A heterogeneous multi-core collaborative management and automated operation and maintenance system

Abstract

The present invention provides a heterogeneous multi-core collaborative management and automated operation and maintenance system, which relates to the field of heterogeneous management technology, including: a resource processing module: used to identify processing core resources and a unified API interface; a scheduling module: used to intelligently schedule the target processing core to complete the current task based on the analysis of the current task requirements and the processing core characteristics; a monitoring module: used to monitor the execution status of the task in real time, and when an abnormal state is detected, take corresponding adjustment measures for processing. By identifying the processing core and the unified API interface, based on the analysis of the current task requirements and the processing core characteristics, the target processing core is intelligently scheduled to complete the current task; the execution status of the task is intelligently scheduled to complete the current task based on the analysis of the current task requirements and the processing core characteristics; the execution status of the task is monitored in real time, and when an abnormal state is detected, the corresponding adjustment measures are taken for processing, which can simplify operation and maintenance management, optimize the task execution process, improve the response speed of the task, and improve the flexibility and utilization of heterogeneous resources.

Description

Technical Field

[0001] The present invention relates to the field of heterogeneous management technology, and in particular to a heterogeneous multi-core collaborative management and automated operation and maintenance system. Background Art

[0002] In recent years, with the rapid development of information technology, the volume of data has expanded dramatically, and the requirements of various tasks have far exceeded the capabilities of traditional CPUs. Single chips often seem "incapable of doing so", and their performance has basically reached the physical technology limit. Therefore, how to intelligently schedule processing cores to complete tasks based on task requirements and processing core characteristics has become one of the current research focuses.

[0003] Therefore, the present invention provides a heterogeneous multi-core collaborative management and automated operation and maintenance system. Summary of the invention

[0004] The present invention provides a heterogeneous multi-core collaborative management and automated operation and maintenance system, which is used to intelligently schedule the target processing core to complete the current task based on the analysis of the current task requirements and the processing core characteristics by identifying the processing core and the unified API interface; monitor the execution status of the task in real time, and when an abnormal status is detected, take corresponding adjustment measures to handle it, which can simplify operation and maintenance management, optimize the task execution process, improve the response speed of the task, and improve the flexibility and utilization of heterogeneous resources.

[0005] The present invention provides a heterogeneous multi-core collaborative management and automated operation and maintenance system, comprising:

[0006] Resource processing module: used to identify the processing core and unify the API interface;

[0007] Scheduling module: used to intelligently schedule the target processing core to complete the task corresponding to the current task request based on the analysis of the current task request and the processing core characteristics;

[0008] Monitoring module: used to monitor the execution status of tasks in real time, and take corresponding adjustment measures to deal with abnormal execution status when it is detected.

[0009] Preferably, the resource processing module includes:

[0010] Resource pool building unit: used to develop the hardware abstraction layer, identify the target processing core in the target cloud platform and register it;

[0011] Virtualize all target processing cores and centrally manage them to form a shared resource pool;

[0012] Interface unification unit: used to provide a unified API interface for the system and develop a software abstraction layer so that applications can use the same interface to access resources.

[0013] Preferably, the target processing core includes a chip server, a distributed storage device, a network device and a security device.

[0014] Preferably, the scheduling module includes:

[0015] Demand analysis unit: used to receive and parse the current task request to obtain task request data;

[0016] Prioritizing tasks in the received task request data based on the first task parameter extracted from the task request data to obtain a target task priority sequence;

[0017] Extracting the first resource requirement and the task execution operation from the task request data, and outputting them together with the target task priority sequence as the requirement analysis result;

[0018] Core analysis unit: used for evaluating the characteristics of each processing core based on the first performance parameters of the processing core at the current moment and the historical moment collected, and obtaining the core analysis result;

[0019] Scheduling unit: used to generate a task scheduling strategy using a scheduling analysis block; and to schedule multiple processing cores to jointly complete a task according to the task scheduling strategy using a scheduling implementation block;

[0020] Scheduling analysis block: It is used to combine the demand analysis results of the task with the core analysis results, allocate the processing cores to the tasks in order according to the priority sequence of the target tasks, and generate the task scheduling strategy;

[0021] Scheduling implementation block: used to generate task execution instructions according to the task scheduling strategy and transmit them to the corresponding processing core for task execution.

