Financial data processing method and system based on cloud platform

Abstract

The invention relates to the technical field of financial data processing, in particular to a financial data processing method and system based on a cloud platform. The method comprises the following steps: collecting multi-source original financial data in a distributed manner; performing data standardization on the multi-source original financial data to obtain standard multi-source financial data; matching the standard multi-source financial data with a financial scene and generating a financial scene identifier; determining a financial scene priority order according to the financial scene identifier to obtain a scene priority value; and monitoring a communication network state of the financial cloud platform based on the scene priority degree value to obtain scene network state data. Through the data processing technology, the network state detection technology and the network fault optimization technology, the fault risk of the server is detected, and the fault risk optimization is carried out on the server side, so that the stability of the cloud platform is improved; the financial task resource allocation data of the cloud platform is dynamically adjusted, and the reliability of financial data is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of financial data processing, and particularly to a financial data processing method and system based on a cloud platform. Background Art

[0002] The cloud platform uses distributed computing technology to disperse the processing of financial data, improving the data processing speed and efficiency; according to the changes in the enterprise's business needs, the cloud platform can quickly expand computing resources to ensure that the data processing capacity always meets the requirements. However, if hardware devices such as servers in the cloud platform fail, such as hard disk damage, memory failure, etc., it will lead to the loss of financial data or the system crash. In addition, the aging or improper maintenance of communication devices causes network failures, which in turn affects the stability of the cloud platform. The performance and stability of communication devices directly affect the availability of the cloud platform. If the network bandwidth is insufficient, the network latency is too high, or the network is interrupted, users will be unable to access the financial data on the cloud platform in a timely manner, affecting the normal conduct of financial work. For example, during a critical financial settlement period, network connection problems cause financial data to be unable to be entered, affecting the normal operation of the enterprise. The transmission speed and processing capacity of communication devices are limited, resulting in delays when financial data is transmitted between the cloud platform and user terminals. For financial data that requires real-time processing and analysis. The hardware resource allocation mechanism of the cloud platform is not perfect enough, resulting in some users being unable to obtain sufficient computing resources and network bandwidth during high-concurrency access, thus affecting the speed and efficiency of financial data processing. Summary of the Invention

[0003] Based on this, it is necessary to provide a financial data processing method and system based on a cloud platform to solve at least one of the above technical problems.

[0004] To achieve the above object, a financial data processing method based on a cloud platform, the method includes the following steps:

[0005] Step S1: Distributively collect multi-source original financial data; perform data standardization on the multi-source original financial data to obtain standard multi-source financial data; match the standard multi-source financial data with financial scenarios and generate financial scenario identifiers;

[0006] Step S2: Determine the priority ranking of financial scenarios according to the financial scenario identifiers to obtain the scenario priority degree value; monitor the communication network status of the financial cloud platform based on the scenario priority degree value to obtain scenario network status data; detect the cloud platform resource demand information with the scenario network status data, and determine the financial task resource allocation data according to the cloud platform resource demand information;

[0007] Step S3: Identify faults in the server side of the financial cloud platform according to the financial task resource allocation data, and generate server fault data; optimize the fault risk of the server side of the financial cloud platform based on the server side fault data to obtain the optimized financial cloud platform;

[0008] Step S4: Upload the standard multi-source financial data to the optimized financial cloud platform and build a financial database; present a financial data report according to the financial database.

[0009] Through the distributed acquisition technology, the present invention can obtain financial data from multiple different data sources, ensuring the extensiveness and diversity of data sources and providing a comprehensive data foundation for subsequent processing. The financial data from different sources and with various formats is standardized, making the data formats unified and standardized, facilitating subsequent matching and processing, improving the usability and consistency of the data, and providing a guarantee for the accurate analysis and application of financial data. By matching the standardized data with specific financial scenarios and generating corresponding financial scenario identifiers, the precise association between data and financial business scenarios is achieved, providing a clear basis for subsequent scenario priority ranking and resource allocation, and enhancing the pertinence and efficiency of financial data processing. According to the financial scenario identifiers, different financial scenarios are ranked in terms of priority to obtain scenario priority degree values, which can reasonably arrange the processing order according to the importance and urgency of financial operations, ensure that key financial tasks are processed first, and improve the efficiency of financial data processing and the rationality of resource utilization. By monitoring the communication network status of the financial cloud platform according to the scenario priority degree values and obtaining scenario network status data, the impact of the network environment on the processing of different financial scenarios can be understood in real time, providing accurate network status information for subsequent resource allocation, and ensuring the stability and reliability of the financial data processing process. By analyzing the scenario network status data, the resource requirement information of the cloud platform in processing different financial scenarios is detected, and based on this, the financial task resource allocation data is determined, realizing the dynamic allocation and optimal configuration of resources, improving the utilization rate of cloud platform resources, and ensuring that the financial data processing tasks can be executed efficiently and stably. According to the financial task resource allocation data, server-side faults are identified to generate server fault data, which can timely detect the fault problems occurring in the server during the processing of financial tasks, provide accurate data support for subsequent fault handling and optimization, and reduce the impact of server faults on financial data processing. According to the server fault data, server-side fault risk optimization is carried out to obtain an optimized financial cloud platform, effectively reducing the probability of server faults occurring, improving the stability and reliability of the financial cloud platform, and ensuring the continuity and security of financial data processing. The multi-source financial data after standardized processing is uploaded to the optimized financial cloud platform, and a financial database is constructed, realizing the centralized management and storage of data, facilitating subsequent data query, analysis, and report generation, and improving the efficiency and convenience of financial data management. Based on the constructed financial database, financial data reports are generated, which can intuitively and accurately display the analysis results and business status of financial data. Therefore, through data processing technology, network status detection technology, and network fault optimization technology, the present invention detects server fault risks and performs server-side fault risk optimization, improving the stability of the cloud platform; dynamically adjusts the financial task resource allocation data of the cloud platform to ensure the reliability of financial data.

[0010] Preferably, step S1 includes the following steps:

[0011] Step S11: Collect financial data from different sources at preset multiple levels of data collection nodes according to the hierarchical time intervals. Among them, the first-level collection nodes collect financial bill data at 1-minute intervals; the second-level collection nodes collect financial object data at 5-minute intervals; the third-level collection nodes collect business type data at 15-minute intervals.

[0012] Step S12: Merge the financial bill data, financial object data, and business type data to generate multi-source original financial data.

[0013] Step S13: Uniformly process the date fields in the multi-source original financial data to obtain standardized date data. Uniformly encode the text fields in the multi-source original financial data to obtain standardized text data. Uniformly process the timestamp fields in the multi-source original financial data to obtain standardized timestamp data. Perform encoding mapping processing on the classification fields in the multi-source original financial data to map the classification labels from different sources to unified digital codes to obtain standardized classification data.

[0014] Step S14: Integrate the standardized date data, standardized text data, standardized timestamp data, and standardized classification data to obtain standard multi-source financial data.

[0015] Step S15: Match the standard multi-source financial data with financial scenarios and generate financial scenario identifiers.

[0016] The present invention collects financial data from different sources at preset multiple levels of data collection nodes at hierarchical time intervals. The first-level collection node collects financial bill data at an interval of 1 minute, the second-level collection node collects financial object data at an interval of 5 minutes, and the third-level collection node collects business type data at an interval of 15 minutes. This hierarchical collection method can accurately match the update frequencies of different financial data, ensure the timely collection of high-frequency data and the reasonable acquisition of low-frequency data, thereby improving the timeliness and integrity of data collection; combines the financial bill data, financial object data, and business type data to achieve data integration from different financial business links; uniformly processes the date fields in the multi-source original financial data, uniformly encodes the text fields, uniformly processes the timestamp fields, and performs encoding mapping processing on the classification fields to map different source classification labels to unified digital codes, ensuring the consistency of data formats and eliminating format differences caused by different data sources; integrates the standardized date data, standardized text data, standardized timestamp data, and standardized classification data. Through this integration process, a structured and standardized financial data set is formed, which is convenient for subsequent financial scenario matching and analysis, and improves the generality and operability of the data; matches the standard multi-source financial data with financial scenarios and generates financial scenario identifiers, realizing the accurate association of data with specific financial business scenarios.

[0017] Preferably, step S15 includes the following steps:

[0018] Step S151: Perform interval division processing on the numerical fields in the standard multi-source data to obtain numerical interval identifiers;

[0019] Step S152: Perform time period division processing on the time fields in the standard multi-source data to obtain time period identifiers;

[0020] Step S153: Perform keyword extraction processing on the text fields in the standard multi-source data, count the occurrence frequencies of the keywords, and classify the text fields according to the occurrence frequencies to obtain category identifiers;

[0021] Step S154: Perform mapping processing on the classification fields in the standard multi-source data to obtain classification field data;

[0022] Step S155: Perform hash processing on the classification field data based on the numerical interval identifiers, time period identifiers, and category identifiers to generate scenario identifiers.

