Cluster state evaluation method, storage medium, electronic device, and program product

Abstract

Embodiments of the present application provide a cluster state evaluation method and a storage medium, an electronic device and a program product, and relate to the field of computers, wherein the method comprises: obtaining a state evaluation request triggered by a cluster, wherein the state evaluation request carries a template request parameter and a data request parameter; obtaining evaluation data of the cluster based on the data request parameter, wherein the evaluation data is cluster data that meets a data range indicated by the data request parameter; in the case of obtaining an evaluation template matching the template request parameter, obtaining a plurality of evaluation elements included in the evaluation template, wherein one evaluation element corresponds to one evaluation style; generating a plurality of evaluation results for the evaluation data according to each evaluation style, wherein one evaluation result corresponds to one evaluation element, and the plurality of evaluation results are used to evaluate the running state of the cluster. The embodiments of the present application solve the technical problem of low state evaluation efficiency of the cluster in the related art.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the computer field, in particular, to a cluster state evaluation method and a storage medium, an electronic device and a program product. BACKGROUND

[0002] In the existing cluster running state evaluation method, the process of obtaining and analyzing cluster state data highly depends on manual operation, which not only consumes a large amount of time and resources, but also easily introduces subjective errors, resulting in low efficiency of cluster running state evaluation.

[0003] Therefore, there is a technical problem of low efficiency of cluster running state evaluation in the related art. SUMMARY

[0004] Embodiments of the present application provide a cluster state evaluation method and a storage medium, an electronic device and a program product to at least solve the technical problem of low efficiency of cluster running state evaluation in the related art.

[0005] According to an embodiment of the present application, a cluster state evaluation method is provided, comprising: obtaining a state evaluation request triggered by a cluster, wherein the state evaluation request carries a template request parameter and a data request parameter; obtaining evaluation data of the cluster based on the data request parameter, wherein the evaluation data is cluster data conforming to a data range indicated by the data request parameter; in the case of obtaining an evaluation template matching the template request parameter, obtaining a plurality of evaluation elements included in the evaluation template, wherein one evaluation element corresponds to one evaluation style; generating a plurality of evaluation results for the evaluation data according to each evaluation style, wherein one evaluation result corresponds to one evaluation element, and the plurality of evaluation results are used to evaluate the running state of the cluster.

[0006] According to another embodiment of the present application, a cluster state evaluation device is provided, comprising: a first obtaining unit configured to obtain a state evaluation request triggered by a cluster, wherein the state evaluation request carries a template request parameter and a data request parameter; a second obtaining unit configured to obtain evaluation data of the cluster based on the data request parameter, wherein the evaluation data is cluster data conforming to a data range indicated by the data request parameter; a third obtaining unit configured to, in the case of obtaining an evaluation template matching the template request parameter, obtain a plurality of evaluation elements included in the evaluation template, wherein one evaluation element corresponds to one evaluation style; and an evaluation unit configured to generate a plurality of evaluation results for the evaluation data according to each evaluation style, wherein one evaluation result corresponds to one evaluation element, and the plurality of evaluation results are used to evaluate the running state of the cluster.

[0007] According to still another embodiment of the present application, a computer readable storage medium is also provided, in which a computer program is stored, wherein the computer program is configured to perform the steps of any of the method embodiments described above when executed.

[0008] According to still another embodiment of the present application, an electronic device is also provided, comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the computer program to perform the steps of any of the method embodiments described above.

[0009] Through the embodiments provided by the present application, by receiving a state evaluation request carrying a template request parameter, a corresponding evaluation template can be flexibly applied, the evaluation template includes evaluation elements and styles that are of interest to users, so that the evaluation result is closer to the actual demand, and the evaluation flexibility of the running state of the cluster is enhanced. Based on the data request parameter, the cluster evaluation data conforming to the specified data range is automatically obtained, avoiding the tedious process of manual data collection. Through automatic data acquisition, the timeliness and integrity of the data are ensured, and the time consumption and human errors of data processing are reduced. Based on the evaluation data, the corresponding evaluation result can be generated according to the evaluation style of each evaluation element, which not only greatly improves the automation level of the cluster state evaluation, but also can achieve diversified evaluation methods (multiple evaluation results, respectively corresponding to multiple evaluation styles) for a piece of evaluation data, enhances the flexibility and comprehensiveness of the evaluation report, realizes the technical effect of comprehensively improving the evaluation efficiency of the running state of the cluster, and solves the technical problem of low evaluation efficiency of the running state of the cluster in related technologies. BRIEF DESCRIPTION OF DRAWINGS

[0010] Figure 1 is a hardware structure block diagram of a cluster state evaluation method according to an embodiment of the present application;

[0011] Figure 2 is a flowchart of a cluster state evaluation method according to an embodiment of the present application;

[0012] Figure 3 is a flowchart of a generation method of a running report of an artificial intelligence platform cluster according to an embodiment of the present application;

[0013] Figure 4 is a processing flow of a generation method of a running report of an artificial intelligence platform cluster according to an embodiment of the present application;

[0014] Figure 5 is a structure block diagram of a cluster state evaluation device according to an embodiment of the present application. DETAILED DESCRIPTION

[0015] In the following, embodiments of the present application will be described in detail with reference to the accompanying drawings and in conjunction with embodiments.

[0016] It should be noted that the terms "first", "second", and the like in the description and in the claims of the present application and above-described accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a particular sequential or chronological order. It should be understood that the data thus used can be interchanged, where appropriate, so that the embodiments of the present application described herein can be carried out in other than the order shown or described herein. Furthermore, the terms "comprise" and "have", and any variations thereof, are intended to cover a non-exclusive inclusion, for example, a process, method, system, product, or apparatus that comprises a list of steps or units is not necessarily limited to those steps or units that are clearly listed, but can include other steps or units that are not clearly listed or inherent to such processes, methods, products, or apparatuses.

[0017] The method embodiments provided in the embodiments of the present application can be executed in a computer terminal or similar computing device. Taking a computer terminal as an example, Figure 1 is a hardware structure block diagram of a computer terminal of a cluster state evaluation method of the embodiments of the present application. As Figure 1 shown, the computer terminal can include one or more (only one is shown in Figure 1 ) processor 102 (the processor 102 can include, but is not limited to, a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data, wherein the above-mentioned computer terminal can further include a transmission device 106 for communication function and an input and output device 108. Those skilled in the art can understand that Figure 1 the structure shown is only schematic, which does not limit the structure of the above-mentioned computer terminal. For example, the computer terminal can further include more or less components than Figure 1 shown, or have a different configuration from Figure 1 shown.

[0018] The memory 104 can be used to store computer programs, for example, software programs of application software and modules, such as the computer program corresponding to the cluster state evaluation method in the embodiments of the present application. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, implements the above-mentioned method. The memory 104 can include a high-speed random access memory, and can also include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some examples, the memory 104 can further include a memory remotely arranged with respect to the processor 102, which can be connected to the computer terminal through a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0019] The transmission device 106 is configured to receive or send data via a network. The network can include a wireless network provided by a communication service provider of a computer terminal. In one example, the transmission device 106 includes a network interface controller (NIC) that can be connected to other network devices through a base station to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module configured to communicate with the Internet wirelessly.

[0020] As an optional solution, the embodiment provides a cluster state evaluation method, as shown in the following table, which includes the following steps: Figure 2

[0021] S202, obtaining a state evaluation request triggered by the cluster, wherein the state evaluation request carries template request parameters and data request parameters;

[0022] S204, obtaining evaluation data of the cluster based on the data request parameters, wherein the evaluation data is cluster data that meets a data range indicated by the data request parameters;

[0023] S206, obtaining a plurality of evaluation elements included in the evaluation template in a case where the evaluation template matching the template request parameters is obtained, wherein one evaluation element corresponds to one evaluation style;

[0024] S208, generating a plurality of evaluation results for the evaluation data according to each evaluation style, wherein one evaluation result corresponds to one evaluation element, and the plurality of evaluation results are used to evaluate the running state of the cluster.

[0025] Optionally, in the embodiment, the state evaluation request is an instruction issued by a user or an automated system, aiming to request a detailed cluster state evaluation report. The request usually carries template request parameters and data request parameters to indicate the structure of the report and the range of required data.

[0026] Optionally, in the embodiment, the template request parameters specify the template style used to generate the report. The template defines the type (such as labels, tables, charts), layout, and possible intelligent analysis rules of each element in the report, ensuring that the report content is comprehensive and meets the specific needs of the user.

[0027] Optionally, in the embodiment, the data request parameters specify the time range, cluster nodes, and specific requirements of the concerned indicators for the evaluation, ensuring that the obtained data accurately reflects the cluster state and time period that the user wants to evaluate.