[0022] Preferably, based on the first task parameter extracted from the task request data, the received tasks are prioritized to obtain a target task priority sequence, including:

[0023] determining a priority coefficient of a current task based on a first task parameter extracted from the task request data;

[0024] Sort the received tasks according to the priority coefficients from large to small to generate a first task priority sequence;

[0025] When there is no dependency relationship between the tasks in the first task priority sequence, marking the current first task priority sequence as the target task priority sequence;

[0026] When there is a dependency relationship between the tasks in the first task priority sequence, a topological sorting algorithm is introduced to adjust the first task priority sequence to obtain a target task priority sequence.

[0027] Preferably, the core analysis unit includes:

[0028] Current analysis block: used to extract characteristic evaluation parameters of the processing core at the current moment from the first performance parameter of the processing core at the current moment;

[0029] Prediction block: used to obtain and use the historical characteristic evaluation parameters of the processing core within a preset time period to construct a historical characteristic change curve;

[0030] Extract trend features of all the historical characteristic change curves, and use the characteristic evaluation parameters of the current processing core as input, input the pre-established characteristic prediction model and output the characteristic prediction parameters;

[0031] The obtained characteristic evaluation parameters of the processing core at the current moment are combined with the characteristic prediction parameters for analysis to obtain a comprehensive characteristic index evaluation coefficient of the characteristic evaluation parameters of the current processing core, and output it as a core analysis result;

[0032] Among them, the calculation formula of the comprehensive characteristic index evaluation coefficient is as follows:

[0033] ; In the formula, denoted as the comprehensive characteristic index evaluation coefficient of the jth characteristic evaluation parameter of the i-th processing core; It is represented as the jth characteristic evaluation parameter of the i-th processing core at the current moment; It is represented as the predicted value of the jth characteristic evaluation parameter of the i-th processing core at the next moment; T is represented as the preset time period; It is represented as the initial configuration value of the jth characteristic evaluation parameter of the i-th processing core; t is more expressed as the time interval between the current moment and the next moment; denoted as the historical characteristic variation coefficient of the jth characteristic evaluation parameter of the i-th processing core; It is represented as the impact weight of the j-th characteristic evaluation parameter of the i-th processing core on the performance of the i-th processing core.

[0034] Preferably, the scheduling analysis block includes:

[0035] For the tasks in the target task priority sequence, the characteristic evaluation parameters are selected from the set execution-characteristic parameter mapping table according to the task type, and marked as the first parameter;

[0036] Extract the corresponding comprehensive characteristic index evaluation coefficients of the first parameters of all processing cores, and calculate the fitness of each processing core with the current task;

[0037] The calculation formula of the fitness is as follows:

[0038] ; In the formula, It is expressed as the suitability of the i-th processing core to the execution operation of the current task; It is expressed as the comprehensive characteristic index evaluation coefficient of the sth first parameter of the i-th processing core, where s=1, 2, 3, , m; m represents the total number of the first parameter; It is represented as the influence weight of the s-th first parameter of the ith processing core on task adaptation; It is expressed as the total task execution failure rate of the i-th processing core; It is expressed as the task execution failure rate of the i-th processing core processing tasks of the same type as the current task;

[0039] The processing core with the highest adaptability to the current task is regarded as the selected processing core, and the current task is allocated to the selected processing core.

[0040] Preferably, the monitoring module includes:

[0041] Monitoring unit: used to monitor the execution status of the current task and the resource usage of the corresponding processing core during the execution of the current task in real time, and obtain the execution status data;

[0042] Compare the execution status data with the corresponding set status threshold, and if there is execution status data exceeding the corresponding set status threshold, determine that the current task execution is abnormal and determine the cause of the abnormality;

[0043] According to the abnormal cause, the corresponding abnormal adjustment strategy is matched to adjust the task execution.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] By identifying the processing core and the unified API interface, based on the analysis of the current task requirements and the processing core characteristics, the target processing core is intelligently scheduled to complete the current task; the execution status of the task is monitored in real time, and when an abnormal status is detected, corresponding adjustment measures are taken to handle it. This can simplify operation and maintenance management, optimize the task execution process, improve the response speed of the task, and improve the flexibility and utilization of heterogeneous resources.