[0023] The generation of the numerical interval identifier of the present invention provides a clear classification basis for the numerical dimension of financial data, facilitating the rapid screening and positioning of data within a specific range and improving the data retrieval efficiency; the division of the time period identifier provides a unified period framework for the time dimension of financial data, facilitating the analysis of the change trends of data in different time periods and enhancing the accuracy of time series analysis; the extraction of keywords and the division of category identifiers quantify and structure the key information in the text data, facilitating classification and analysis according to the text content and enhancing the characteristics of the text data in financial analysis; the mapping process of the classification field data standardizes the classification fields in multi-source data, ensuring data consistency and providing a reliable basis for subsequent data integration and analysis; the generation of the scenario identifier based on the hash processing of multi-dimensional identifiers can efficiently integrate the multi-dimensional characteristics of financial data, realize the rapid positioning and scenario differentiation of data, and meet the data processing requirements in complex financial scenarios.

[0024] Preferably, the determination of the financial scenario priority ranking according to the financial scenario identifier in step S2 includes:

[0025] Perform decoding processing on the scenario identifier to extract multiple feature fields; the feature fields include a time period field, a numerical interval field, a category field, and a scenario type field; the decoding process includes decomposing the scenario identifier into multiple parts, and each part corresponds to a feature field;

[0026] Assign weights to the feature fields to obtain feature field weight values; the weight of the time period field is 0.3, the weight of the numerical interval field is 0.4, the weight of the category field is 0.2, and the weight of the scenario type field is 0.1;

[0027] Perform weighted calculation on the feature field weight values to obtain a priority score; the weighted calculation formula is: priority score = time period field value × time period weight + numerical interval field value × numerical interval weight + category field value × category weight + scenario type field value × scenario type weight;

[0028] Sort the data items according to the priority score and arrange them in descending order to obtain the scenario priority value.

[0029] By decomposing the scene identifier into a time period field, a numerical range field, a category field, and a scene type field, the present invention can accurately extract multi-dimensional feature information of financial data, providing a clear feature basis for subsequent refined analysis and ensuring the integrity and traceability of data features. The weights of the time period field, the numerical range field, the category field, and the scene type field are clearly set to 0.3, 0.4, 0.2, and 0.1 respectively, realizing a quantitative distinction of the importance of different feature fields, ensuring that the contribution degrees of each feature field in subsequent calculations meet the requirements of actual application scenarios, and providing an objective weight basis for priority calculation. Weighted calculation is performed according to the set weight values to obtain accurate priority scores, which can comprehensively consider the multi-dimensional features of financial data, ensure that the calculation results of priorities are comprehensive and accurate, provide a scientific quantitative basis for data sorting, and avoid the limitations brought by single-dimensional analysis. The data items are sorted from high to low according to the priority scores to obtain the scene priority degree values, realizing the dynamic priority sorting of financial data, facilitating the quick identification and processing of important financial scenarios, improving the efficiency and pertinence of data processing, and ensuring that key financial data can be processed first;

[0030] Preferably, the monitoring of the communication network status of the financial cloud platform in step S2 includes:

[0031] Determine the communication network status monitoring range corresponding to each scene according to the scene priority degree value, where the higher the priority degree value, the larger the monitoring range;

[0032] On multiple nodes of the cloud platform, periodically send test signals according to the communication network status monitoring range and record the timestamps of sending and receiving; the test signal is set as a data packet of a fixed size, containing a unique timestamp, and detecting the signal round-trip time, the number of signals, and the actual transmitted data volume of the network;

[0033] Collect network status data, including network latency, packet loss rate, and bandwidth utilization rate; among them, the network latency is evaluated by measuring the round-trip time of the test signal; the packet loss rate is evaluated by counting the number of test signals that do not receive responses; the bandwidth utilization rate is evaluated by monitoring the actual transmitted data volume;

[0034] Upload the received network status data to the central monitoring system of the cloud platform and summarize and store it to form scene network status data.

[0035] The present invention dynamically adjusts the monitoring range of the communication network status according to the scene priority value, ensuring that scenes with higher priorities obtain a wider monitoring coverage, thereby accurately allocating monitoring resources, optimizing the efficiency and pertinence of network status monitoring, and providing a more reliable network guarantee for the transmission of important financial data. By regularly sending test signals of a fixed size containing a unique timestamp at multiple nodes and recording the sending and receiving timestamps, the round-trip time of the signals can be accurately measured, the number of signals and the actual data volume transmitted can be counted, providing basic data for the quantitative evaluation of the network status and ensuring the accuracy and real-time nature of network performance monitoring. By measuring the round-trip time of the test signals to evaluate network latency, counting the number of test signals without responses to evaluate the packet loss rate, and monitoring the actual data volume transmitted to evaluate the bandwidth utilization rate, a comprehensive quantitative monitoring of the network status is achieved, providing detailed and accurate data support for subsequent network optimization and fault troubleshooting, and ensuring that the network performance meets the requirements of financial data transmission. The collected network status data is uploaded to the central monitoring system of the cloud platform and summarized and stored to form complete scene network status data, which is convenient for centralized management and analysis, providing a unified data basis for network resource allocation, fault warning, and optimization strategy formulation of the cloud platform, and improving the network management efficiency and reliability of the cloud platform.

[0036] Preferably, the detecting the cloud platform resource requirement information from the scene network status data and determining the financial task resource allocation data in step S2 includes:

[0037] Extracting network latency, packet loss rate, and bandwidth utilization rate from the scene network status data to obtain network parameter indicators;

[0038] Detecting the storage resource requirement of the cloud platform according to the network parameter indicators; determining the CPU utilization rate, memory utilization rate, and disk I / O utilization rate of the financial task through the storage resource requirement to generate storage resource allocation data;

[0039] Detecting the computing resource requirement of the cloud platform according to the network parameter indicators; determining the number of tasks and resource occupancy of the financial task during operation through the computing resource requirement to generate computing resource allocation data;

[0040] Detecting the network resource requirement of the cloud platform according to the network parameter indicators; determining the network bandwidth of the financial task through the network resource requirement to generate network resource allocation data;

[0041] Combining the storage resource allocation data, computing resource allocation data, and network resource allocation data to obtain the financial task resource allocation data.

[0042] By extracting key network parameter indicators such as network latency, packet loss rate, and bandwidth utilization, the present invention provides an accurate quantitative basis for subsequent resource demand detection, ensuring that resource allocation decisions are based on actual network performance; determining the CPU utilization rate, memory utilization rate, and disk I / O utilization rate of financial tasks based on network parameter indicators can accurately evaluate the storage resource requirements, generate targeted storage resource allocation data, optimize the storage resource allocation, and improve the utilization efficiency of storage resources; determining the number of running tasks and resource occupancy of financial tasks through network parameter indicators, generating computing resource allocation data, realizing the on-demand allocation of computing resources, and ensuring the efficient operation of financial tasks; determining the network bandwidth requirements of financial tasks based on network parameter indicators, generating network resource allocation data, realizing the dynamic adjustment of network resources, and ensuring the stability and efficiency of financial data transmission; by integrating storage, computing, and network resource allocation data, forming comprehensive financial task resource allocation data, realizing the unified management and optimal configuration of resources, and improving the overall support ability of the cloud platform for financial data processing.

[0043] Preferably, the fault identification of the server side of the financial cloud platform according to the financial task resource allocation data in step S3 includes:

[0044] Detect the storage hardware structure in the server side of the financial cloud platform according to the storage resource allocation data: identify the disk type of the storage hardware structure and record the disk read and write speed; perform read and write latency detection on the disk read and write speed to obtain disk read and write fault data;

[0045] Detect the computing hardware structure in the server side of the financial cloud platform according to the computing resource allocation data: identify the number of CPU cores of the computing hardware structure and record the CPU cache usage; perform cache occupancy fault detection on the CPU cache usage to obtain CPU cache fault data;

[0046] Detect the network hardware structure in the server side of the financial cloud platform according to the network resource allocation data: identify the network interface type of the network hardware structure and record the network interface exchange volume; perform exchange volume fault detection on the network interface exchange volume to obtain exchange volume fault data;

[0047] Integrate the disk read and write fault data, CPU cache fault data, and exchange volume fault data to generate server side fault data.

[0048] By identifying the disk types of the storage hardware structure and recording the disk read and write speeds, the present invention can accurately grasp the actual performance of the storage hardware. By detecting the read and write latency of the disk read and write speeds to obtain disk read and write failure data, potential failures of the storage hardware can be discovered in a timely manner, ensuring the reliability of financial data storage; by identifying the number of CPU cores of the computing hardware structure and recording the CPU cache usage, the usage status of computing resources can be clarified. By detecting the cache occupancy failure of the CPU cache usage to obtain CPU cache failure data, computing resource failures can be effectively prevented, ensuring the efficient operation of financial tasks; by identifying the network interface types of the network hardware structure and recording the network interface traffic, the operation of the network hardware can be comprehensively understood. By detecting the traffic failure of the network interface traffic to obtain traffic failure data, network transmission problems can be discovered in a timely manner, ensuring the stability of financial data transmission; by integrating multi-dimensional failure data to form comprehensive server-side failure data, it is convenient for centralized management and analysis, providing unified data support for fault troubleshooting and resource optimization of the cloud platform, and improving the operation and maintenance efficiency and reliability of the cloud platform.

