flow chart or otherwise described herein can be understood to represent a module, segment or portion of code comprising one or more executable instructions for implementing the steps of a specific logical function or process. The scope of the preferred embodiments of the present application includes additional implementations in which the functions may be performed in a different order than shown or discussed, including in a substantially simultaneous manner or in a reverse order depending on the functions involved.

[0104] The logic and / or steps represented in the flowchart or otherwise described herein may be considered, for example, as an ordered list of executable instructions for implementing logical functions, and may be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device).

[0105] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily conceive of various modifications or substitutions within the technical scope disclosed in the present application, and such modifications or substitutions should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A collaborative optimization method for real-time cloud service scheduling and multi-server system configuration, characterized in that: include: Obtain system hardware configuration information of each cloud server in real time, including CPU utilization, memory usage, and disk IOPS; Obtaining the task to be executed and the system hardware request information corresponding to the task to be executed, and determining the task type of the task to be executed based on the system hardware request information; According to the task type, the current CPU weight, current memory weight, and current disk weight are dynamically adjusted to determine the target CPU weight, target memory weight, and target disk weight; The performance score of each cloud server is determined based on CPU utilization, memory usage, disk IOPS, target CPU weight, target memory weight, and target disk weight. The target cloud server is determined based on the performance score, and the tasks to be executed are scheduled to the target cloud server for execution.

2. The collaborative optimization method for real-time cloud service scheduling and multi-server system configuration according to claim 1, characterized in that: Determining the task type of the task to be executed according to the system hardware request information includes: Determine the CPU demand based on the number of CPU cores requested in the system hardware request information; Determine the memory requirement based on the memory size requested in the system hardware request information; Determine the IOPS requirement based on the disk IOPS requested in the system hardware request information; Determine the task type to be executed based on the CPU demand, memory demand, and IOPS demand.

3. The collaborative optimization method for real-time cloud service scheduling and multi-server system configuration according to claim 2, characterized in that: Determine the CPU demand based on the ratio of the requested number of CPU cores to the baseline number of cores; Determine the memory requirement based on the ratio of the requested memory size to the baseline memory size; Determine the IOPS requirement based on the ratio of the requested disk IOPS to the benchmark disk IOPS.

4. The collaborative optimization method for real-time cloud service scheduling and multi-server system configuration according to claim 2 or 3, characterized in that: Determining the task type of the task to be executed based on the CPU demand, memory demand, and IOPS demand includes: When the CPU demand is greater than the first threshold, the memory demand is less than the second threshold, and the IOPS demand is less than the second threshold, determining that the task type to be executed is computationally intensive; When the CPU demand is less than the third threshold and the IOPS demand is greater than the third threshold, determining that the task type of the task to be executed is IO intensive; When the CPU demand is less than the third threshold, the memory demand is greater than the first threshold, and the IOPS demand is less than the second threshold, determining that the task type to be executed is memory intensive; Otherwise, the task type of the task to be executed is determined to be other types; The first threshold is greater than the third threshold, and the third threshold is greater than the second threshold.

5. The collaborative optimization method for real-time cloud service scheduling and multi-server system configuration according to claim 4, characterized in that: The dynamically adjusting the current CPU weight, the current memory weight, and the current disk weight according to the task type to determine the target CPU weight, the target memory weight, and the target disk weight includes: Obtain historical task execution data corresponding to compute-intensive, IO-intensive, and memory-intensive tasks respectively; Analyze historical task execution data to determine the computation quality value corresponding to computation-intensive tasks, the IO quality value corresponding to IO-intensive tasks, and the memory quality value corresponding to memory-intensive tasks. Based on the task type of the task to be executed, the target CPU weight after adjustment is determined according to the computing quality value and the current CPU weight. The target disk weight after adjustment is determined according to the IO quality value and the current disk weight. The target memory weight after adjustment is determined according to the memory quality value and the current memory weight.