[0046] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0047] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0049] Figure 1 This is a structural diagram of a heterogeneous multi-core collaborative management and automated operation and maintenance system in an embodiment of the present invention. DETAILED DESCRIPTION

[0050] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0051] The embodiment of the present invention provides a heterogeneous multi-core collaborative management and automated operation and maintenance system, such as Figure 1 As shown, including:

[0052] Resource processing module: used to identify the processing core and unify the API interface;

[0053] Scheduling module: used to intelligently schedule the target processing core to complete the task corresponding to the current task request based on the analysis of the current task request and the processing core characteristics;

[0054] Monitoring module: used to monitor the execution status of tasks in real time, and take corresponding adjustment measures to deal with abnormal execution status when it is detected.

[0055] In this embodiment, the processing core resources refer to the chip servers, distributed storage devices, network devices and security devices included in the platform; the unified API interface refers to providing a set of standard and unified interface specifications for the system, so that applications use the same interface to access different processing cores or resources; the current task request refers to the received business task request, which contains basic information about the task, such as task type and expected task execution time; the characteristics of the processing core refer to the characteristic evaluation parameters of different processing cores, including processing speed, power consumption, execution throughput and workload; the target processing core includes chip servers, distributed storage devices, network devices and security devices; the execution status refers to the task execution time, task execution progress and the resource usage of the corresponding allocated processing core during the execution of the current task, such as CPU usage, memory usage, I / O read and write rate, and network bandwidth.

[0056] The beneficial effects of the above technical solution are: by identifying the processing core and the unified API interface, based on the analysis of the current task requirements and the processing core characteristics, the target processing core is intelligently scheduled to complete the current task; the execution status of the task is monitored in real time, and when an abnormal status is detected, corresponding adjustment measures are taken to handle it, which can simplify operation and maintenance management, optimize the task execution process, improve the response speed of the task, and improve the flexibility and utilization of heterogeneous resources.

[0057] An embodiment of the present invention provides a heterogeneous multi-core collaborative management and automated operation and maintenance system, wherein the resource processing module includes:

[0058] Resource pool building unit: used to develop the hardware abstraction layer, identify the target processing core in the target cloud platform and register it;

[0059] Virtualize all target processing cores and centrally manage them to form a shared resource pool;

[0060] Interface unification unit: used to provide a unified API interface for the system and develop a software abstraction layer so that applications can use the same interface to access resources.

[0061] In this embodiment, the purpose of developing a hardware abstraction layer is to abstract the physical hardware resources in the target cloud platform, such as CPU, memory, storage and other resources, so as to decouple them from the upper-level applications; the upper-level applications refer to software components or services for end users or clients; the target cloud platform refers to a specific cloud computing environment or platform; the shared resource pool is a resource pool generated by virtualizing the target processing core and then centrally managing it, wherein virtualization processing refers to using a virtual machine to virtualize the target processing core resources into logical virtual resources; the unified API interface refers to providing a set of standard and unified interface specifications for the system, so that applications use the same interface to access different processing cores or resources; the purpose of developing the software layer is to further encapsulate and abstract the API interface, so that the application can interact with the resource pool without having to worry about the specific implementation details of the underlying API.

[0062] The beneficial effects of the above technical solution are: by building a resource pool, resource management can be simplified and resource utilization can be improved; providing the system with a unified API interface and software abstraction layer can reduce the difficulty of application development and achieve more extensive task integration and collaborative management.

[0063] An embodiment of the present invention provides a heterogeneous multi-core collaborative management and automated operation and maintenance system, wherein the scheduling module includes:

[0064] Demand analysis unit: used to receive and parse the current task request to obtain task request data;

[0065] Prioritizing tasks in the received task request data based on the first task parameter extracted from the task request data to obtain a target task priority sequence;

[0066] Extracting the first resource requirement and the task execution operation from the task request data, and outputting them together with the target task priority sequence as the requirement analysis result;

[0067] Core analysis unit: used for evaluating the characteristics of each processing core based on the first performance parameters of the processing core at the current moment and the historical moment collected, and obtaining the core analysis result;

[0068] Scheduling unit: used to generate a task scheduling strategy using a scheduling analysis block; and to schedule multiple processing cores to jointly complete a task according to the task scheduling strategy using a scheduling implementation block;

[0069] Scheduling analysis block: It is used to combine the demand analysis results of the task with the core analysis results, allocate the processing cores to the tasks in order according to the priority sequence of the target tasks, and generate the task scheduling strategy;

[0070] Scheduling implementation block: used to generate task execution instructions according to the task scheduling strategy and transmit them to the corresponding processing core for task execution.