[0049] Preferably, the fault risk optimization of the server side of the financial cloud platform based on the server-side failure data in step S3 includes:

[0050] Locate the read and write failure disk blocks on the server side of the financial cloud platform according to the disk read and write failure data, and migrate the financial data of the read and write failure disk blocks to generate disk failure optimization measures;

[0051] Determine the cache kernel control group according to the CPU cache failure data, limit the cache usage of non-critical tasks based on the cache kernel control group, and set critical tasks for real-time scheduling to generate cache failure optimization measures;

[0052] Detect the switching path on the server side of the financial cloud platform according to the traffic failure data, and dynamically adjust the switching path volume to generate traffic failure optimization measures;

[0053] Perform fault optimization adjustment on the server side of the financial cloud platform based on the disk failure optimization measures, cache failure optimization measures, and traffic failure optimization measures to obtain the optimized financial cloud platform.

[0054] By accurately locating and migrating relevant financial data from faulty disk blocks, the present invention can effectively avoid data loss and read / write errors, ensuring the integrity and reliability of financial data. By restricting the cache usage of non-critical tasks and optimizing the scheduling of critical tasks, it can improve the utilization efficiency of CPU resources and ensure the efficient operation of financial tasks. By detecting the switching path and dynamically adjusting the switching volume, it can optimize network resource allocation, reduce network latency and packet loss rate, and ensure the stability of financial data transmission. By comprehensively adjusting the server side with various optimization measures, it can comprehensively improve the performance and stability of the financial cloud platform and ensure the reliability of financial data processing.

[0055] Preferably, step S4 includes the following steps:

[0056] Step S41: Split the standard multi-source financial data into multiple financial data blocks, each with a size not exceeding 1 MB; upload the financial data blocks to the financial cloud platform one by one through the network interface to obtain financial data upload information;

[0057] Step S42: Define the database table structure based on the financial data upload information, including table name, field name, and field type, and create a database table, initialize the table structure and index; import the financial data blocks into the database table one by one to build a financial database;

[0058] Step S43: Execute an SQL query on the financial database and extract data that meets the query conditions;

[0059] Step S44: Set the title, header, and footer structures of the report for the data that meets the query conditions to obtain report data, and export the report data as a financial data report.

[0060] The present invention splits the standard multi-source financial data into data blocks not exceeding 1 MB and uploads them to the financial cloud platform one by one, which can ensure the efficiency and stability of data transmission, avoid transmission failures or network congestion caused by overly large data blocks, and facilitate the monitoring and management of the upload process. Defining and creating the database table structure based on the financial data upload information, and initializing the table structure and index can provide a standardized and efficient storage environment for financial data, ensure fast data retrieval and query performance, and lay a foundation for subsequent data processing and analysis. Executing an SQL query on the financial database and extracting data that meets the query conditions can efficiently filter out target financial data, meet diverse data analysis needs, and improve the flexibility and accuracy of data processing. Setting the title, header, and footer structures of the report for the data that meets the query conditions and exporting it as a financial data report can present the data in a standardized format, facilitate users to quickly understand and use it, improve the readability and practicality of financial data, and meet the output requirements of financial reports.

[0061] In this specification, a financial data processing system based on a cloud platform is provided for implementing the above-mentioned financial data processing method based on a cloud platform. The financial data processing system based on a cloud platform includes:

[0062] A financial data collection module for distributively collecting multi-source original financial data; performing data standardization on the multi-source original financial data to obtain standard multi-source financial data; matching the standard multi-source financial data with financial scenarios and generating financial scenario identifiers;

[0063] A communication network status detection module for determining the priority ranking of financial scenarios according to the financial scenario identifiers to obtain the scenario priority degree value; monitoring the communication network status of the financial cloud platform based on the scenario priority degree value to obtain scenario network status data; detecting the cloud platform resource demand information from the scenario network status data, and determining the financial task resource allocation data according to the cloud platform resource demand information;

[0064] A server failure optimization module for identifying server failures on the server side of the financial cloud platform according to the financial task resource allocation data to generate server failure data; performing failure risk optimization on the server side of the financial cloud platform based on the server side failure data to obtain an optimized financial cloud platform;

[0065] A financial database management module for uploading the standard multi-source financial data to the optimized financial cloud platform and constructing a financial database; presenting a financial data report according to the financial database.

[0066] Through the financial data collection module, the present invention realizes the distributive collection and standardized processing of multi-source original financial data, and generates financial scenario identifiers, providing a high-quality basis and clear classification basis for data processing; the communication network status detection module determines the scenario priority degree value according to the financial scenario identifiers, monitors the communication network status and generates scenario network status data, and then accurately evaluates the cloud platform resource requirements and determines the financial task resource allocation data to ensure the efficiency and stability of data processing; the server failure optimization module identifies server failures according to the resource allocation data and performs optimization to improve the system performance and reliability and reduce the failure risk; the financial database management module uploads the standardized data to the optimized cloud platform and constructs a financial database, and finally generates a financial data report to realize the centralized management and intuitive presentation of data and meet the financial reporting requirements. Overall, the system realizes the full-process optimization of financial data processing, improving the data processing efficiency, stability and reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 It is a schematic diagram of the step flow of a financial data processing method based on a cloud platform;

[0068] Figure 2 ForFigure 1 Schematic diagram of the detailed implementation steps of step S1 in

[0069] Figure 3 is Figure 1 Schematic diagram of the detailed implementation steps of step S4 in

[0070] The realization, functional features and advantages of the object of the present invention will be further described in conjunction with the embodiments with reference to the accompanying drawings. Specific implementation manners

[0071] The technical method of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0072] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and thus the repeated description thereof will be omitted. Some of the block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities may be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0073] It should be understood that although the terms "first", "second", etc. may be used here to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0074] To achieve the above object, please refer to Figures 1 to 3 , a financial data processing method based on a cloud platform, the method comprising the following steps:

[0075] Step S1: Distributedly collect multi-source original financial data; perform data standardization on the multi-source original financial data to obtain standard multi-source financial data; match the standard multi-source financial data with a financial scenario and generate a financial scenario identifier;

[0076] Step S2: Determine the priority ranking of financial scenarios according to the financial scenario identifier to obtain the scenario priority value; monitor the communication network status of the financial cloud platform based on the scenario priority value to obtain scenario network status data; detect the cloud platform resource requirement information from the scenario network status data, and determine the financial task resource allocation data according to the cloud platform resource requirement information;

[0077] Step S3: Identify faults in the server side of the financial cloud platform according to the financial task resource allocation data to generate server fault data; optimize the fault risk of the server side of the financial cloud platform based on the server side fault data to obtain the optimized financial cloud platform;

[0078] Step S4: Upload the standard multi-source financial data to the optimized financial cloud platform and build a financial database; present a financial data report according to the financial database.

[0079] In the embodiment of the present invention, with reference to Figure 1 shown in the figure, it is a schematic diagram of the step flow of a financial data processing method based on a cloud platform according to the present invention. In this example, the financial data processing method based on a cloud platform includes the following steps:

[0080] Step S1: Distributedly collect multi-source original financial data; perform data standardization on the multi-source original financial data to obtain standard multi-source financial data; match the standard multi-source financial data with financial scenarios and generate financial scenario identifiers;

[0081] In the embodiments of the present invention, a distributed data acquisition system architecture is adopted, including multiple acquisition nodes (Agents) and a central management node (Master). The acquisition nodes are responsible for acquiring data from their respective data sources, and the central management node is responsible for task scheduling, status monitoring, and data aggregation. According to the sources of financial data, multiple acquisition nodes are configured to connect to different data sources. These data sources include the financial system, ERP system, bank reconciliation system within the enterprise, and external financial data interfaces, etc. Each acquisition node selects an appropriate connection protocol and authentication method according to the type of data source. The central management node distributes acquisition tasks to each acquisition node according to the distribution of data sources and the requirements of acquisition tasks. The acquisition nodes acquire data from the specified data sources according to the assigned tasks. The acquired original financial data is transmitted through the network to a central processing unit or a distributed storage system (such as HDFS) for centralized storage. During the transmission process, data compression and encryption technologies are adopted to improve transmission efficiency and data security. Format conversion is performed on the acquired multi-source financial data to ensure that all data adopts a unified format. For example, the date format is unified to "YYYY-MM-DD", and the amount format is unified to two decimal places. Data cleaning tools are used to preprocess the original data, including removing duplicate data, filling in missing values, correcting outliers, etc. For example, for missing amount data, it can be filled in by interpolation or using default values. A unified coding rule is established for various items and accounts in the financial data. For example, different expense items are coded as "EXPENSE_001", "EXPENSE_002", etc., to facilitate subsequent data processing and analysis. A financial scenario rule library is preset to define the matching conditions for different financial scenarios. For example, the matching conditions for the sales business scenario include that the "income type" is "sales income" and the "transaction status" is "completed". According to the preset rules, the standardized financial data is matched. For the data that matches the sales business scenario, a financial scenario identifier "SALES_SCENE_001" is generated; for the data that matches the purchase business scenario, an identifier "PURCHASE_SCENE_002" is generated. The matched financial data and its corresponding scenario identifiers are stored in a database, and an index is established to facilitate subsequent query and analysis.