[0028] ​Optionally, in this embodiment, the evaluation data is historical or real-time running data that meets specific data ranges, which is obtained from the cluster according to the data request parameters. The evaluation data is the basis for generating evaluation results, and its accuracy directly affects the credibility of the evaluation report.

[0029] Optionally, in this embodiment, the evaluation template is a predefined structure used to guide the content, layout, and presentation form of the evaluation report. The evaluation template contains multiple evaluation elements and their styles customized by the user, ensuring that the report is generated in accordance with standard specifications and personalized needs.

[0030] Optionally, in this embodiment, the evaluation element is the basic unit of the report, and each element represents a specific type of information display, such as labels for text descriptions, tables for data listing, and charts (bar charts, line charts, pie charts, etc.) for intuitive data visualization.

[0031] Optionally, in this embodiment, the evaluation style is the presentation form and rules associated with the evaluation element, such as data color coding, chart layout design, and abnormal data highlighting methods. The evaluation style enhances the readability of the evaluation results and the efficiency of information transmission.

[0032] Optionally, in this embodiment, the evaluation result is various evaluation information generated from the evaluation data according to the evaluation style, and each result corresponds to an evaluation element, aiming to comprehensively reflect the running state of the cluster in a specific time period, including but not limited to performance indicators, usage trends, and abnormal situations.

[0033] Optionally, in this embodiment, a state evaluation request containing template request parameters and data request parameters is received. The request can be initiated manually by the user through the interface or automatically triggered by the system according to the pre-set timing evaluation strategy. The template request parameters and data request parameters jointly define the appearance and content boundaries of the report, which are important basis for subsequent data acquisition and report generation.

[0034] According to the time range, data type, and specific indicators in the data request parameters, the required data is extracted from the cluster's monitoring system or logs. The evaluation data here is filtered and preprocessed to ensure that only data points that match the request parameters are included, facilitating subsequent analysis and display.

[0035] Once the data is obtained, the evaluation template matching the template request parameters will be found. After finding the correct template, all pre-defined evaluation elements are parsed from the template, ready to combine these elements with the evaluation data to generate specific evaluation results. The diversity of evaluation elements ensures that the report can comprehensively and meticulously reflect the cluster state.

[0036] After all the evaluation elements and their styles are determined, the evaluation data is analyzed and the corresponding evaluation results are generated according to the style rules of each evaluation element. For example, if the style rule is to make special markings when a certain performance indicator exceeds the threshold, the evaluation results will be visually highlighted. The generation of evaluation results takes into account the depth analysis and intelligent perception of data, and presents the evaluation focus and findings in a more intuitive and effective way.

[0037] It can be understood that the embodiment can highly customize the style and content of the report according to the user's needs, while automatically obtaining data from the cluster and conducting in-depth analysis. This method not only reduces the manual burden of the operation and maintenance personnel and improves the generation efficiency of the evaluation report, but also enhances the reaction speed of the report to the abnormal state of the cluster through the application of intelligent analysis rules.

[0038] Through the embodiments provided by the present application, by receiving a state evaluation request carrying a template request parameter, the corresponding evaluation template can be flexibly applied, the evaluation template includes evaluation elements and styles that the user is interested in, so that the evaluation results are more close to the actual needs, and the evaluation flexibility of the running state of the cluster is enhanced. Based on the data request parameter, the cluster evaluation data conforming to the specified data range is automatically obtained, avoiding the tedious process of manual data collection. Through automatic data acquisition, the timeliness and completeness of the data are ensured, and the time consumption and human error of data processing are reduced. Based on the evaluation data, the corresponding evaluation results can be generated according to the evaluation style of each evaluation element, which not only greatly improves the automation level of the cluster state evaluation, but also can achieve diversified evaluation methods (multiple evaluation results, respectively corresponding to multiple evaluation styles) for a piece of evaluation data, enhances the flexibility and comprehensiveness of the evaluation report, and realizes the technical effect of comprehensively improving the evaluation efficiency of the running state of the cluster.

[0039] As an optional solution, after generating multiple evaluation results for the evaluation data according to each evaluation style, the method further includes:

[0040] The multiple evaluation results are integrated to obtain an evaluation report for evaluating the running state of the cluster, wherein the first arrangement order of the multiple evaluation results on the evaluation report corresponds to the second arrangement order of the multiple evaluation elements on the evaluation template.

[0041] Optionally, in the embodiment, the first arrangement order is the display order of each evaluation result in the generated evaluation report. This order usually follows the arrangement logic of the evaluation elements in the evaluation template, ensuring the coherence of the report content and the convenience of user reading.

[0042] Optionally, in this embodiment, the second arrangement order is the arrangement order of the evaluation elements defined in the evaluation template, which determines the order of the contents in the evaluation report, helping to clarify the report structure and rationalize the content organization.

[0043] Optionally, in this embodiment, after completing the specific analysis and result generation of the evaluation data according to each evaluation style, the next important step is to integrate these scattered evaluation results to build a complete evaluation report. This process includes sequentially combining all evaluation results according to the arrangement order of elements in the pre-defined evaluation template, ensuring the logicality of the report content and the visual coherence.

[0044] The integration process not only involves the physical merging of data, but more importantly, it ensures that the display method of the evaluation results matches the definition in the evaluation template, that is, the presentation form of each evaluation result in the evaluation report (such as label, table, or chart) is consistent with the layout in the evaluation template. The benefit of this processing is that the evaluation report can clearly and intuitively show the running status of the cluster in all aspects, making it easy for users to quickly review and understand.

[0045] For example, if the arrangement order of evaluation elements in the evaluation template is CPU usage (label), memory occupation (table), and network traffic (column chart) in sequence, then in the generated evaluation report, these three evaluation results will also appear in the same order, that is, first a label showing CPU usage, then a table showing memory occupation, and finally a column chart showing network traffic changes. This integration based on the preset template order helps users quickly locate the information of interest, improving the practical value and user experience of the report.

[0046] Through the embodiments provided by the present application, the evaluation report not only collects the depth analysis results for each key indicator of the cluster, but also ensures that the organization logic and visual presentation of the report content perfectly match the design intention of the original template, thereby maximizing the integrity of the evaluation information and the visual style customized by the user, providing a cluster state evaluation solution that combines professionalism and personalization.

[0047] As an optional solution, after integrating the multiple evaluation results to obtain an evaluation report for evaluating the running status of the cluster, the method further includes:

[0048] displaying the first evaluation content in the evaluation report according to a first display style, and displaying the second evaluation content in the evaluation report according to a second display style, wherein the first evaluation content is the evaluation content in the multiple evaluation results that meets the expected abnormal condition, the second evaluation content is the evaluation content in the multiple evaluation results other than the first evaluation content, and the first display style is different from the second display style.

[0049] Optionally, in this embodiment, the first evaluation content is the information identified as meeting the expected abnormal condition in the evaluation report based on the evaluation results. These contents generally represent the abnormal or deviated normal range of the cluster running state, which requires special attention from the operation and maintenance personnel or managers.

[0050] Optionally, in this embodiment, the second evaluation content is the other evaluation information in the evaluation report except the first evaluation content, that is, the state data of the cluster within the normal operation range. This part of the content provides a comprehensive perspective of the cluster performance, although it is not highlighted as an abnormal situation, but is also important for the overall evaluation.

[0051] Optionally, in this embodiment, the first display style is a visual presentation method specially used to highlight the first evaluation content. Generally, the first display style will use more eye-catching colors, bold fonts, special symbols or background highlights, etc. to make the abnormal information stand out in the report, so as to quickly locate.

[0052] Optionally, in this embodiment, the second display style is a visual style used to display the second evaluation content, which is significantly different from the first display style. The design of the second display style pays more attention to the clear display of information and the aesthetics of the overall report, using conventional fonts, colors and layouts to ensure the reading fluency and professionalism of the report content.

[0053] Optionally, in this embodiment, after the integration of the evaluation report is completed, the system will identify and mark all the first evaluation content, that is, the evaluation results that meet the expected abnormal condition. Then, the system uses the pre-defined first display style to specially process these abnormal information, so that they are presented in a more eye-catching form in the report. The first display style may include but is not limited to using red highlights, bold text, adding warning icons, etc. to ensure that the operation and maintenance personnel can quickly notice the abnormal situation when reviewing the report.

[0054] For the evaluation results of the normal running state other than the first evaluation content, that is, the second evaluation content, the system will display them according to the second display style. The second display style is more conventional compared to the first display style, aiming to provide clear information display while maintaining the overall visual harmony of the report. This may include using default fonts, standard color coding, appropriate spacing, etc. to ensure that all evaluation information can be reasonably and professionally presented in the report, facilitating users to comprehensively understand the running status of the cluster.