[0071] In this embodiment, the current task request refers to the received business task request, which contains basic information about the task, such as the task type and the expected execution time of the task; the task request data is obtained by parsing the current business task request, and includes the task execution type, input data, output requirements, resource requirements, etc., wherein the output requirements include the expected execution time, output data format, processing speed, etc.; the task execution types include data query, text analysis, database update, etc.; the first resource requirement refers to the resource requirement required to execute the current task, including computing resources, storage resources, and network resources, etc.

[0072] In this embodiment, the demand analysis result refers to the first resource demand, task execution operation and target task priority sequence, wherein the target task priority sequence is obtained by sorting the received tasks from large to small according to the priority coefficient, and then adjusting the sorting according to the analysis of the mutual dependence of the tasks; the first task parameters include the computing amount, communication volume, response time and processing speed required for the task, etc.; the task execution operation includes calculation, data analysis, and data interaction.

[0073] In this embodiment, the first performance parameter includes processing speed, power consumption, cache size, execution throughput, workload, failure rate and number of restarts, etc., wherein the workload refers to the number of tasks currently being executed by the processing core and the resource occupancy rate; the core analysis result refers to the corresponding comprehensive characteristic index evaluation coefficient of the first parameter of the processing core, wherein the first parameter refers to the characteristic evaluation parameter of the task in the target task priority sequence, which is selected from the set execution-characteristic parameter mapping table according to the execution operation of the task in turn; the task scheduling strategy is composed of task execution order, task execution time and processing core allocation, and is used for task allocation; the task execution instruction refers to the execution instruction generated according to the task scheduling strategy.

[0074] The beneficial effects of the above technical solution are: by parsing the received task request, the analysis result of the resource demand analysis is combined with the evaluation result of the processing core characteristics to determine the task scheduling strategy; the task scheduling strategy is used to assign the current task to the appropriate processing core for execution, effectively realizing the efficiency of task processing, and improving resource utilization and task processing efficiency.

[0075] An embodiment of the present invention provides a heterogeneous multi-core collaborative management and automated operation and maintenance system, which prioritizes received tasks based on a first task parameter extracted from the task request data to obtain a target task priority sequence, including:

[0076] determining a priority coefficient of a current task based on a first task parameter extracted from the task request data;

[0077] Sort the received tasks according to the priority coefficients from large to small to generate a first task priority sequence;

[0078] When there is no dependency relationship between the tasks in the first task priority sequence, marking the current first task priority sequence as the target task priority sequence;

[0079] When there is a dependency relationship between the tasks in the first task priority sequence, a topological sorting algorithm is introduced to adjust the first task priority sequence to obtain a target task priority sequence.

[0080] In this embodiment, the task request data is obtained by parsing the current business task request, including the task execution type, input data, output requirements, resource requirements, etc., wherein the task execution type includes data query, text analysis, database update, etc.; the priority coefficient is obtained by weighted average of the first task parameters, wherein the first operating parameters include the amount of computing required for the task, the amount of communication, the response time, and the processing speed, etc.; the weight assigned to the first operating parameter is obtained by solving a matrix constructed by using the hierarchical analysis method to compare the first operating parameters pairwise and score their relative importance.

[0081] In this embodiment, for example, if there is a task A that depends on the output of task B, then there is a dependency relationship between task A and task B, and task B must be executed before task A.

[0082] In this embodiment, a topological sorting algorithm is used to simplify complex task dependencies into a linear execution order, so that task management becomes simpler; the task execution time refers to the expected execution time of the current task.

[0083] The beneficial effects of the above technical solution are: obtaining a first task set by dividing the received tasks according to task types; after prioritizing the tasks within the first task set, different types of tasks are mixed and sorted according to the expected execution time of the tasks to obtain a target task priority sequence, providing a basis for subsequent task scheduling and allocation, thereby improving resource utilization.