[0082] Step S2: Determine the priority ranking of financial scenarios according to the financial scenario identifiers to obtain the scenario priority degree value; monitor the communication network status of the financial cloud platform based on the scenario priority degree value to obtain the scenario network status data; detect the cloud platform resource requirement information from the scenario network status data, and determine the financial task resource allocation data according to the cloud platform resource requirement information;

[0083] In the embodiments of the present invention, a weighted scoring model is used to prioritize financial scenarios. First, multiple evaluation dimensions are defined, including business value, urgency, technical complexity, resource requirements, etc. Each dimension is assigned a corresponding weight. For example, the weight of business value is 0.4, the weight of urgency is 0.3, the weight of technical complexity is 0.2, and the weight of resource requirements is 0.1. For each financial scenario, relevant data is extracted from the database according to its identifier (such as "SALES_SCENE_001"), and financial experts score each dimension according to preset criteria. For example, the sales scenario gets a high score (90 points) in the business value dimension, a medium score (60 points) in the urgency dimension, a low score (30 points) in the technical complexity dimension, and a medium score (50 points) in the resource requirements dimension. Through weighted calculation, the priority value of this scenario is obtained: Priority value = (90×0.4)+(60×0.3)+(30×0.2)+(50×0.1) = 62; In this way, a priority value is calculated for each financial scenario, and all scenarios are sorted from high to low according to the priority value. A network performance monitoring tool (such as Zabbix or Prometheus) is used to monitor the communication network status of the financial cloud platform in real time. According to the priority value of the financial scenario, a higher monitoring frequency is assigned to high-priority scenarios. For example, scenarios with a priority value higher than 80 are monitored for network status every minute, scenarios with a priority value between 60 and 80 are monitored every 5 minutes, and scenarios with a priority value lower than 60 are monitored every 15 minutes. The monitoring metrics include network bandwidth utilization, latency, packet loss rate, etc. The monitored network status data is stored in the monitoring database of the cloud platform, and the data format includes fields such as scenario identifier, monitoring timestamp, bandwidth utilization, latency, and packet loss rate; The network status data of the scenario is analyzed to determine the resource requirement information of the cloud platform. When the network bandwidth utilization exceeds 80% or the latency exceeds 200 milliseconds, a resource requirement alert is triggered. According to the alert level and the priority value of the financial scenario, the resource allocation requirement is calculated. For example, for high-priority scenarios (priority value higher than 80), when the network bandwidth is insufficient, additional bandwidth resources are automatically allocated from the resource pool; for medium-priority scenarios (priority value between 60 and 80), moderate allocation is performed when resources permit; for low-priority scenarios (priority value lower than 60), allocation is only performed when resources are sufficient. The resource allocation data includes information such as scenario identifier, resource type (such as CPU, memory, bandwidth), allocation quantity, and allocation time, and this data is stored in the resource management database of the cloud platform.

[0084] Step S3: Identify faults in the server side of the financial cloud platform based on the financial task resource allocation data, and generate server fault data; optimize the fault risk of the server side of the financial cloud platform based on the server side fault data to obtain the optimized financial cloud platform.

[0085] In the embodiment of the present invention, according to the financial task resource allocation data, a real-time monitoring tool (such as Prometheus) is used to identify faults in the server side of the financial cloud platform. The monitoring metrics include CPU usage rate, memory occupancy rate, disk I / O read and write speed, network bandwidth utilization rate, latency, and the response time and error rate of the application program, etc. When the monitoring metrics exceed the preset thresholds, such as the CPU usage rate exceeds 85%, the memory occupancy rate exceeds 90%, or the network latency exceeds 200 milliseconds, the system will trigger a fault alarm and record the relevant data to generate server fault data; use the automated root cause analysis function of the AIOps platform to deeply mine and correlate analyze the fault data. For example, by analyzing the error codes in the system log, the bottleneck points in the performance metrics (such as disk I / O blockage), and the abnormal behaviors in the application program log, determine the root cause of the fault; according to the fault root cause analysis results, combined with the preset risk assessment model, evaluate the impact of the fault on the financial cloud platform, and formulate an optimization strategy. For example, if the fault is caused by disk I / O blockage, the optimization strategy includes increasing the disk capacity, optimizing the database index, or adjusting the disk scheduling strategy; execute the optimization strategy through the automated tools (such as Ansible, Terraform) of the AIOps platform. For example, automatically adjust the server resource allocation, restart the relevant services, or switch to the standby instance. At the same time, record the detailed information of the optimization operation in the optimization log, including the operation time, the executed optimization measures and their results; after the optimization operation is completed, continuously monitor the running state of the server to evaluate the optimization effect. If the performance metrics return to normal and run stably after optimization, mark the optimization strategy as successful; if the expected effect is not achieved, readjust the optimization strategy according to the new fault data.

[0086] Step S4: Upload the standard multi-source financial data to the optimized financial cloud platform, and build a financial database; present financial data reports according to the financial database.

[0087] In the embodiments of the present invention, first, the standard multi-source financial data is transmitted from the local storage or the data lake to the optimized financial cloud platform by using a distributed data transmission system. The transmission system adopts a highly available architecture. By configuring multiple data transmission nodes, the stability and reliability of data transmission are ensured. During the transmission process, the data is divided into multiple small pieces, and each small piece is verified after the transmission is completed to ensure the integrity of the data. After the transmission is completed, the data is temporarily stored in the distributed storage system of the cloud platform for subsequent processing and storage. Next, a relational database management system is deployed on the cloud platform to construct a financial database. Through database management tools, a financial database instance is created, and a reasonable table structure is designed according to the characteristics of the financial data. The table structure design follows the normalization principle to reduce data redundancy and improve data consistency. For example, an account information table is designed to store the basic information of the accounts, and a transaction details table is designed to record the specific information of each transaction. During the table structure design process, primary key and foreign key constraints are set for each table to ensure the relevance and integrity of the data. Subsequently, the ETL tool is used to import the standard multi-source financial data temporarily stored in the distributed storage system into the financial database. The ETL tool performs data cleaning, transformation, and verification during the data import process. Data cleaning operations include removing duplicate data, filling in missing values, correcting format errors, etc. Data transformation operations uniformly adjust the data format according to the requirements of the table structure of the financial database. Data verification operations ensure that the data imported into the database complies with the preset business rules and data integrity constraints. Through the automated processing of the ETL tool, the data is efficiently and accurately imported into the financial database. After the financial database is constructed, a professional report generation tool is used to connect to the financial database to generate and display reports. The report generation tool extracts the required data from the financial database through preset report templates and query conditions. The report templates are designed according to the needs of financial analysis, including various common financial statement formats such as balance sheets, income statements, and cash flow statements. The report templates define the data display methods, formatting rules, and dynamic parameter settings to generate corresponding report content according to different query conditions. When the report generation tool extracts data, it performs corresponding query operations from the financial database according to the dynamic parameters set by the user, such as the time range and business type, to obtain the data required for the report. The query operations are performed through optimized SQL query statements to ensure the efficiency and accuracy of data extraction. The extracted data is formatted by the report generation tool and typeset and displayed according to the preset report template style. The report display supports multiple formats, including HTML web format, PDF document format, and Excel spreadsheet format, to meet the needs of different users. Finally, the report generation tool provides an automated report distribution function. Users can configure scheduled tasks to set the report generation and distribution time.For example, at the beginning of each month, a financial statement for the previous month is automatically generated and automatically sent to the relevant personnel in the finance department via email or the enterprise internal messaging system. The report distribution function supports multiple distribution channels to ensure that the reports can be delivered to the target users in a timely and accurate manner.

[0088] As an example of the present invention, refer to Figure 2 As shown, in this example, step S1 includes:

[0089] Step S11: At a plurality of preset hierarchical data collection nodes, collect financial data from different sources at hierarchical time intervals; wherein, the first-layer collection node collects financial bill data at 1-minute intervals; the second-layer collection node collects financial object data at 5-minute intervals; the third-layer collection node collects business type data at 15-minute intervals;

[0090] Step S12: Merge the financial bill data, financial object data, and business type data to generate multi-source raw financial data;

[0091] Step S13: Uniformly process the date fields in the multi-source raw financial data to obtain standardized date data, uniformly process the text fields in the multi-source raw financial data to obtain standardized text data; uniformly process the timestamp fields in the multi-source raw financial data to obtain standardized timestamp data; perform coding mapping processing on the classification fields in the multi-source raw financial data, map different source classification labels to a unified digital code, and obtain standardized classification data;

[0092] Step S14: Integrate the standardized date data, standardized text data, standardized timestamp data, and standardized classification data to obtain standard multi-source financial data;

[0093] Step S15: Match the standard multi-source financial data with the financial scenario and generate a financial scenario identifier.