[0055] Through the embodiments provided in the present application, the abnormal and normal information in the evaluation report is efficiently and intuitively conveyed through differentiated display styles. By adopting different first display styles and second display styles, the present application can significantly enhance the information differentiation of the evaluation report, so that the user can quickly identify which indicators need special attention and which belong to the data in the normal operation state of the cluster.

[0056] As an optional solution, before displaying the first evaluation content in the evaluation report according to the first display style and displaying the second evaluation content in the evaluation report according to the second display style, the method further comprises:

[0057] obtaining the parameter type carried in the state evaluation request and the parameter threshold associated with the parameter type;

[0058] determining an expected abnormal condition according to the parameter type and the parameter threshold, wherein the expected abnormal condition is used to indicate that the parameter type is greater than the parameter threshold;

[0059] determining the first evaluation content meeting the expected abnormal condition from the plurality of evaluation results.

[0060] Optionally, in the present embodiment, the parameter type is a performance indicator category associated with a specific evaluation result in the evaluation report, such as CPU usage, memory occupancy, network transmission rate, etc. These parameter types define the types and focus points of the evaluation content.

[0061] Optionally, in the present embodiment, the parameter threshold is a numerical limit associated with the parameter type, which is used to determine whether the evaluation result is abnormal. For example, if the parameter threshold of CPU usage is set to 85%, then any usage rate exceeding 85% will be considered abnormal.

[0062] Optionally, in the present embodiment, before the generation and display of the evaluation report, the system first needs to parse the parameter type and the set parameter threshold of the user's attention from the state evaluation request. These information is usually obtained through the template request parameter and the data request parameter in the request, which defines the structure and data filtering standard of the evaluation report.

[0063] The system automatically generates an expected abnormal condition according to the received parameter type and the user-specified parameter threshold. The expected abnormal condition is a logical judgment rule used to identify which indicators in the evaluation result are abnormal. For example, if the parameter type is CPU usage and the parameter threshold is 85%, then the expected abnormal condition is "CPU usage greater than 85%".

[0064] According to the defined expected abnormal condition, the system filters the multiple evaluation results completed by integration, identifies all evaluation information meeting the condition, i.e., the first evaluation content. This process automatically completes the marking and classification of abnormal data, ensuring the subsequent highlighting and detailed analysis of abnormal information.

[0065] Through the embodiments provided in the present application, it is ensured that the evaluation report can not only be efficiently generated, but also visually and intuitively distinguish abnormal indicators and normal indicators, significantly improving the efficiency and accuracy of operation and maintenance personnel and managers when evaluating the running state of the cluster. The setting of parameter types and thresholds allows users to define the boundaries of abnormalities according to their own needs and experience, enabling the system to more intelligently detect and highlight abnormal situations, which helps to quickly respond and solve problems and maintain the stable operation of the cluster.

[0066] As an optional solution, multiple evaluation results are generated for the evaluation data according to each evaluation style, including at least one of the following:

[0067] In the case where the multiple evaluation elements include the label evaluation element, a label evaluation result corresponding to the label evaluation element is generated for the evaluation data, and the multiple evaluation results include the label evaluation result;

[0068] In the case where the multiple evaluation elements include the table evaluation element, a table evaluation result corresponding to the table evaluation element is generated for the evaluation data, and the multiple evaluation results include the table evaluation result;

[0069] In the case where the multiple evaluation elements include the graphical evaluation element, a graphical evaluation result corresponding to the graphical evaluation element is generated for the evaluation data, and the multiple evaluation results include the graphical evaluation result.

[0070] Optionally, in this embodiment, when the evaluation element is a label, the generated evaluation result generally contains a brief textual description for intuitively displaying a specific evaluation indicator or running state, such as a cluster health overview.

[0071] Optionally, in this embodiment, in the case where the evaluation element is a table, the evaluation result is a detailed data set in the form of rows and columns, which facilitates users to view and compare performance indicators and state data of multiple nodes or multiple time points in the cluster, such as CPU usage and memory occupancy.

[0072] Optionally, in this embodiment, when the evaluation element is a graph, the evaluation result is intuitively presented in the form of a chart, including a column chart, a line chart, a pie chart, etc., for displaying the distribution, trend and proportion of the evaluation data, so that complex data can be understood at a glance, for example, the trend of cluster resource usage over time.

[0073] Optionally, in this embodiment, if the label evaluation element is included in the evaluation template, the next step of the method will be to process and generate the corresponding label evaluation result. This process usually involves extracting key indicator values from the evaluation data and converting them into the form of labels, such as "Cluster Average CPU Usage: 82%". Label evaluation results are concise and clear, and can quickly convey key information, suitable for the beginning or summary part of the report, providing users with an overview of the current status of the cluster.

[0074] When a table evaluation element is defined in the evaluation template, the system will select the appropriate data set from the evaluation data according to the data request parameters, and organize it into a table evaluation result. The table evaluation result lists the performance indicators and status data of the cluster in detail, allowing users to conduct in-depth analysis and comparison of the data. The system will ensure that the format of the table meets the requirements of the template, such as specific column titles, row data sorting, and possible intelligent analysis markers (such as highlighting abnormal data).

[0075] If a graphical evaluation element such as a line chart, bar chart or pie chart is used in the evaluation template, the system will generate the corresponding graphical evaluation result according to the evaluation data. This process involves statistical analysis and graphical display of data. The system will map the evaluation data to the X-axis, Y-axis or other dimensions of the chart to generate an intuitive graphical display, while applying user-defined intelligent analysis rules (such as marking points that exceed the normal range on the chart), so that the graphical evaluation result is not only intuitive, but also highlights important or abnormal information.

[0076] Through the embodiments provided in this application, corresponding evaluation results are generated according to different types of evaluation elements (labels, tables, graphics). This process ensures that the evaluation report can present the evaluation data in the most suitable form according to the specific needs and preferences of the user. Optionally, label evaluation results provide a way to quickly review the report, allowing users to have a preliminary understanding of the overall health of the cluster; table evaluation results provide a detailed view of the data, suitable for data comparison and historical trend analysis; graphical evaluation results display the distribution and changes of evaluation data in an intuitive visual form, helping to quickly grasp key trends and abnormal situations. By flexibly using different types of evaluation elements, not only the readability and information transmission efficiency of the report are improved, but also the professionalism and comprehensiveness of the evaluation report are ensured.

[0077] As an optional solution, in the case of multiple evaluation elements including a graphical evaluation element, a graphical evaluation result corresponding to the graphical evaluation element is generated for the evaluation data, including at least one of the following:

[0078] In the case of a line chart element as a graphical evaluation element, a line chart evaluation result corresponding to the line chart element is generated for the evaluation data, and the graphical evaluation result includes the line chart evaluation result.

[0079] In the case that the graphical evaluation element is a pie chart element, the system generates a pie chart evaluation result corresponding to the pie chart element for the evaluation data. The graphical evaluation result includes the pie chart evaluation result.

[0080] In the case that the graphical evaluation element is a column chart element, the system generates a column chart evaluation result corresponding to the column chart element for the evaluation data. The graphical evaluation result includes the column chart evaluation result.

[0081] Optionally, in this embodiment, when the graphical evaluation element is set to a line chart, the system generates a dynamic trend chart based on the evaluation data. The line chart evaluation result typically shows the trend of a specific indicator over time, such as the fluctuation of cluster resource usage.

[0082] Optionally, in this embodiment, when the graphical evaluation element is a pie chart, the system generates an evaluation result that shows the proportion of each component in the overall. The pie chart evaluation result is often used to reflect the allocation of cluster resources, such as the proportion of CPU usage of different nodes.

[0083] Optionally, in this embodiment, when the evaluation uses a column chart element, the generated evaluation result visually displays the comparison relationship between data in the form of a column chart. The column chart evaluation result is suitable for displaying the performance indicator differences between different time periods or different nodes, such as the average memory usage of each node.

[0084] Optionally, in this embodiment, if the evaluation template contains a line chart element, the system will generate a line chart evaluation result based on the evaluation data. This process involves mapping the time series data in the data set to the X-axis (time) and Y-axis (indicator value) of the line chart, and drawing a continuous line to show the trend of the indicator over time. The system will ensure that the generation of the line chart evaluation result meets the display requirements set by the user in the template, such as the time accuracy of the X-axis and the numerical range of the Y-axis, while possibly applying intelligent analysis rules, such as special marking of points exceeding the threshold, so that the line chart evaluation result is not only intuitive but also highlights abnormal data points.