[0084] An embodiment of the present invention provides a heterogeneous multi-core collaborative management and automated operation and maintenance system, wherein the core analysis unit includes:

[0085] Current analysis block: used to extract characteristic evaluation parameters of the processing core at the current moment from the first performance parameter of the processing core at the current moment;

[0086] Prediction block: used to obtain and use the historical characteristic evaluation parameters of the processing core within a preset time period to construct a historical characteristic change curve;

[0087] Extract trend features of all the historical characteristic change curves, and use the characteristic evaluation parameters of the current processing core as input, input the pre-established characteristic prediction model and output the characteristic prediction parameters;

[0088] The obtained characteristic evaluation parameters of the processing core at the current moment are combined with the characteristic prediction parameters for analysis to obtain a comprehensive characteristic index evaluation coefficient of the characteristic evaluation parameters of the current processing core, and output it as a core analysis result;

[0089] Among them, the calculation formula of the comprehensive characteristic index evaluation coefficient is as follows:

[0090] ; In the formula, denoted as the comprehensive characteristic index evaluation coefficient of the jth characteristic evaluation parameter of the i-th processing core; It is represented as the jth characteristic evaluation parameter of the i-th processing core at the current moment; It is represented as the predicted value of the jth characteristic evaluation parameter of the i-th processing core at the next moment; T is represented as the preset time period; It is represented as the initial configuration value of the jth characteristic evaluation parameter of the i-th processing core; t is more expressed as the time interval between the current moment and the next moment; denoted as the historical characteristic variation coefficient of the jth characteristic evaluation parameter of the i-th processing core; It is represented as the impact weight of the j-th characteristic evaluation parameter of the i-th processing core on the performance of the i-th processing core.

[0091] In this embodiment, the first performance parameters include processing speed, power consumption, cache size, execution throughput, workload, failure rate and number of restarts, etc., wherein the workload refers to the number of tasks being executed by the current processing core and the resource occupancy rate; the characteristic evaluation parameters include processing speed, power consumption, execution throughput, cache size and workload, etc.; the preset time period is pre-set.

[0092] In this embodiment, the historical characteristic change curve is a characteristic evaluation parameter change curve constructed in time series with historical characteristic evaluation parameters as dependent variables; trend characteristics include curvature, slope, acceleration and the sum of the difference between extreme values ​​and average values; the historical characteristic change coefficient is obtained by weighted averaging the trend characteristics, and is used to quantify the degree of historical change of the characteristic evaluation parameters within a preset time period, wherein the weight assigned to the trend characteristics is obtained by solving a matrix constructed by pairwise comparison and relative importance scoring of the trend characteristics using the hierarchical analysis method; the characteristic prediction model is a model obtained by training a neural network based on pre-acquired training data, and is used to predict the characteristic evaluation parameters of the processing core at future moments; the characteristic prediction parameter refers to the predicted value of the characteristic evaluation parameter of the processing core at the next moment; the comprehensive characteristic index evaluation coefficient is used to evaluate the characteristic status of the processing core.

[0093] The beneficial effect of the above technical solution is that by real-time evaluation and prediction of the performance characteristics of the processing core, accurate and effective data support can be provided for subsequent reasonable and optimized task scheduling strategies, thereby improving resource utilization.

[0094] An embodiment of the present invention provides a heterogeneous multi-core collaborative management and automated operation and maintenance system, wherein the scheduling analysis block includes:

[0095] For the tasks in the target task priority sequence, characteristic evaluation parameters are selected from the set execution-characteristic parameter mapping table according to the execution operations of the tasks, and marked as the first parameters;

[0096] Extracting the corresponding comprehensive characteristic index evaluation coefficients of the first parameters of all processing cores, and calculating the fitness of each processing core with the execution operation of the current task;

[0097] The calculation formula of the fitness is as follows:

[0098] ; In the formula, It is expressed as the suitability of the i-th processing core to the execution operation of the current task; It is expressed as the comprehensive characteristic index evaluation coefficient of the sth first parameter of the i-th processing core, where s=1, 2, 3, , m; m represents the total number of the first parameter; It is represented as the influence weight of the s-th first parameter of the ith processing core on task adaptation; It is expressed as the total task execution failure rate of the i-th processing core; It is expressed as the task execution failure rate of the i-th processing core processing tasks of the same type as the current task;

[0099] The processing core having the highest degree of fitness for the execution operation of the current task is regarded as the selected processing core, and the execution operation of the current task is allocated to the selected processing core.

[0100] In this embodiment, the execution-characteristic parameter mapping table is set to be composed of the task execution operation, the task performance requirement determined based on the task execution operation, and the characteristic evaluation parameters of the processing core adapted to the task performance requirement.