[0094] In an embodiment of the present invention, a distributed data acquisition system is deployed, and multiple levels of data acquisition nodes are set up. Each node acquires financial data from different sources. The first-level acquisition nodes are configured to acquire financial bill data every 1 minute, connect to the bill module of the financial system, and obtain real-time bill information through the API interface. The second-level acquisition nodes acquire financial object data at intervals of 5 minutes, connect to the ERP system, and receive object data through the message queue. The third-level acquisition nodes acquire business type data at intervals of 15 minutes. The acquired financial bill data, financial object data, and business type data are transmitted to the data processing center. The data processing center uses data fusion technology to merge different types of data according to the preset field mapping rules. During the merging process, the data is associated through a unique identifier (such as a transaction ID) to achieve data consistency and integrity. The merged multi-source original financial data is stored in a distributed storage system. The multi-source original financial data is standardized. For the date field, all dates are uniformly converted to the "YYYY-MM-DD" format using a date formatting tool. For the text field, the Unicode encoding scheme is used for unified encoding. For the timestamp field, it is uniformly converted to the UTC time format, accurate to milliseconds. For the classification field, a classification label mapping table is established to map classification labels from different sources to a unified digital code. For example, "income" is mapped to "1" and "expense" is mapped to "2". The standardized date data, text data, timestamp data, and classification data are integrated. During the integration process, a data verification tool is used to check the integrity and consistency of the data, and each piece of data contains the necessary fields. The integrated standard multi-source financial data is stored in a data warehouse. The data warehouse adopts a star schema and organizes data in the form of fact tables and dimension tables for subsequent query and analysis. The standard multi-source financial data is matched with preset financial scenarios. Through a scenario matching engine, the data is assigned to the corresponding financial scenarios according to the business attributes and financial characteristics of the data. For example, data related to sales business is matched to the "sales scenario", and data related to procurement business is matched to the "procurement scenario". After the matching is completed, a unique financial scenario identifier is generated for each piece of data, and the identifier adopts the UUID format. The generated financial scenario identifier is stored in the data warehouse together with the data for subsequent scenario-based analysis and processing.

[0095] Preferably, step S15 includes the following steps:

[0096] Step S151: Perform interval division processing on the numerical fields in the standard multi-source data to obtain numerical interval identifiers;

[0097] Step S152: Perform time period division processing on the time fields in the standard multi-source data to obtain time period identifiers;

[0098] Step S153: Extract keywords from the text fields in the standard multi-source data, count the occurrence frequencies of the keywords, and classify the text fields according to the occurrence frequencies to obtain category identifiers;

[0099] Step S154: Perform mapping processing on the classification fields in the standard multi-source data to obtain classification field data;

[0100] Step S155: Perform hashing processing on the classification field data based on the numerical range identifier, time period identifier, and category identifier to generate a scenario identifier.

[0101] In the embodiments of the present invention, a data processing framework (such as Apache Spark) is used to divide the numerical fields in the standard multi-source financial data into intervals. According to the statistical distribution characteristics of the numerical fields, a suitable interval division method is selected. For example, for a certain numerical field, a division strategy based on quantiles is adopted to divide it into several intervals. By calculating the quantiles of the field (such as 25%, 50%, 75% quantiles), the interval boundaries are determined. Using the distributed computing power of Spark, large-scale data sets are efficiently processed, mapping the values of each numerical field to the corresponding intervals, and assigning a unique numerical interval identifier to each interval. For example, the interval with a value less than the 25% quantile is marked as "Interval 1", the interval between the 25% and 50% quantiles is marked as "Interval 2", and so on. The time fields in the standard multi-source financial data are divided into time periods. First, the time fields are uniformly converted into a standard time format (such as ISO 8601 format). According to the business requirements and the time granularity of the data, the division method of the time period is determined. For example, for the transaction time field, it is divided according to periods such as days, weeks, and months. Using a time processing library (such as the datetime module in Python or the java.time package in Java), the year, month, day, week, etc. information in the time field is extracted to generate time period identifiers. Keywords are extracted from the text fields in the standard multi-source financial data. Natural language processing tools (such as NLTK or spaCy) are used to tokenize, part-of-speech tag, and extract keywords from the text fields. Through statistical analysis, the occurrence frequency of each keyword in the text field is calculated. According to the frequency distribution of the keywords, the text fields are divided into different categories. For example, the TF-IDF algorithm is used to calculate the importance of the keywords, and the text fields are divided into "high-frequency keyword categories" and "low-frequency keyword categories" according to their values. Through keyword extraction and frequency statistics, a corresponding category identifier is assigned to each text field. Mapping processing is performed on the classification fields in the standard multi-source financial data. Mapping rules for the classification fields are established to uniformly map the classification labels from different sources to predefined digital encodings. For example, "sales" is mapped to "1", "purchase" is mapped to "2", and "reimbursement" is mapped to "3". Through the mapping rules, the original labels of the classification fields are converted into standardized digital encodings to obtain classification field data. This process can be implemented through a configuration file or a database table to ensure that the classification labels from different sources can be accurately mapped to a unified coding system. Hash processing is performed on the classification field data based on the numerical interval identifier, time period identifier, and category identifier. The numerical interval identifier, time period identifier, and category identifier are combined with the classification field data into a unique string. A hash algorithm (such as SHA-256) is used to perform a hash calculation on this string to generate a unique scenario identifier.For example, for a piece of data, its numerical range identifier is "Range 2", the time period identifier is "Month X, Year XX", the category identifier is "High-frequency keyword category", and the classification field data is "Sales", then the combined string is "Range 2_Month X, Year XX_High-frequency keyword category_Sales". The scenario identifier for this piece of data is calculated through a hashing algorithm. The generated scenario identifier is stored in the data warehouse and associated with the original data for subsequent scenario analysis and processing.

[0102] Preferably, the determining of the financial scenario priority ranking according to the financial scenario identifier in step S2 includes:

[0103] Perform a decoding process on the scenario identifier to extract multiple feature fields; the feature fields include a time period field, a numerical range field, a category field, and a scenario type field; the decoding process includes decomposing the scenario identifier into multiple parts, each part corresponding to a feature field;

[0104] Assign weights to the feature fields to obtain feature field weight values; assign a weight of 0.3 to the time period field, a weight of 0.4 to the numerical range field, a weight of 0.2 to the category field, and a weight of 0.1 to the scenario type field;

[0105] Perform a weighted calculation on the feature field weight values to obtain a priority score; the weighted calculation formula is: Priority score = Time period field value × Time period weight + Numerical range field value × Numerical range weight + Category field value × Category weight + Scenario type field value × Scenario type weight;

[0106] Sort the data items according to the priority score and arrange them in descending order to obtain the scenario priority value.

[0107] In the embodiments of the present invention, a data processing tool (such as the re module of Python or a regular expression tool) is used to decode the scene identifier. The scene identifier is a unique string generated by a hash algorithm. The decoding process requires reverse parsing of the string to extract multiple feature fields. The specific operation steps are as follows: Decompose the scene identifier into multiple parts, each part corresponding to a feature field. For example, if the scene identifier is "Period A_Interval B_Category C_Type D", through regular expression matching and splitting operations, the time period field "Period A", the numerical interval field "Interval B", the category field "Category C", and the scene type field "Type D" are extracted. According to the preset weight assignment rules, weight values are assigned to each feature field. The specific weight assignment is as follows: The weight of the time period field is 0.3; the weight of the numerical interval field is 0.4; the weight of the category field is 0.2; the weight of the scene type field is 0.1. Use the weighted calculation formula to calculate the weight values of the feature fields to obtain the priority score. The formula is as follows: Priority score = Time period field value × Time period weight + Numerical interval field value × Numerical interval weight + Category field value × Category weight + Scene type field value × Scene type weight. The specific operation steps are as follows: Convert the feature field values into numerical forms (such as through encoding or mapping) for weighted calculation. For example, the time period field "Period A" can be mapped to the numerical value 1, the numerical interval field "Interval B" can be mapped to the numerical value 2, the category field "Category C" can be mapped to the numerical value 1, and the scene type field "Type D" can be mapped to the numerical value 1. Calculate according to the weighted formula to obtain the priority score of each piece of data. Associate the calculated priority score with the data item, and sort the data items in descending order of the priority score. The specific operation steps are as follows: Create a data structure (such as a list or an array) to store the data items and their corresponding priority scores. Use a sorting algorithm (such as quicksort or mergesort) to sort the data structure to ensure that the data items are arranged in descending order of the priority score. Store the sorted data items and their priority scores in a database or a data warehouse for subsequent analysis and processing.