[0085] In the case that a pie chart element is defined in the evaluation template, the system will generate a pie chart evaluation result based on the evaluation data. This process requires mapping the classification data and its proportion in the data set to the pie chart, with each classification corresponding to a sector in the pie chart, and the size reflecting the proportion of the classification in the overall data. The design of the pie chart evaluation result will follow the user's requirements in the template, such as the display method of classification labels and the color scheme of the pie chart, to ensure the accuracy of information transmission and the aesthetics of the report.

[0086] If the evaluation template contains a histogram element, the system will generate a histogram evaluation result to show the comparison or distribution of the evaluation data. The generation of the histogram evaluation result involves mapping the classification data or time data in the data set to each column of the histogram, and the height or length of the column reflects the numerical value of the data. The system will set appropriate classification labels, numerical axis ranges, and column styles for the histogram evaluation result according to the template requirements, and may apply intelligent analysis rules such as highlighting numerical abnormal columns to enhance the visualization of information and the ability to detect abnormalities.

[0087] Through the embodiments provided in the present application, the evaluation report can not only provide an intuitive view of various types of data, but also select the most suitable chart type for display according to the characteristics of the data and the preferences of the user, thereby improving the readability of the report and the depth of data analysis. Whether it is to show the time series trend, the proportion relationship of each part, or the comparison of data of different categories, line charts, pie charts, and histograms can all convey key information of evaluation data in the most intuitive way, helping users quickly understand the details of the cluster running state and providing strong data support for subsequent operation and maintenance decisions and fault troubleshooting.

[0088] As an optional solution, the evaluation data of the cluster is obtained based on the data request parameter, including:

[0089] In the case where the data request parameter includes a time request sub-parameter, the first data generated by the cluster in the time range indicated by the time request sub-parameter is determined as the evaluation data; or,

[0090] In the case where the data request parameter includes a node request sub-parameter, the second data generated by the target node indicated by the node request sub-parameter in the cluster is determined as the evaluation data.

[0091] Optionally, in the present embodiment, the time request sub-parameter is a sub-parameter in the data request parameter used to specify the time range of the evaluation data. It includes a start time and an end time, which are used to limit the time window of the data in the evaluation report, ensuring that the report reflects the running state of the cluster in a specific time period.

[0092] Optionally, in the present embodiment, the first data is the evaluation data collected by the system in the specified time range according to the time request sub-parameter in the case where the data request parameter contains the time request sub-parameter, which is used to generate report content reflecting the change of the cluster running state over time.

[0093] Optionally, in the present embodiment, the node request sub-parameter is a sub-parameter in the data request parameter used to specify the target node. It allows the user or system to focus on the running data of a specific node when generating the evaluation report, rather than the entire cluster.

[0094] Optionally, in this embodiment, the second data is the evaluation data of the specific target node collected by the system when the data request parameter includes the node request sub-parameter, which is used to generate the report content reflecting the unique running state of the node.

[0095] Optionally, in this embodiment, when the time request sub-parameter is included in the data request parameter, the system first parses the parameter values of the start time and the end time, and then obtains the first data generated in the time range from the cluster. This process involves filtering historical data and monitoring real-time data, ensuring that the evaluation report reflects the running status of the cluster in the time window concerned by the user. For example, if the time request sub-parameter is last week, the system will collect the CPU usage, memory usage, network traffic, and other data of all nodes in that week to generate a running state evaluation report for the last week.

[0096] If the node request sub-parameter is included in the data request parameter, the system will focus on the specific target node and collect the second data generated by the node during the evaluation period. This step allows the report generation process to be more targeted, facilitating in-depth understanding and analysis of the running status of the specific node by the user.

[0097] Through the embodiments provided in this application, the process of obtaining the first data based on the time request sub-parameter ensures the timeliness of the evaluation report content, allowing the user to obtain the running state information in a specific period, which helps to discover and solve the performance problems of the cluster in a timely manner. Obtaining the second data based on the node request sub-parameter highlights the individualization and targeting of the evaluation report, allowing the operation and maintenance personnel to deeply analyze the running status of a single or multiple specific nodes, providing data support for the diagnosis and optimization of node-level problems.

[0098] As an optional solution, based on the data request parameter, the evaluation data of the cluster is obtained, and the method further includes:

[0099] When the data request parameter includes the time request sub-parameter and the node request sub-parameter, the third data generated by the target node in the cluster within the time range is determined as the evaluation data.

[0100] Optionally, in this embodiment, when the data request parameter includes the time request sub-parameter and the node request sub-parameter, the next step of the evaluation report generation method is to collect and determine the third data. This process involves accurate acquisition of the running data of the specific node within the specified time. The system first identifies the time request sub-parameter to determine the time range for data collection, and then locates the target node indicated by the node request sub-parameter.

[0101] The system will filter out the running indicators such as CPU usage, memory occupation, disk I / O, network bandwidth, etc. in a set time range from the historical and real-time data stream of the target node, and the series of filtered data is the third data. Through accurate time and node positioning, the third data can provide detailed running conditions of the target node in a specific time period, which is extremely crucial for in-depth analysis of node-level performance problems, resource bottlenecks or fault sources.

[0102] Through the combination of the time request sub-parameter and the node request sub-parameter, the multi-dimensional customization of the evaluation report generation is realized, the support of the operation and maintenance decision is greatly improved, and the efficiency of problem solving is also improved. At the same time, the application also highlights the strong function of the application to provide fine operation and maintenance assistance.

[0103] As an optional solution, the method further comprises, before obtaining the state evaluation request triggered by the cluster:

[0104] The at least two evaluation elements are combined in order to generate at least two evaluation templates, wherein the number of evaluation elements contained in different evaluation templates is different, or the combination order of the evaluation elements is different;

[0105] A one-to-one mapping relationship is established between the at least two template request parameters and the at least two evaluation templates;

[0106] Before obtaining the plurality of evaluation elements included in the evaluation template, the method further comprises:

[0107] The evaluation template matched by the template request parameter is determined from the mapping relationship.

[0108] Optionally, in the embodiment, before the evaluation report is generated, the system needs to define a plurality of evaluation templates in advance, each template being combined by at least two evaluation elements in a certain order and format. This process may involve permutation and combination of a plurality of evaluation elements, to ensure that different evaluation templates can meet the needs of different scenarios, such as resource utilization efficiency analysis, system stability evaluation, etc. The system will carefully design the structure and content of each evaluation template according to the expected report type, display requirement and data analysis purpose.

[0109] In order to enable the user to call the corresponding evaluation template according to the specific requirement, the system needs to establish a mapping mechanism of the template request parameter and the evaluation template. This usually requires defining a unique template request parameter for each evaluation template, which contains the identification information of the template and possible customization requirements. For example, an evaluation template focusing on node performance may be assigned a specific template request parameter, indicating that the system calls the template when generating the report.

[0110] When a user or system triggers a state evaluation request, a template request parameter is carried, which is used to indicate which evaluation template is used to generate the report. Before actually processing the request, the system will parse the template request parameter according to the preset mapping relationship, so as to determine the matching evaluation template. This process ensures that the subsequent data acquisition, analysis and report generation steps can be carried out according to the template specified by the user, meeting the customization needs and professional requirements of the user.

[0111] Through the embodiments provided in the present application, the design of the evaluation template considers the diversity of the evaluation elements and the flexibility of the combination, allowing the user to generate reports focusing on different aspects and structures according to specific concerns and data display preferences. The mapping between the template request parameter and the evaluation template provides the user with a mechanism to select a specific template, ensuring the accuracy and customization level of the report generation. When the user initiates a state evaluation request, by parsing the template request parameter and determining the corresponding evaluation template, the system realizes the seamless connection from user demand to report generation logic. This series of steps not only improves the generation efficiency of the evaluation report, but also ensures the diversification and individualization of the report content and format, meeting the user's diverse needs for the depth and breadth of cluster state evaluation.

[0112] As an optional solution, the method further comprises:

[0113] In the case that the state evaluation request carries the period request parameter and the template request parameter only includes the first template request parameter, the first evaluation template matched with the first template request parameter is obtained, and a plurality of first evaluation elements included in the first evaluation template are obtained;

[0114] According to the period number indicated by the period request parameter, a plurality of first evaluation results corresponding to the plurality of first evaluation elements are generated for the evaluation data.

[0115] Optionally, in the present embodiment, the period request parameter is a parameter in the state evaluation request, used to indicate the generation frequency of the evaluation report, such as daily, weekly, monthly, etc.

[0116] Optionally, in the present embodiment, when the state evaluation request carries the period request parameter and the template request parameter only contains the first template request parameter, the system will parse the first template request parameter in the request, so as to determine the matching first evaluation template. Subsequently, the system will identify all the first evaluation elements contained in the first evaluation template, which will guide the subsequent data acquisition and evaluation result generation process. For example, the first evaluation template may contain a resource utilization table, an abnormal event statistics label and a node state column chart, and the system will prepare the corresponding data and analysis logic according to the definition of these elements.