[0101] In this embodiment, the task execution operation includes parallel computing, data analysis and data interaction operations, wherein, when the task execution operation is parallel computing, the task performance requirement refers to computing performance, and the characteristic evaluation parameters of the processing core adapted to the computing performance include computing speed, cache size, power consumption and parallel processing capability; when the task execution operation is data interaction, the task performance requirement is communication performance, and the characteristic evaluation parameters of the processing core adapted to the communication performance include communication bandwidth, communication delay and concurrent communication processing capability; when the task execution operation is data analysis, the task performance requirement is I / O performance, and the characteristic evaluation parameters of the processing core adapted to the I / O performance include I / O bandwidth, I / O delay and concurrent I / O processing capability.

[0102] The beneficial effect of the above technical solution is: by screening the characteristic evaluation parameters and calculating the adaptability, it is ensured that the task is assigned to the most suitable processing core to jointly promote the completion of the task, thereby improving the execution efficiency of the task.

[0103] An embodiment of the present invention provides a heterogeneous multi-core collaborative management and automated operation and maintenance system, wherein the monitoring module includes:

[0104] Monitoring unit: monitors the execution status of the current task and the resource usage of the corresponding processing core during the execution of the current task in real time to obtain execution status data;

[0105] Compare the execution status data with the corresponding set status threshold, and if there is execution status data exceeding the corresponding set status threshold, determine that the current task execution is abnormal and determine the cause of the abnormality;

[0106] According to the abnormal cause, the corresponding abnormal adjustment strategy is matched to adjust the task execution.

[0107] In this embodiment, the execution status data includes the task execution time, the execution progress and the resource usage of the current task, such as CPU usage, memory usage, I / O read and write rate, and network bandwidth; the setting status threshold includes setting the time threshold and the setting usage range of various resources; the abnormal reasons include task execution timeout, abnormal increase in resource usage, task suspension and task deadlock, etc., which are based on the real-time monitoring of the execution status of the current task and the resource usage of the corresponding processing core during the execution of the current task, and are obtained by analyzing the abnormal recognition model, wherein the abnormal recognition model is a model for identifying the abnormal cause obtained by training the neural network with the historical abnormal data of different processing cores as training data; the abnormal adjustment strategy includes task readjustment, processing core resource reallocation and task configuration parameter adjustment, etc., which is obtained by taking the acquired abnormal cause as the input, inputting it into the strategy matching model and then outputting it, wherein the strategy matching model is a model for matching the strategy for handling abnormalities obtained by training the neural network with the historical abnormal processing data of different processing cores as training data.

[0108] In this embodiment, determining whether there is an abnormality in the execution of the current task includes:

[0109] Subtract the resource usage of the processing core at the previous moment from the resource usage at the current moment to obtain a first usage difference;

[0110] If the first usage difference exceeds the set usage range, it is determined that the current processing core resource utilization is abnormal and the current task execution is abnormal;

[0111] If the task execution time of the current task is greater than the set time threshold, the current task execution is determined to be abnormal.

[0112] The beneficial effect of the above technical solution is: by real-time monitoring of the execution status of the current task and the resource usage of the processing core, and taking corresponding adjustment strategies when anomalies are found, it is beneficial to improve the success rate of task execution, optimize resource utilization and enhance user experience.