[0108] Preferably, monitoring the communication network status of the financial cloud platform based on the scene priority value in step S2 includes:

[0109] Determine the communication network status monitoring range corresponding to each scene according to the scene priority value, where the higher the priority value, the larger the monitoring range;

[0110] On multiple nodes of the cloud platform, regularly send test signals according to the communication network status monitoring range, and record the timestamps of sending and receiving; where the test signal is set as a data packet of a fixed size, containing a unique timestamp, to detect the signal round-trip time, the number of signals, and the actual data volume transmitted by the network;

[0111] Collect network status data, including network latency, packet loss rate, and bandwidth utilization; among them, network latency is evaluated by measuring the round-trip time of test signals; packet loss rate is evaluated by counting the number of test signals without received responses; bandwidth utilization is evaluated by monitoring the actual amount of transmitted data;

[0112] Upload the collected network status data to the central monitoring system of the cloud platform, and conduct summary storage to form scenario network status data.

[0113] In the embodiments of the present invention, a resource management system of a cloud platform dynamically allocates the monitoring range of the communication network status according to the scene priority value. The system pre-sets the mapping rules between the priority value and the monitoring range, divides the priority value into multiple intervals, and each interval corresponds to a different monitoring range. For example, scenarios with a priority value between 80 and 100 are allocated the maximum monitoring range, covering all core nodes and key edge nodes of the cloud platform; scenarios with a priority value between 60 and 79 are allocated a medium monitoring range, covering some core nodes and key edge nodes; scenarios with a priority value below 60 are allocated the minimum monitoring range, covering only key edge nodes. According to these rules, the resource management system generates a specific monitoring task configuration file for each scenario, specifying parameters such as the monitoring node list and monitoring frequency. A network status monitoring tool (such as a Ping tool based on the ICMP protocol or a custom UDP / TCP test tool) is deployed on multiple nodes of the cloud platform. These tools can generate data packets of a fixed size, and the size of the data packet is set according to actual needs, for example, set to 128 bytes. Each data packet contains a unique timestamp, and the timestamp format adopts the standard UTC time format, accurate to milliseconds. According to the monitoring range determined by the scene priority value, the monitoring tool periodically sends test signals at a preset frequency. For example, the monitoring frequency for high-priority scenarios is once every 30 seconds, for medium-priority scenarios is once every 2 minutes, and for low-priority scenarios is once every 5 minutes. When the monitoring tool sends a test signal, it records the sending timestamp, and when it receives a response signal, it records the receiving timestamp, and at the same time counts the number of sent signals and the number of signals without receiving a response. The collection of network status data includes three key indicators: network latency, packet loss rate, and bandwidth utilization. Network latency is evaluated by calculating the round-trip time of the test signal, that is, the receiving timestamp minus the sending timestamp. The packet loss rate is evaluated by statistically calculating the ratio of the number of test signals without receiving a response to the total number of sent test signals, and the calculation formula is: packet loss rate = (number of test signals without receiving a response / total number of sent test signals) × 100%. Bandwidth utilization is evaluated by monitoring the actual amount of transmitted data. The monitoring tool records the amount of successfully transmitted data within a certain time interval and compares it with the maximum available bandwidth within this time interval. The calculation formula is: bandwidth utilization = (actual amount of transmitted data / maximum available bandwidth) × 100%. The monitoring tool stores the collected network latency, packet loss rate, and bandwidth utilization data in a local log file, and the log file adopts a structured format, such as CSV format, for easy subsequent data uploading and parsing. A data transfer tool (such as a RESTful API interface based on the HTTP protocol or a message queue system such as Kafka) is used to upload the collected network status data from each monitoring node to the central monitoring system of the cloud platform. The uploaded data includes fields such as scenario identifier, timestamp, network latency, packet loss rate, and bandwidth utilization.After the central monitoring system receives the uploaded data, it performs data parsing and verification to ensure the integrity and accuracy of the data. The parsed data is stored in a distributed database (such as Cassandra or HBase), and the database table structure is designed to include fields such as scenario identifiers, timestamps, network latency, packet loss rate, and bandwidth utilization, facilitating subsequent data query and analysis. The central monitoring system regularly summarizes the stored network status data to generate a time series of scenario network status data, providing data support for network management and optimization of the cloud platform.

[0114] Preferably, the step of detecting the cloud platform resource requirement information from the scenario network status data and determining the financial task resource allocation data according to the cloud platform resource requirement information includes:

[0115] Extract the network latency, packet loss rate, and bandwidth utilization from the scenario network status data to obtain network parameter indicators;

[0116] Detect the storage resource requirements of the cloud platform based on the network parameter indicators; determine the CPU utilization rate, memory utilization rate, and disk I / O utilization rate of the financial task through the storage resource requirements to generate storage resource allocation data;

[0117] Detect the computing resource requirements of the cloud platform based on the network parameter indicators; determine the number of running tasks and resource occupancy of the financial task through the computing resource requirements to generate computing resource allocation data;

[0118] Detect the network resource requirements of the cloud platform based on the network parameter indicators; determine the network bandwidth of the financial task through the network resource requirements to generate network resource allocation data;

[0119] Merge the storage resource allocation data, computing resource allocation data, and network resource allocation data to obtain the financial task resource allocation data.

[0120] In the embodiments of the present invention, a data processing tool (such as Apache Spark or Pandas) is used to extract key metrics such as network latency, packet loss rate, and bandwidth utilization from the scene network status data. These data are stored in a distributed database on the cloud platform, and the data of corresponding fields are extracted from the database through a preset query statement. For example, the SQL statement SELECT network_latency, packet_loss_rate, bandwidth_utilization FROM network_status_data WHERE scene_id = 'a specific scene identifier' is used to obtain the network parameter metrics of a specific scene; by analyzing the network parameter metrics, machine learning algorithms (such as linear regression or decision tree) are used to predict the storage resource requirements. For example, when the network latency is high or the packet loss rate increases, it means that the storage resources are insufficient and the disk I / O utilization needs to be increased. According to the thresholds of the network parameter metrics (such as the network latency exceeding 100 milliseconds or the packet loss rate exceeding 5%), the evaluation of the storage resource requirements is triggered. The specific operations include: using a storage resource monitoring tool (such as Prometheus) to monitor the CPU utilization, memory utilization, and disk I / O utilization in real time. According to the changes in the network parameter metrics, the storage resource allocation is dynamically adjusted. For example, when the network latency exceeds the threshold, the disk I / O utilization is increased to alleviate the storage bottleneck. By analyzing the network parameter metrics, machine learning algorithms are used to predict the computing resource requirements. For example, when the bandwidth utilization approaches the upper limit, it means that the computing resource requirements increase. The specific operations include: using a computing resource monitoring tool (such as Ganglia) to monitor the number of tasks and resource occupancy in real time. According to the changes in the network parameter metrics, the computing resource allocation is dynamically adjusted. For example, when the bandwidth utilization exceeds 80%, computing resources are increased to support the running of more tasks. By analyzing the network parameter metrics, machine learning algorithms are used to predict the network resource requirements. For example, when the network latency or packet loss rate increases, it means that the network bandwidth is insufficient. The specific operations include: using a network resource monitoring tool (such as Zabbix) to monitor the network bandwidth in real time. According to the changes in the network parameter metrics, the network resource allocation is dynamically adjusted. For example, when the network latency exceeds the threshold, the network bandwidth is increased to alleviate the network bottleneck. A data integration tool (such as Apache NiFi) is used to merge the storage resource allocation data, computing resource allocation data, and network resource allocation data. The specific operations include: extracting the storage resource allocation data, computing resource allocation data, and network resource allocation data from the distributed database. Using the data integration tool to merge these data into a unified data structure, for example, creating a table structure that includes the scene identifier, storage resource allocation information, computing resource allocation information, and network resource allocation information. The merged data is stored in the central resource management system of the cloud platform for subsequent resource allocation and management.

[0121] Preferably, the failure identification of the server side of the financial cloud platform according to the financial task resource allocation data in step S3 includes:

[0122] Detect the storage hardware structure in the server side of the financial cloud platform according to the storage resource allocation data: identify the disk type of the storage hardware structure and record the disk read and write speed; perform read and write latency detection on the disk read and write speed to obtain disk read and write failure data;

[0123] Detect the computing hardware structure in the server side of the financial cloud platform according to the computing resource allocation data: identify the number of CPU cores of the computing hardware structure and record the CPU cache usage; perform cache ratio failure detection on the CPU cache usage to obtain CPU cache failure data;

[0124] Detect the network hardware structure in the server side of the financial cloud platform according to the network resource allocation data: identify the network interface type of the network hardware structure and record the network interface exchange volume; perform exchange volume failure detection on the network interface exchange volume to obtain exchange volume failure data;

[0125] Integrate the disk read and write failure data, CPU cache failure data, and exchange volume failure data to generate server side failure data.