[0117] After determining the first evaluation template, the system will periodically repeat the process of data acquisition and evaluation result generation according to the cycle number set in the cycle request parameter. For example, if the cycle request parameter indicates that a report should be generated once a week, the system will obtain the latest state data from the cluster according to the definition of the first evaluation element at a fixed time point every week, and generate the corresponding first evaluation result. This process will continue until the total number specified in the cycle request parameter is met. In this way, the system can automatically and periodically generate evaluation reports with the same structure and content, helping users to continuously monitor and analyze the running state of the cluster and timely discover potential problems and trend changes.

[0118] Through the embodiments provided in the present application, when the state evaluation request triggered by the user or the system automatically contains the cycle request parameter, and the template request parameter only specifies one evaluation template (i.e., only contains the first template request parameter), the system will automatically identify and lock this specific evaluation template, while understanding all the first evaluation elements contained in the template. Subsequently, the system will automatically execute the data acquisition and evaluation result generation process periodically according to the cycle and number set in the cycle request parameter. This mechanism ensures that the evaluation report can be automatically updated within the time interval defined by the user without manual intervention, greatly saving the time and effort of the operation and maintenance personnel, while ensuring the timeliness and consistency of the report.

[0119] As an optional solution, the method further comprises:

[0120] In the case where the state evaluation request carries the cycle request parameter and the template request parameter includes the first template request parameter and the second template request parameter, the first evaluation template matched with the first template request parameter, the plurality of first evaluation elements included in the first evaluation template are obtained, and the second evaluation template matched with the second template request parameter, the plurality of second evaluation elements included in the second evaluation template are obtained;

[0121] The plurality of first evaluation results corresponding to the plurality of first evaluation elements and the plurality of second evaluation results corresponding to the plurality of second evaluation elements are generated for the evaluation data for a plurality of times according to the cycle number indicated by the cycle request parameter, wherein the evaluation results corresponding to every two times in the cycle number are the plurality of first evaluation results and the plurality of second evaluation results.

[0122] Optionally, in the present embodiment, when the state evaluation request contains both the cycle request parameter and the first template request parameter and the second template request parameter, the system first parses the two template request parameters to determine the first evaluation template and the second evaluation template matched with them. Next, the system further identifies all the evaluation elements (first evaluation elements and second evaluation elements) contained in the first evaluation template and the second evaluation template. These elements define the content structure of the report, including the type, format and analysis rules of data display.

[0123] After the periodic request parameter and the two template request parameters are determined, the system will generate a composite evaluation report containing the first evaluation result and the second evaluation result according to the indication of the periodic number. Specifically, in each periodic evaluation data acquisition process, the system will generate a first evaluation result according to the first evaluation template and a second evaluation result according to the second evaluation template, and integrate them into the same periodic comprehensive evaluation report. This processing method is particularly suitable for scenarios that require simultaneous monitoring and analysis of multiple cluster state data, ensuring that each report generation contains all the key information users care about, improving the comprehensiveness and timeliness of operation and maintenance decisions.

[0124] Through the embodiments provided in the present application, when the state evaluation request contains not only the periodic request parameter, but also two template request parameters (i.e. the first and second template request parameters), the system will perform multi-level data collection and report generation according to these parameters. First, the system will identify and obtain the first evaluation template and the second evaluation template to ensure that the content of the subsequent report can cover the two different dimensions of data specified by the user. Then, within each period, the system will generate corresponding first evaluation results and second evaluation results for the evaluation data according to the evaluation elements in the two templates, and integrate them into a periodic comprehensive evaluation report. This process not only embodies the efficiency and automation characteristics of the evaluation report generation method, but also demonstrates its high customization and data integration capabilities. By alternating or simultaneously presenting the results of different templates and evaluation elements in each period, the system can provide a comprehensive and dynamically updated cluster state view, significantly enhancing the monitoring and analysis capabilities of operation and maintenance personnel on the health status of the cluster, and thus improving the accuracy and efficiency of operation and maintenance decisions.

[0125] As an optional solution, after obtaining the evaluation data of the cluster based on the data request parameter, the method further includes:

[0126] performing data verification on the evaluation data;

[0127] in the case where the data verification determines that the evaluation data includes missing data, obtaining a proportion of the missing data in the evaluation data;

[0128] in the case where the proportion is less than a preset proportion threshold, performing a data filling operation matching the type of the missing data according to the type of the missing data.

[0129] Optionally, in this embodiment, after obtaining the evaluation data, the system will perform comprehensive data verification to check the integrity, accuracy and compliance of the data. The verification process includes but is not limited to field integrity verification, record integrity verification, data accuracy verification, etc., to ensure that each piece of data meets the preset data quality standards.

[0130] If missing data is found during data verification, the system will further calculate the proportion of missing data in the evaluation data. This proportion is crucial because it directly affects whether the evaluation report can continue to be generated and how to handle the missing data.

[0131] When it is determined that the proportion of missing data is below the preset proportion threshold, the system will perform corresponding data filling operations according to the type of missing data. Data filling strategies may include linear interpolation, average value filling, nearest neighbor filling, etc. The specific strategy will be selected according to the nature of the missing data (such as time series data or statistical data) and contextual information to maximize the authenticity and integrity of the data.

[0132] For example, for time series type data, if the missing is a continuous data point in a certain period, the system can use linear interpolation to estimate the missing value based on the data points before and after the missing period; for statistical data, if a statistical data point is missing, the average value or median of the past 7 days can be used for filling.

[0133] Through the embodiments provided in the present application, if data verification finds that there is missing data in the evaluation data, the system will further calculate its proportion in the overall data. If this proportion is below the preset proportion threshold, it means that the impact of missing data on the overall data quality is within a controllable range. At this time, the system will automatically perform data filling operations based on the type of missing data, using the preset filling strategy to supplement the missing data points, thereby maintaining the continuity and accuracy of the evaluation report generation.

[0134] As an optional solution, the above-mentioned cluster state evaluation method is used for artificial intelligence platform cluster operation report generation. With the rapid development of artificial intelligence technology, the scale and complexity of artificial intelligence platform clusters are constantly improving. The detection and analysis of cluster running status become crucial. Cluster operation report as an important carrier reflecting the running status of the cluster can provide key information for operation and management personnel, helping them to discover problems in time and optimize cluster performance.

[0135] Currently, there are many deficiencies in the way of generating artificial intelligence platform cluster operation report. On the one hand, traditional report generation relies on manual production, and operation and maintenance personnel need to extract data from the platform and arrange it into a report according to a certain format, which not only consumes a lot of time and effort, but also is prone to data errors or format inconsistencies due to human operation. On the other hand, some existing automatic report generation tools have fixed report templates, and users cannot define the element types and data sets included in the report according to their own needs, making it difficult to meet the reporting needs in different scenarios. In addition, there is insufficient support for the function of generating reports at regular intervals, which cannot realize the automatic generation of weekly, monthly and annual reports, and lacks intelligent perception and prominent display of data anomalies, bringing inconvenience to operation and maintenance work.

[0136] To solve the above-mentioned defects, the embodiment provides a generation method for artificial intelligence platform cluster operation report based on the above-mentioned cluster state evaluation method. Through this method, the definition of cluster report template, data generation, report export, and configuration generation of timed report can be realized, which can greatly improve the flexibility of cluster operation report definition, and the efficiency and accuracy of report generation.

[0137] Optionally, in the embodiment, an intelligent report template is defined. The template content includes report element types (including labels, tables, column charts, line charts, pie charts), data sets and intelligent analysis rules. Among them, the intelligent analysis rules can be set by the user, for example, special marking when a certain index exceeds the preset threshold. The data set is derived from the data set of the artificial intelligence platform report statistics function, which can be selected by the user as needed.

[0138] Optionally, in the embodiment, the dynamic data set is obtained and analyzed. When the user triggers the report export function or the timing task execution, this module obtains the corresponding report data set from the report statistics function of the artificial intelligence platform according to the data set defined in the report template, the user-selected data range (such as a specific time range) and the intelligent analysis rules. In the data acquisition process, not only the integrity and accuracy of the data are checked, and if there is data missing or abnormal, a prompt or a preset processing method (such as filling in the default value) is used, but also the data is analyzed in depth according to the intelligent analysis rules to identify key information such as abnormal fluctuations and trend changes in the data, and generate analysis markers.