6. The collaborative optimization method for real-time cloud service scheduling and multi-server system configuration according to claim 5, characterized in that: The analyzing of historical task execution data to determine the computing quality value corresponding to the computing-intensive task, the IO quality value corresponding to the IO-intensive task, and the memory quality value corresponding to the memory-intensive task includes: Analyze historical task execution data to determine the SLA time limit, the average completion time of the first task corresponding to compute-intensive tasks, the proportion of idle CPU resources, the average completion time of the second task corresponding to IO-intensive tasks, the proportion of idle disk resources, and the average completion time of the third task corresponding to memory-intensive tasks, and the proportion of idle memory resources. Determine a first ratio, a second ratio, and a third ratio of the average completion time of the first task, the average completion time of the second task, and the average completion time of the third task to the time upper limit of the SLA respectively; Respectively determining differences between the preset value and the first ratio, the second ratio, the third ratio, the CPU idle resource ratio, the disk idle resource ratio, and the memory idle resource ratio to obtain a first difference, a second difference, a third difference, a fourth difference, a fifth difference, and a sixth difference; Based on the sum of the first difference and the fourth difference, the computing quality value corresponding to the computing intensive type is determined; based on the sum of the second difference and the fifth difference, the IO quality value corresponding to the IO intensive type is determined; based on the sum of the third difference and the sixth difference, the memory quality value corresponding to the memory intensive type is determined.

7. The collaborative optimization method for real-time cloud service scheduling and multi-server system configuration according to claim 5, characterized in that: The method of determining the adjusted target CPU weight based on the task type to be executed, the calculated quality value and the current CPU weight, the adjusted target disk weight based on the IO quality value and the current disk weight, and the adjusted target memory weight based on the memory quality value and the current memory weight includes: When the task type of the task to be executed is compute-intensive, the first amplitude is determined according to the product of the computation quality value and the preset learning rate, the adjusted target CPU weight is determined according to the sum of the current CPU weight and the first amplitude, and the current disk weight and the current memory weight are respectively reduced by half of the first amplitude to obtain the adjusted target disk weight and the adjusted target memory weight; When the task type of the task to be executed is IO intensive, the second amplitude is determined based on the product of the IO quality value and the preset learning rate. The adjusted target disk weight is determined based on the sum of the current disk weight and the second amplitude. The current CPU weight is reduced by the second amplitude to obtain the adjusted target CPU weight, and the current memory weight is used as the adjusted target memory weight. When the task type of the task to be executed is memory intensive, the third amplitude is determined based on the product of the memory quality value and the preset learning rate. The adjusted target memory weight is determined based on the sum of the current memory weight and the third amplitude. The current disk weight is reduced by the third amplitude to obtain the adjusted target disk weight. The current CPU weight is used as the adjusted target CPU weight. When the task type of the task to be executed is other types, the current CPU weight, the current disk weight, and the current memory weight are used as the adjusted target CPU weight, the adjusted target disk weight, and the adjusted target memory weight, respectively.

8. The collaborative optimization method for real-time cloud service scheduling and multi-server system configuration according to claim 1, characterized in that: Determining the performance score of each cloud server based on CPU utilization, memory occupancy, disk IOPS, target CPU weight, target memory weight, and target disk weight includes: Determine the CPU availability score, memory availability score, and disk availability score based on CPU utilization, memory usage, and disk IOPS respectively; The performance score of each cloud server is obtained by taking a weighted sum of the CPU availability score, target CPU weight, memory availability score, target memory weight, disk availability score, and target disk weight.

9. The collaborative optimization method for real-time cloud service scheduling and multi-server system configuration according to claim 8, characterized in that: Determining the CPU availability score value, the memory availability score value, and the disk availability score value according to the CPU utilization rate, the memory occupancy rate, and the disk IOPS respectively includes: Determine a fourth ratio of the CPU utilization to the CPU overload threshold, a fifth ratio of the memory occupancy to the memory overload threshold, and a sixth ratio of the disk IOPS to the disk overload IOPS value; The CPU availability score value, the memory availability score value, and the disk availability score value are determined according to the differences between the preset value and the fourth ratio, the fifth ratio, and the sixth ratio, respectively.

10. The collaborative optimization method for real-time cloud service scheduling and multi-server system configuration according to claim 1, characterized in that: Determining the target cloud server based on the performance score and dispatching the task to be executed to the target cloud server includes: The cloud server with the highest performance score is selected as the target cloud server, and the tasks to be executed are scheduled to the target cloud server.

Images