[0126] In the embodiments of the present invention, a monitoring tool is used to detect the number of CPU cores of the server-side computing hardware, and the usage of the CPU cache is recorded through the monitoring metric "CPU cache usage". The monitoring tool displays the CPU cache usage in the form of an average value, with the unit of MB. When the CPU cache usage exceeds a preset threshold (such as exceeding 80% of the total cache), it is recorded as CPU cache failure data. The failure data includes the timestamp of the failure occurrence, the number of CPU cores, the current cache usage, and the duration of the failure. A monitoring tool is used to detect the network interface type of the server-side network hardware (such as 1Gbps, 10Gbps, etc.), and the traffic of the network interface is recorded through the monitoring metrics "inbound packet rate of network card" and "outbound packet rate of network card". The monitoring tool displays the traffic of the network interface in the form of an average value, with the unit of pps (packets per second). When the traffic of the network interface exceeds a preset threshold (such as the inbound packet rate or the outbound packet rate exceeding 90% of the interface capacity), it is recorded as traffic failure data. The failure data includes the timestamp of the failure occurrence, the network interface type, the current traffic, and the duration of the failure. A data processing tool (such as Apache NiFi) is used to integrate the disk read / write failure data, the CPU cache failure data, and the traffic failure data. The integrated data is stored in the central failure management system of the cloud platform, and the data structure includes fields: failure type (disk read / write failure, CPU cache failure, traffic failure), failure timestamp, failure duration, failure description, etc. The integrated server-side failure data is stored in a distributed database (such as Cassandra) for subsequent failure analysis and troubleshooting. Through the query function provided by the failure management system, the failure data can be viewed and analyzed in real time to quickly locate and solve the hardware problems on the server side.

[0127] Especially importantly, identify the disk type of the storage hardware structure and record the disk read / write speed;

[0128] Detect the physical interface type of each disk in the storage hardware structure and record its corresponding interface version number;

[0129] In the storage hardware structure, for each disk, perform read / write speed tests in the inner, middle, and outer regions of the disk respectively, and record the read / write speed differences of the disk in different regions through the hardware controller;

[0130] Read the SMART data of the disk, including the number of read / write errors of the disk, the number of remapped sectors, the number of sectors to be remapped, and the disk temperature, and record the cumulative read / write times and power-on time of the disk;

[0131] During the test process, according to the read / write speed differences, use the cache mechanism in the storage hardware structure to test the read / write speeds of the disk with the cache enabled and disabled respectively, and record them as the disk read / write speeds.

[0132] In the embodiments of the present invention, a hardware management tool that supports the detection of multiple storage interfaces is used, such as the built-in hardware detection tool of the system or a third-party tool (such as CrystalDiskInfo). Through this tool, all disk devices in the storage hardware structure are scanned to identify and record the physical interface type of each disk. Common disk interface types include SATA, SAS, M.2, etc. For the SATA interface, its version number is further detected, such as SATA 1.0, SATA 2.0, SATA 3.0, etc.; for the SAS interface, its version number is recorded, such as SAS1.0, SAS2.0, SAS 3.0, etc.; for the M.2 interface, the protocol version it supports is detected, such as PCIe x2, PCIe x4, etc. The interface type of each detected disk and its corresponding version number are stored in the database of the cloud platform. A professional disk performance testing tool is used, such as CrystalDiskMark. First, connect the testing tool to the storage hardware structure and ensure that it can access all disks. Then, for each disk, read and write speed tests are respectively performed in the inner, middle, and outer regions of the disk. The specific operation is as follows: set the test areas of the testing tool to the inner region of the disk (such as the part close to the disk center), the middle region of the disk (the middle part of the disk), and the outer region of the disk (the part close to the disk edge). Read and write operations are respectively performed in each region, and the size of the read and write data blocks can be selected according to the disk type and test requirements. For example, for a mechanical hard disk, larger data blocks can be selected for continuous read and write tests, and for a solid-state drive, smaller data blocks can be selected for random read and write tests. The read and write speeds of each region are recorded by the hardware controller, including sequential read and write speeds and random read and write speeds. The test results are stored in the database of the cloud platform to analyze the read and write speed differences of the disk in different regions. A tool that supports the reading of SMART data is used, such as CrystalDiskMark. After connecting this tool to the storage hardware structure, the SMART data of each disk is read. SMART (Self-Monitoring, Analysis and Reporting Technology) is the self-monitoring, analysis, and reporting technology of the disk, which can provide the health status information of the disk. The specific data read includes the number of read and write errors of the disk, the number of remapped sectors, the number of sectors to be remapped, and the disk temperature. In addition, the cumulative number of read and writes and the power-on time of the disk are also read. These data are usually recorded by the disk controller and provided to the SMART data reading tool. The read SMART data is stored in the database of the cloud platform to monitor and analyze the health status of the disk. Through the management interface or command-line tool of the storage hardware structure, the cache mechanism is configured. First, enable the cache mechanism of the disk, and then use a disk performance testing tool (such as CrystalDiskMark) to perform read and write speed tests on the disk.During testing, select appropriate test parameters such as data block size, test file size, etc. to ensure that the test results can accurately reflect the read and write performance of the disk with caching enabled. Then, disable the caching mechanism of the disk and perform read and write speed tests on the disk again using the same test tools and parameters. Analyze the impact of the caching mechanism on disk performance by comparing the read and write speeds with and without caching enabled. Record the test results in the database of the cloud platform for subsequent evaluation and optimization of disk performance.

[0133] Of particular importance is to identify the network interface type of the network hardware structure and record the network response interface traffic volume;

[0134] Detect the physical interface types of network devices one by one, including RJ45 Ethernet interfaces, fiber optic interfaces, wireless Wi-Fi interfaces, etc., and record the physical connection status of each interface; at the same time, detect the network protocol types supported by each network interface, record the bandwidth specifications of the interface and the supported transmission modes.

[0135] On each network interface, set up a packet monitoring point to record in real time the number of packets, the packet size, and the packet transmission direction passing through the interface, and mark the protocol type of the packets;

[0136] Statistically calculate the total data transmission volume of each network interface within a preset time interval, including the upload data volume and the download data volume, record the network latency and packet loss rate within this time interval, and record the peak and trough values of the traffic of each network interface at different time periods to obtain the network response interface traffic volume.

[0137] In the embodiments of the present invention, the lshw command is used to detect the physical interface type of network devices. Enter lshw - class network in the command line, and the system will list the detailed information of all network interfaces, including the interface type (such as RJ45 Ethernet interface, fiber optic interface, wireless Wi - Fi interface, etc.). At the same time, the ethtool command is used to detect the physical connection status. For example, execute sudo ethtool eth0, and look for Link detected: yes / no in the output result, where yes indicates a normal network connection and no indicates no connection. In addition, the physical connection status can also be detected by reading the / sys / class / net / INTERFACE / carrier file. The file content being 1 indicates a normal connection, and 0 indicates no connection. The protocol types and bandwidth specifications supported by the network interface are obtained through the ethtool command. For example, execute sudo ethtool eth0, and the output result will show the protocol types supported by the interface (such as IPv4, IPv6, etc.), as well as the interface rate (such as 1000Mb / s, 10000Mb / s, etc.) and transmission mode (such as full - duplex, half - duplex). For wireless Wi - Fi interfaces, information such as the wireless protocol type and signal strength can also be viewed through the iwconfig command. Tools such as tcpdump or Wireshark are used to set up packet monitoring points on the network interface. Taking tcpdump as an example, execute sudo tcpdump - i eth0, and it can capture the packets passing through the eth0 interface in real - time. The tool will record the size of each packet, the transmission direction (inbound or outbound), and parse the protocol type of the packet (such as TCP, UDP, ICMP, etc.). The captured packet information is saved to a log file for subsequent analysis. Tools such as iftop or nethogs are used to count the total data transmission volume of the network interface within a preset time interval. For example, execute sudo iftop - i eth0, and the tool will display the upload and download data volumes of the eth0 interface in real - time. At the same time, the network latency is tested through the ping command. Execute ping - c 10 8.8.8.8, and calculate the average round - trip time (RTT) to obtain the network latency. The iperf tool is used to test the network bandwidth and packet loss rate. Execute iperf - c <server_ip>, and the packet loss rate in the test result can be obtained. In addition, by analyzing the packet information in the log file, the peak and trough values of the traffic of each network interface at different time periods are recorded to obtain the traffic exchange volume of the network interface.

[0138] Preferably, the optimization of the server - side fault risk of the financial cloud platform based on the server - side fault data in step S3 includes:

[0139] Locate the disk blocks with read / write failures on the server side of the financial cloud platform according to the disk read / write failure data, and migrate the financial data of the disk blocks with read / write failures to generate disk failure optimization measures;

[0140] Determine the cache kernel control group according to the CPU cache failure data, limit the cache usage of non-critical tasks based on the cache kernel control group, and set critical tasks for real-time scheduling to generate cache failure optimization measures;

[0141] Detect the swap path on the server side of the financial cloud platform according to the swap volume failure data, and dynamically adjust the swap path volume to generate swap volume failure optimization measures;

[0142] Perform failure optimization adjustment on the server side of the financial cloud platform based on the disk failure optimization measures, cache failure optimization measures, and swap volume failure optimization measures to obtain the optimized financial cloud platform.