[0139] Optionally, in this embodiment, adaptive report generation and export are implemented. After obtaining the corresponding report dataset and analysis tags, the module fills in the data and analysis tags according to the element types and layouts defined in the report template. For tag elements, in addition to filling in the corresponding data description text, analysis tags are also displayed if applicable. For table elements, the data is arranged and filled in according to the row and column structure, and abnormal data cells are marked with a special color. For chart elements such as bar charts, line charts, and pie charts, corresponding graphs are automatically generated based on the data, and abnormal data points are highlighted with special symbols. After filling is complete, the report layout is automatically adjusted according to the data volume and element layout to ensure that the content is clear and aesthetically pleasing. Then, the report is converted to PDF format, and an export function is provided, allowing users to choose to save the report to a specified local path.

[0140] Optionally, in this embodiment, intelligent scheduled task management is implemented. A definable intelligent scheduled task function is provided, allowing users to set the execution cycle (e.g., weekly, monthly, yearly), execution time, and corresponding report templates and data ranges for scheduled tasks according to their needs. This module has intelligent task adjustment capabilities, automatically optimizing task execution time based on historical report generation data and cluster load, avoiding task execution during periods of high cluster load. Simultaneously, the dynamic dataset acquisition and analysis module and the adaptive report generation and export module are automatically triggered according to the set time to complete the scheduled generation of cluster operation reports. The notification method for task execution results (e.g., email notification, platform message notification) can be configured to inform users that the report has been generated and its storage location. If there are anomaly analysis markers in the report, they will be highlighted in the notification.

[0141] To further illustrate, here is a flowchart illustrating how an artificial intelligence platform cluster operation report is generated, such as... Figure 3 As shown, the specific steps include: intelligent report template definition, dataset acquisition and analysis, report generation and export, and scheduled task management. Optionally, the intelligent report template definition step is the foundation of the entire process, providing the report presentation format, including tags, tables, line charts, bar charts, pie charts, etc., for subsequent data visualization. The dataset acquisition and analysis step first reads the report template to clarify the data display requirements; then, it acquires the data required for the report from the reporting service; next, it analyzes the acquired data to provide data support for report generation. The report generation and export step first populates the analyzed dataset into the report template; then, it automatically formats the report to ensure a standardized and aesthetically pleasing format; next, it generates a PDF report; finally, it outputs the generated report to a specified path, completing the entire report generation process. The scheduled task management step manages the report generation tasks, allowing you to define the task execution cycle, determine the frequency of report generation, define associated report templates, specify the templates used by the task, and define task notification methods to provide notifications when relevant task execution statuses occur.

[0142] Optionally, in the embodiment, the intelligent report template definition function is developed to realize the definition of the report template. The definition content includes the report name, the report description, and the report data. The report data is defined in sequence from top to bottom. The corresponding report data set is obtained in the defined order during report generation, and the report data generation is completed. Each part of the report data includes data type, title, data source, data set, and other attributes. According to different data types, the attribute columns that need to be defined are dynamically displayed. The attribute information that needs to be defined for different data types is shown in Table 1:

[0143] Table 1

[0144]

[0145] For the column attributes described above, including data column, index column, X-axis, Y-axis, classification, and value, intelligent analysis rules can be set, such as “CPU usage rate > 90% is marked as abnormal”. When marked as abnormal, the generated report is displayed in the set eye-catching font and color.

[0146] Optionally, in the embodiment, by calling the report statistical data interface provided by the artificial intelligence platform, according to the data source and data set identifier in the report template, and the data range parameters (such as start time startTime and end time endTime) selected by the user, the data request is sent. After the interface returns the data, the data acquisition module parses the data, checks the field integrity and data format of the data, and whether they meet the template requirements. If there are missing necessary fields, an error prompt is returned to the user; if the data format is not consistent, format conversion processing is performed, such as converting a string type date to a date format. Finally, a data set suitable for the corresponding data type is generated to fill in the components in the subsequent report.

[0147] Specifically, the embodiment obtains the corresponding report data by calling the report statistical function interface of the artificial intelligence platform. The interface data is obtained in sequence by calling the corresponding interface according to the data type defined in the report.

[0148] Field integrity verification: first, during template definition, mark the mandatory fields. The system automatically compares the fields of the pulled data set with the mandatory field list. If there are missing fields, the prompt mechanism is triggered immediately - a warning window is popped up on the interface, displaying: “Missing field: field name, this field is the core indicator of “template name”, which may affect the integrity of the report”, and providing the options of “complete pulling” (re-requesting data containing missing fields) or “ignore and continue” (marking as missing).

[0149] Record integrity check: For time range type data, check the continuity and coverage of time series. Pull data by timestamp sorting, calculate the time interval of adjacent records, and if the interval exceeds the preset threshold (default 15 minutes), judge as "time segment missing".

[0150] For numerical fields, a reasonable value range is preset. For text fields, a legal format is preset, such as IP address matching according to regular expressions. Based on business logic, the constraint relationship of associated fields is preset, such as cluster accelerator card number = accelerator card used number + accelerator card available number. Real-time calculation is performed on the associated fields. If the equation is not established, it is marked as "logical exception".

[0151] For time series type data, linear interpolation method is used for filling, such as missing 10:00 data, which is filled with the average value of 9:00 to 11:00 data. For statistical data, the average or median of the last 7 days is used for filling. If the missing data ratio exceeds the preset threshold (default 10%, configurable), the system will pause processing and give a prompt, and the user can choose "continue processing", "re-pull data" or "manual input".

[0152] This embodiment can but not limited to use PDF generation library itext (an open source Java library for PDF document generation and operation) to build PDF document structure, and draw labels, tables and charts in PDF pages according to the element layout defined in the template. The specific cases are as follows:

[0153] For label elements, the itext library Paragraph object is used to generate label elements in PDF documents, and abnormal elements are marked.

[0154] For table elements, the itext library Table object is used to fill the parsed interface data to generate table elements in PDF documents, and the elements marked as abnormal are highlighted in cells.

[0155] For chart elements, according to the defined chart type, the parsed interface data is passed to the corresponding chart object through the JFreeChart library (an open source Java library for generating various statistical charts) to generate the corresponding chart (bar chart, line chart, pie chart), and the abnormal data points are set with special symbols and colors. Then the chart object is converted into a picture, which is used as an itext picture element for display.

[0156] After completing the construction of the document, the entire PDF document object is output and saved to the specified server directory for users to download and view through the interface.

[0157] The embodiment is based on an intelligent task scheduling framework to implement management of a timing task. A user sets basic information (task name, period, time, etc.), an associated report template ID, and a data range parameter of the timing task on an interface. The system stores the information into a task database. The scheduling framework periodically scans the tasks in the database. When a task execution time is reached, an execution thread is triggered to call an interface of a dynamic data set acquisition and analysis module and an adaptive report generation and export module to complete report generation. After the task execution is completed, notification information including a report storage address and an exception prompt (if any) is sent according to a notification mode configured by the user.

[0158] Regarding the definition of the timing task, as shown in Table 2:

[0159] Table 2

[0160]

[0161] It can be understood that the embodiment supports a highly customized intelligent report template definition mechanism, allows a user to flexibly configure a report element type (a label, a table, a column chart, a line chart, a pie chart, etc.), a data set source, and an intelligent analysis rule (such as an index threshold exception marking), breaks through the limitations of a traditional fixed template, and meets personalized report requirements in different scenarios.

[0162] The embodiment automatically acquires data through an interface and performs integrity (field, time sequence) and accuracy (numerical range, format, business logic) verification. Missing or abnormal data is processed by using a preset rule (such as linear interpolation or mean value filling). In combination with an intelligent analysis rule, data trends and exceptions are identified, analysis marks are generated, and data reliability and analysis depth are improved.

[0163] The embodiment automatically fills data according to a template, differentially displays abnormal information (a special color of a table cell or a special symbol of an abnormal point of a chart), and automatically optimizes layout based on a data volume and layout. Finally, the report is exported in a PDF format, the beauty of the report is realized, and the efficiency of information transmission is high.

[0164] The embodiment supports custom report generation periods (week / month / year), times, and templates. The core is that a task execution time can be automatically adjusted to avoid a high load period according to historical load data of a cluster. Meanwhile, a report and an exception prompt are timely pushed through a notification mechanism (email or a platform message), and stability and operation and maintenance response efficiency of the timing task are improved.

[0165] The embodiment deeply integrates four modules of template definition, data processing, report generation, and timing management to form a full-process automation solution from “user requirements” to “intelligent report output”. The solution solves problems of low efficiency of traditional manual production, poor flexibility of fixed tools, and weak exception perception.