[0113] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A heterogeneous multi-core collaborative management and automated operation and maintenance system, characterized in that: include: Resource processing module: used to identify the processing core and unify the API interface; Scheduling module: used to intelligently schedule the target processing core to complete the task corresponding to the current task request based on the analysis of the current task request and the processing core characteristics; Monitoring module: used to monitor the execution status of tasks in real time, and take corresponding adjustment measures when abnormal execution status is detected; Among them, the target processing cores include chip servers, distributed storage devices, network devices and security devices; Wherein, the scheduling module includes: Demand analysis unit: used to receive and parse the current task request to obtain task request data; Prioritizing tasks in the received task request data based on the first task parameter extracted from the task request data to obtain a target task priority sequence; Extracting the first resource requirement and the task execution operation from the task request data, and outputting them together with the target task priority sequence as the requirement analysis result; Core analysis unit: used for evaluating the characteristics of each processing core based on the first performance parameters of the processing core at the current moment and the historical moment collected, and obtaining the core analysis result; Scheduling unit: used to generate a task scheduling strategy using a scheduling analysis block; and to schedule multiple processing cores to jointly complete a task according to the task scheduling strategy using a scheduling implementation block; Scheduling analysis block: It is used to combine the demand analysis results of the task with the core analysis results, allocate the processing cores to the tasks in order according to the priority sequence of the target tasks, and generate the task scheduling strategy; Scheduling implementation block: used to generate task execution instructions according to the task scheduling strategy and transmit them to the corresponding processing core for task execution; Wherein, the core analysis unit includes: Current analysis block: used to extract characteristic evaluation parameters of the processing core at the current moment from the first performance parameter of the processing core at the current moment; Prediction block: used to obtain and use the historical characteristic evaluation parameters of the processing core within a preset time period to construct a historical characteristic change curve; Extract trend features of all the historical characteristic change curves, and use the characteristic evaluation parameters of the current processing core as input, input the pre-established characteristic prediction model and output the characteristic prediction parameters; The obtained characteristic evaluation parameters of the processing core at the current moment are combined with the characteristic prediction parameters for analysis to obtain a comprehensive characteristic index evaluation coefficient of the characteristic evaluation parameters of the current processing core, and output it as a core analysis result; Among them, the calculation formula of the comprehensive characteristic index evaluation coefficient is as follows: ; In the formula, denoted as the comprehensive characteristic index evaluation coefficient of the jth characteristic evaluation parameter of the i-th processing core; It is represented as the jth characteristic evaluation parameter of the i-th processing core at the current moment; It is represented as the predicted value of the jth characteristic evaluation parameter of the i-th processing core at the next moment; T is represented as the preset time period; It is represented as the initial configuration value of the jth characteristic evaluation parameter of the i-th processing core; t is more expressed as the time interval between the current moment and the next moment; denoted as the historical characteristic variation coefficient of the jth characteristic evaluation parameter of the i-th processing core; It is represented as the weight of the jth characteristic evaluation parameter of the ith processing core on the performance of the ith processing core; Wherein, the scheduling analysis block includes: For the tasks in the target task priority sequence, characteristic evaluation parameters are selected from the set execution-characteristic parameter mapping table according to the execution operations of the tasks, and marked as the first parameters; Extracting the corresponding comprehensive characteristic index evaluation coefficients of the first parameters of all processing cores, and calculating the fitness of each processing core with the execution operation of the current task; The calculation formula of the fitness is as follows: ; In the formula, It is expressed as the suitability of the i-th processing core to the execution operation of the current task; It is expressed as the comprehensive characteristic index evaluation coefficient of the sth first parameter of the i-th processing core, where s=1, 2, 3, , m; m represents the total number of the first parameter; It is represented as the influence weight of the s-th first parameter of the ith processing core on task adaptation; It is expressed as the total task execution failure rate of the i-th processing core; It is expressed as the task execution failure rate of the i-th processing core processing tasks of the same type as the current task; The processing core having the highest degree of fitness for the execution operation of the current task is regarded as the selected processing core, and the execution operation of the current task is allocated to the selected processing core.

2. A heterogeneous multi-core collaborative management and automated operation and maintenance system according to claim 1, characterized in that: The resource processing module includes: Resource pool building unit: used to develop the hardware abstraction layer, identify the target processing core in the target cloud platform and register it; Virtualize all target processing cores and centrally manage them to form a shared resource pool; Interface unification unit: used to provide a unified API interface for the system and develop a software abstraction layer so that applications can use the same interface to access resources.

3. A heterogeneous multi-core collaborative management and automated operation and maintenance system according to claim 1, characterized in that: Based on the first task parameter extracted from the task request data, the received tasks are prioritized to obtain a target task priority sequence, including: Determining a priority coefficient of a task corresponding to the current task request data based on a first task parameter extracted from the task request data; Sort the tasks corresponding to the received task request data according to the priority coefficients from large to small to generate a first task priority sequence; When there is no dependency relationship between the tasks in the first task priority sequence, marking the current first task priority sequence as the target task priority sequence; When there is a dependency relationship between the tasks in the first task priority sequence, a topological sorting algorithm is introduced to adjust the first task priority sequence to obtain a target task priority sequence.

4. The heterogeneous multi-core collaborative management and automated operation and maintenance system according to claim 1, characterized in that: The monitoring module comprises: Monitoring unit: used to monitor the execution status of the current task and the resource usage of the corresponding processing core during the execution of the current task in real time, and obtain the execution status data; Compare the execution status data with the corresponding set status threshold, and if there is execution status data exceeding the corresponding set status threshold, determine that the current task execution is abnormal and determine the cause of the abnormality; According to the abnormal cause, the corresponding abnormal adjustment strategy is matched to adjust the task execution.

Images