[0166] To further illustrate, the processing flow for generating an AI platform cluster operation report is as follows: Figure 4 As shown, the specific steps include: First, the "user or scheduled task," acting as the trigger for report generation, sends a request to the "report template" to obtain it. This step determines the basic framework of the report, such as its format and structure. After the "report template" returns the result, the "user or scheduled task" calls the "report service" interface to obtain the data needed to generate the report. Upon receiving the relevant request, the "report service" populates the "report data" with report data, preparing the data content for the subsequent generation of the complete report. After receiving the data population result from the "report service," the "report template" interacts with "document generation" to assemble the report template and the populated data to generate the document. Finally, "document generation" returns document information (and document status information) to the "user or scheduled task," informing them of the report generation result (success or failure status, and the generated document).

[0167] The embodiments provided in this application significantly improve report generation efficiency and quality: intelligent template recommendations reduce user template definition time, dynamic data acquisition and analysis eliminate manual analysis, and adaptive layout ensures report aesthetics, resulting in a substantial overall improvement in report generation efficiency and quality, allowing operations and maintenance personnel to focus on problem solving. Report flexibility is significantly enhanced: dynamic data analysis automatically identifies key information, making reports more aligned with actual needs and providing intelligent decision support capabilities. Report accuracy and anomaly detection capabilities are effectively guaranteed: data is directly acquired from the platform, reducing human error, and intelligent analysis rules can promptly detect and highlight data anomalies, facilitating rapid cluster operation problem detection for users. Intelligent optimization of scheduled report generation is achieved: intelligent scheduled task management automatically avoids high cluster load periods, ensuring smooth task execution, while providing timely alerts for abnormal reports, improving operations and maintenance response speed.

[0168] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of this application.

[0169] A cluster state evaluation apparatus is also provided in the embodiments, which is configured to implement the above embodiments and preferred embodiments, and details of which have been described above. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the apparatus described in the following embodiments is preferably implemented in software, implementation in hardware, or a combination of software and hardware, is also possible and contemplated.

[0170] Figure 5 is a structural block diagram of a cluster state evaluation apparatus according to an embodiment of the present application, as shown in Figure 5 the apparatus comprises:

[0171] a first obtaining unit 502 configured to obtain a state evaluation request triggered for a cluster, wherein the state evaluation request carries a template request parameter and a data request parameter;

[0172] a second obtaining unit 504 configured to obtain evaluation data of the cluster based on the data request parameter, wherein the evaluation data is cluster data that meets a data range indicated by the data request parameter;

[0173] a third obtaining unit 506 configured to obtain a plurality of evaluation elements included in an evaluation template in a case where the evaluation template matching the template request parameter is obtained, wherein one evaluation element corresponds to one evaluation style;

[0174] an evaluation unit 508 configured to generate a plurality of evaluation results for the evaluation data according to each evaluation style, wherein one evaluation result corresponds to one evaluation element, and the plurality of evaluation results are used to evaluate a running state of the cluster.

[0175] As an optional solution, the apparatus further comprises:

[0176] a consolidation module configured to consolidate the plurality of evaluation results after the plurality of evaluation results are generated for the evaluation data according to each evaluation style, to obtain an evaluation report used to evaluate the running state of the cluster, wherein a first arrangement order of the plurality of evaluation results on the evaluation report corresponds to a second arrangement order of the plurality of evaluation elements on the evaluation template.

[0177] As an optional solution, the apparatus further comprises:

[0178] a display module configured to display first evaluation content in the evaluation report in a first display style and second evaluation content in the evaluation report in a second display style after the plurality of evaluation results are consolidated to obtain the evaluation report used to evaluate the running state of the cluster, wherein the first evaluation content is evaluation content in the plurality of evaluation results that meets an expected abnormal condition, the second evaluation content is evaluation content in the plurality of evaluation results other than the first evaluation content, and the first display style is different from the second display style.

[0179] As an optional solution, the apparatus further comprises:

[0180] The first obtaining module is configured to obtain a parameter type and a parameter threshold associated with the parameter type carried in the state evaluation request before displaying the first evaluation content in the evaluation report in the first display style and displaying the second evaluation content in the evaluation report in the second display style;

[0181] The first determining module is configured to determine an expected abnormal condition according to the parameter type and the parameter threshold before displaying the first evaluation content in the evaluation report in the first display style and displaying the second evaluation content in the evaluation report in the second display style, wherein the expected abnormal condition is used to indicate that the parameter type is greater than the parameter threshold;

[0182] The second determining module is configured to determine the first evaluation content meeting the expected abnormal condition from the plurality of evaluation results before displaying the first evaluation content in the evaluation report in the first display style and displaying the second evaluation content in the evaluation report in the second display style.

[0183] As an optional solution, the evaluation unit 508 comprises at least one of the following:

[0184] The first evaluation module is configured to generate a label evaluation result corresponding to a label evaluation element for the evaluation data in a case where the plurality of evaluation elements comprises the label evaluation element, and the plurality of evaluation results comprises the label evaluation result;

[0185] The second evaluation module is configured to generate a table evaluation result corresponding to a table evaluation element for the evaluation data in a case where the plurality of evaluation elements comprises the table evaluation element, and the plurality of evaluation results comprises the table evaluation result;

[0186] The third evaluation module is configured to generate a graph evaluation result corresponding to a graph evaluation element for the evaluation data in a case where the plurality of evaluation elements comprises the graph evaluation element, and the plurality of evaluation results comprises the graph evaluation result.

[0187] As an optional solution, the third evaluation module comprises at least one of the following:

[0188] The first evaluation submodule is configured to generate a line graph evaluation result corresponding to a line graph element for the evaluation data in a case where the graph evaluation element is the line graph element, and the graph evaluation result comprises the line graph evaluation result;

[0189] The second evaluation submodule is configured to generate a pie chart evaluation result corresponding to a pie chart element for the evaluation data in a case where the graph evaluation element is the pie chart element, and the graph evaluation result comprises the pie chart evaluation result;

[0190] The third evaluation submodule is configured to generate a column chart evaluation result corresponding to the column chart element for the evaluation data when the graphic evaluation element is the column chart element, and the graphic evaluation result comprises the column chart evaluation result.

[0191] As an optional solution, the second obtaining unit 504 comprises:

[0192] The third determination module is configured to determine, when the data request parameter comprises the time request sub-parameter, first data generated by the cluster within a time range indicated by the time request sub-parameter as the evaluation data; or,

[0193] The fourth determination module is configured to determine, when the data request parameter comprises the node request sub-parameter, second data generated by a target node indicated by the node request sub-parameter in the cluster as the evaluation data.

[0194] As an optional solution, the second obtaining unit 504 further comprises:

[0195] The fifth determination module is configured to determine, when the data request parameter comprises the time request sub-parameter and the node request sub-parameter, third data generated by the target node within the time range in the cluster as the evaluation data.

[0196] As an optional solution, the apparatus further comprises:

[0197] The combination module is configured to, before obtaining the state evaluation request triggered by the cluster, sequentially combine the at least two evaluation elements to generate at least two evaluation templates, wherein the number of evaluation elements contained in different evaluation templates is different, or the combination order of the evaluation elements contained is different.

[0198] The establishment module is configured to, before obtaining the state evaluation request triggered by the cluster, establish a one-to-one mapping relationship between the at least two template request parameters and the at least two evaluation templates.

[0199] The apparatus further comprises:

[0200] The sixth determination module is configured to, before obtaining the plurality of evaluation elements comprised by the evaluation template, determine the evaluation template matched by the template request parameter from the mapping relationship.

[0201] As an optional solution, the apparatus further comprises:

[0202] The second obtaining module is configured to, when the state evaluation request carries the period request parameter and the template request parameter only comprises the first template request parameter, obtain the first evaluation template matched by the first template request parameter, and obtain the plurality of first evaluation elements comprised by the first evaluation template.

[0203] The first generation module is configured to generate, according to the periodic number indicated by the periodic request parameter, a plurality of first evaluation results corresponding to a plurality of first evaluation elements for the evaluation data for a plurality of times.

[0204] As an optional solution, the apparatus further includes:

[0205] The third acquisition module is configured to, in a case where the state evaluation request carries the periodic request parameter and the template request parameter includes the first template request parameter and the second template request parameter, acquire a first evaluation template matched with the first template request parameter, a plurality of first evaluation elements included in the first evaluation template, and acquire a second evaluation template matched with the second template request parameter, a plurality of second evaluation elements included in the second evaluation template.

[0206] The second generation module is configured to, according to the periodic number indicated by the periodic request parameter, generate, for the evaluation data for a plurality of times, a plurality of first evaluation results corresponding to a plurality of first evaluation elements and a plurality of second evaluation results corresponding to a plurality of second evaluation elements, wherein the evaluation results corresponding to each two times in the periodic number are the plurality of first evaluation results and the plurality of second evaluation results.

[0207] As an optional solution, the apparatus further includes:

[0208] The verification module is configured to, after acquiring the evaluation data of the cluster based on the data request parameter, perform data verification on the evaluation data.

[0209] The fourth acquisition module is configured to, after acquiring the evaluation data of the cluster based on the data request parameter, acquire a proportion of missing data in the evaluation data in a case where it is determined through the data verification that the evaluation data includes the missing data.

[0210] The filling module is configured to, after acquiring the evaluation data of the cluster based on the data request parameter, perform, in a case where the proportion is less than a preset proportion threshold, a data filling operation matched with a type of the missing data according to the type of the missing data.

[0211] The specific examples in this embodiment can refer to the examples described in the above embodiments and exemplary embodiments, and will not be described herein again.

[0212] Those skilled in the art can clearly understand that the method according to the above-mentioned embodiments can be realized by means of software on a general hardware platform, and of course, can also be realized by hardware, but in many cases, the former is a better implementation. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, which is stored in a storage medium (such as a ROM / RAM, a magnetic disk, or an optical disk) and includes a plurality of instructions for causing a terminal device (which can be a mobile phone, a computer, a server, or a network device) to execute the methods of the various embodiments of the present application.

[0213] It should be noted that the above-mentioned modules can be realized by software or hardware, and for the latter, the following implementation manners can be used, but are not limited thereto: all the modules are located in the same processor; or the modules are located in different processors in any combination.

[0214] The embodiments of the present application also provide a computer readable storage medium, which stores a computer program, and the computer program is configured to execute the steps in any of the above-mentioned method embodiments when running.

[0215] In an example embodiment, the above-mentioned computer readable storage medium can include, but is not limited to, a U disk, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk, and various media that can store computer programs.

[0216] The embodiments of the present application also provide an electronic device, which includes a memory and a processor, the memory stores a computer program, and the processor is configured to execute the computer program to execute the steps in any of the above-mentioned method embodiments.

[0217] In an example embodiment, the above-mentioned electronic device can also include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0218] The embodiments of the present application also provide a computer program product, which includes a non-volatile computer readable storage medium, the non-volatile computer readable storage medium stores a computer program product, and the computer program is executed by a processor to realize the steps in the method of the various embodiments of the present application.

[0219] The specific examples in the present embodiment can refer to the examples described in the above-mentioned embodiments and example embodiments, which will not be described herein again.

[0220] It should be apparent to those skilled in the art that the modules or steps of the application described above can be implemented with general computing devices, which can be centralized on a single computing device or distributed on a network of multiple computing devices, which can be implemented with program codes executable by the computing devices, so that they can be stored in storage devices and executed by the computing devices, and in some cases, the steps shown or described can be executed in different orders than shown, or made into individual integrated circuit modules, or made into a single integrated circuit module. Thus, the present application is not limited to any particular combination of hardware and software.

[0221] The preferred embodiments of the present application are only used to illustrate the technical solutions of the present application and not used to limit the present application. Any modifications, equivalent replacements, improvements, and the like made within the principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for assessing the state of a cluster, characterized in that, include: At least two evaluation elements are combined in an ordered manner to generate at least two evaluation templates, wherein the number of evaluation elements contained in different evaluation templates is different, or the order in which the evaluation elements are combined is different. Establish a one-to-one mapping relationship between at least two template request parameters and the at least two evaluation templates; Obtain a status assessment request triggered on the cluster, wherein the status assessment request carries template request parameters and data request parameters; Based on the historical report generation status and cluster operating load of the cluster, the task execution time for responding to the status assessment request is determined, wherein the task execution time is during a period of low cluster load. Perform the following operations during the task execution time: When the data request parameter includes a time request sub-parameter, the first data generated by the cluster within the time range indicated by the time request sub-parameter is determined as the evaluation data of the cluster; or, when the data request parameter includes a node request sub-parameter, the second data generated by the target node indicated by the node request sub-parameter in the cluster is determined as the evaluation data; or, when the data request parameter includes both the time request sub-parameter and the node request sub-parameter, the third data generated by the target node in the cluster within the time range is determined as the evaluation data, wherein the evaluation data is cluster data that conforms to the data range indicated by the data request parameter; The evaluation template matching the template request parameters is determined from the mapping relationship, and multiple evaluation elements included in the evaluation template are obtained, wherein one evaluation element corresponds to one evaluation style. According to each evaluation style, multiple evaluation results are generated for the evaluation data, wherein each evaluation result corresponds to an evaluation element, and the multiple evaluation results are used to evaluate the operating status of the cluster. When the status assessment request carries a periodic request parameter and the template request parameter only includes a first template request parameter, a first assessment template matching the first template request parameter is obtained, and multiple first assessment elements included in the first assessment template are obtained, wherein the periodic request parameter is used to indicate the generation frequency of the assessment report; according to the number of periods indicated by the periodic request parameter, multiple first assessment results corresponding to the multiple first assessment elements are generated multiple times for the assessment data; When the status assessment request carries the periodic request parameter and the template request parameter includes the first template request parameter and the second template request parameter, the first assessment template matching the first template request parameter and the plurality of first assessment elements included in the first assessment template are obtained, and the second assessment template matching the second template request parameter and the plurality of second assessment elements included in the second assessment template are obtained; for the number of periods indicated by the periodic request parameter, a plurality of first assessment results corresponding to the plurality of first assessment elements and a plurality of second assessment results corresponding to the plurality of second assessment elements are generated multiple times for the assessment data, and the first assessment results and second assessment results of the same period among the plurality of first assessment results and the plurality of second assessment results are integrated into the same periodic comprehensive assessment report.

2. The method according to claim 1, characterized in that, After generating multiple evaluation results for the evaluation data according to each evaluation style, the method further includes: The multiple evaluation results are integrated to obtain an evaluation report for evaluating the operating status of the cluster, wherein the first arrangement order of the multiple evaluation results on the evaluation report corresponds to the second arrangement order of the multiple evaluation elements on the evaluation template.

3. The method according to claim 2, characterized in that, After integrating the multiple evaluation results to obtain an evaluation report for assessing the operational status of the cluster, the method further includes: The evaluation report displays the first evaluation content according to a first display style and the second evaluation content according to a second display style. The first evaluation content is the evaluation content that meets the expected abnormal conditions among the plurality of evaluation results, and the second evaluation content is the evaluation content other than the first evaluation content among the plurality of evaluation results. The first display style and the second display style are different.

4. The method according to claim 3, characterized in that, Before displaying the first evaluation content in the evaluation report according to the first display style, and the second evaluation content in the evaluation report according to the second display style, the method further includes: Obtain the parameter types carried in the state assessment request and the parameter thresholds associated with those parameter types; The expected abnormal condition is determined based on the parameter type and the parameter threshold, wherein the expected abnormal condition is used to indicate that the parameter type is greater than the parameter threshold; The first evaluation content that meets the expected abnormal conditions is determined from the multiple evaluation results.

5. The method according to claim 1, characterized in that, The process of generating multiple evaluation results for the evaluation data according to each evaluation style includes at least one of the following: When the plurality of evaluation elements include a label evaluation element, a label evaluation result corresponding to the label evaluation element is generated for the evaluation data, wherein the plurality of evaluation results include the label evaluation result; When the plurality of evaluation elements include table evaluation elements, a table evaluation result corresponding to the table evaluation element is generated for the evaluation data, and the plurality of evaluation results include the table evaluation result; When the plurality of evaluation elements include graphical evaluation elements, a graphical evaluation result corresponding to the graphical evaluation element is generated for the evaluation data, and the plurality of evaluation results include the graphical evaluation result.

6. The method according to claim 5, characterized in that, When the plurality of evaluation elements include graphical evaluation elements, generating a graphical evaluation result corresponding to the graphical evaluation elements for the evaluation data includes at least one of the following: When the graphical evaluation element is a line chart element, a line chart evaluation result corresponding to the line chart element is generated for the evaluation data, and the graphical evaluation result includes the line chart evaluation result; When the graphic evaluation element is a pie chart element, a pie chart evaluation result corresponding to the pie chart element is generated for the evaluation data, and the graphic evaluation result includes the pie chart evaluation result; When the graphical evaluation element is a bar chart element, a bar chart evaluation result corresponding to the bar chart element is generated for the evaluation data, and the graphical evaluation result includes the bar chart evaluation result.

7. The method according to any one of claims 1 to 6, characterized in that, After obtaining the cluster's evaluation data based on the data request parameters, the method further includes: The evaluation data is validated. If the data verification determines that the evaluation data includes missing data, the proportion of the missing data in the evaluation data is obtained. If the percentage is less than a preset percentage threshold, a data filling operation matching the type of missing data is performed.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein the computer program, when executed by a processor, implements the steps of the method according to any one of claims 1 to 7.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images