[0143] In the embodiment of the present invention, a disk detection tool is used to check the health status of the disks on the server side, and key attention is paid to the error count field in the SMART information of the disks. For the detected disk blocks with read / write failures, record their location information; for the financial data on the disk blocks with read / write failures, use a data migration tool to migrate it to a healthy disk. After the migration is completed, update the mount point of the disk to ensure that the system can normally access the new data storage location. Use the cgroups function of Linux to create a cache kernel control group, and assign non-critical tasks to this control group. Set the CPU cache usage limit of the control group to limit the excessive use of the CPU cache by non-critical tasks. Set critical tasks for real-time scheduling to ensure that critical tasks can obtain the execution right first when the CPU resources are tense. Use a network monitoring tool to monitor the network interfaces on the server side in real time to detect whether there is abnormal network interface traffic. Use a network management tool to dynamically adjust the traffic of the network interfaces, set the maximum traffic limit of the network interfaces, and avoid swap volume failures caused by excessive traffic. Integrate the disk failure optimization measures, cache failure optimization measures, and swap volume failure optimization measures into the operation and maintenance management system of the cloud platform. Through automated scripts or operation and maintenance management tools, regularly check the status of the disks, CPU caches, and network interfaces on the server side. Once a failure is detected, automatically execute the corresponding optimization measures. Through the above failure optimization adjustment, ensure that the server side of the financial cloud platform can operate normally in terms of disks, CPU caches, and network interfaces, and improve the stability and reliability of the system.

[0144] As an example of the present invention, refer to Figure 3 As shown, in this example, step S4 includes:

[0145] Step S41: Split the standard multi-source financial data into multiple financial data blocks, with the size of each data block not exceeding 1 MB; upload the financial data blocks to the financial cloud platform one by one through the network interface to obtain financial data upload information;

[0146] Step S42: Define the database table structure based on the financial data upload information, including the table name, field names, and field types, and create a database table, initialize the table structure and indexes; import the financial data blocks into the database table one by one to build a financial database;

[0147] Step S43: Execute an SQL query on the financial database and extract the data that meets the query conditions;

[0148] Step S44: Set the title, table headers, and table footer structure of the report for the data that meets the query conditions to obtain report data, and export the report data as a financial data report.

[0149] In the embodiments of the present invention, a data processing tool (such as the OSS tool of Alibaba Cloud) is used to split the standard multi-source financial data into multiple financial data blocks, and the size of each data block does not exceed 1 MB. The specific operation is as follows: According to the total size of the data file, calculate the number of data blocks after splitting to ensure that the size of each data block does not exceed 1 MB. Upload the financial data blocks one by one to the financial cloud platform through a network interface (such as the HTTP protocol). Use the SDK or API of Alibaba Cloud OSS to upload each data block to the specified OSS bucket. During the upload process, record the upload timestamp, upload status (success or failure), and upload node information of each data block to form financial data upload information. Based on the financial data upload information, use a cloud database platform (such as Huawei Cloud RDS) to define the database table structure. The table structure includes table name (such as financial_data), field names (such as data_block_id, upload_timestamp, upload_status, etc.), and field types (such as data_block_id is INT, upload_timestamp is TIMESTAMP, upload_status is VARCHAR). Use SQL statements to create a database table and initialize the table structure and indexes. For example, use the SQL statement CREATE TABLE financial_data(data_block_id INT PRIMARY KEY,upload_timestamp TIMESTAMP,upload_status VARCHAR(10)) to create a table and create an index for the data_block_id field. Import the financial data blocks into the database table one by one. Use a data import tool (such as the data import function of Huawei Cloud RDS) to import the data blocks into the financial_data table to ensure data integrity and consistency. Execute an SQL query on the financial database to extract data that meets the query conditions. For example, use the SQL statement SELECT*FROM financial_data WHERE upload_status='success' to extract the financial data blocks with successful uploads. Store the query results in a temporary table or in memory for subsequent processing. The extracted data includes information such as the ID of the data block, upload timestamp, and upload status. Set the title, table header, and table footer structure of the report for the data that meets the query conditions. For example, the report title is "Financial Data Upload Report", the table header includes fields such as "Data Block ID", "Upload Timestamp", "Upload Status", etc., and the table footer includes statistical information (such as the total number of successful uploads). Use a report generation tool (such as FineBI) to export the report data as a financial data report. The report can be exported in multiple formats, such as PDF, Excel, or HTML, for users to view and use.

[0150] In this specification, a financial data processing system based on a cloud platform is provided for implementing the above-mentioned financial data processing method based on a cloud platform. The financial data processing system based on a cloud platform includes:

[0151] A financial data collection module, configured to distributively collect multi-source original financial data; perform data standardization on the multi-source original financial data to obtain standard multi-source financial data; match the standard multi-source financial data with financial scenarios and generate financial scenario identifiers;

[0152] A communication network status detection module, configured to determine the priority ranking of financial scenarios according to the financial scenario identifiers to obtain a scenario priority value; monitor the communication network status of the financial cloud platform based on the scenario priority value to obtain scenario network status data; detect the cloud platform resource requirement information with the scenario network status data, and determine financial task resource allocation data according to the cloud platform resource requirement information;

[0153] A server failure optimization module, configured to identify failures of the server side of the financial cloud platform according to the financial task resource allocation data to generate server failure data; perform failure risk optimization on the server side of the financial cloud platform based on the server side failure data to obtain an optimized financial cloud platform;

[0154] A financial database management module, configured to upload the standard multi-source financial data to the optimized financial cloud platform and construct a financial database; present a financial data report according to the financial database.

[0155] Therefore, from any perspective, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the application document are intended to be included in the present invention.

[0156] The above are only specific embodiments of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but rather to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A financial data processing method based on a cloud platform, characterized in that Including the following steps: Step S1: Distributedly collect multi-source original financial data; perform data standardization on the multi-source original financial data to obtain standard multi-source financial data; match the standard multi-source financial data with financial scenarios and generate financial scenario identifiers; Step S2: Determine the financial scenario priority ranking according to the financial scenario identifiers to obtain the scenario priority degree value; monitor the communication network status of the financial cloud platform based on the scenario priority degree value to obtain scenario network status data; detect the cloud platform resource demand information from the scenario network status data, and determine the financial task resource allocation data according to the cloud platform resource demand information; Step S3: Identify faults in the server side of the financial cloud platform according to the financial task resource allocation data, and generate server fault data; Optimize the fault risk of the server side of the financial cloud platform based on the server side fault data to obtain the optimized financial cloud platform; Step S4: Upload the standard multi-source financial data to the optimized financial cloud platform and build a financial database; present a financial data report according to the financial database.

2. The financial data processing method based on a cloud platform according to claim 1, wherein Step S1 includes the following steps: Step S11: Collect financial data from different sources at preset multiple levels of data collection nodes at hierarchical time intervals; among them, the first-level collection nodes collect financial bill data at an interval of 1 minute; the second-level collection nodes collect financial object data at an interval of 5 minutes; the third-level collection nodes collect business type data at an interval of 15 minutes; Step S12: Merge the financial bill data, financial object data, and business type data to generate multi-source original financial data; Step S13: Uniformly process the date fields in the multi-source original financial data to obtain standardized date data, uniformly process the text fields in the multi-source original financial data to obtain standardized text data; uniformly process the timestamp fields in the multi-source original financial data to obtain standardized timestamp data; perform encoding mapping processing on the classification fields in the multi-source original financial data, map different source classification labels to a unified digital code, and obtain standardized classification data; Step S14: Integrate the standardized date data, standardized text data, standardized timestamp data, and standardized classification data to obtain standard multi-source financial data; Step S15: Match the standard multi-source financial data with financial scenarios and generate financial scenario identifiers.

3. The financial data processing method based on a cloud platform according to claim 2, wherein Step S15 includes the following steps: Step S151: Perform interval division processing on the numerical fields in the standard multi-source data to obtain numerical interval identifiers; Step S152: Perform time period division processing on the time fields in the standard multi-source data to obtain time period identifiers; Step S153: Extract keywords from the text fields in the standard multi-source data, count the occurrence frequencies of the keywords, and classify the text fields according to the occurrence frequencies to obtain category identifiers; Step S154: Perform mapping processing on the classification fields in the standard multi-source data to obtain classification field data; Step S155: Perform hash processing on the classification field data based on the numerical interval identifiers, time period identifiers, and category identifiers to generate scenario identifiers.

4. The financial data processing method based on a cloud platform according to claim 1, wherein, The determination of the financial scenario priority ranking described in step S2 includes: Decode the scenario identifier to extract multiple feature fields; the feature fields include a time period field, a numerical range field, a category field, and a scenario type field; the decoding process includes decomposing the scenario identifier into multiple parts, each part corresponding to a feature field; Assign weights to the feature fields to obtain feature field weight values; assign a weight of 0.3 to the time period field, a weight of 0.4 to the numerical range field, a weight of 0.2 to the category field, and a weight of 0.1 to the scenario type field; Perform weighted calculation on the feature field weight values to obtain a priority score; the weighted calculation formula is: priority score = time period field value × time period weight + numerical range field value × numerical range weight + category field value × category weight + scenario type field value × scenario type weight; Sort the data items according to the priority score and arrange them in descending order to obtain the scenario priority value.

5. The financial data processing method based on a cloud platform according to claim 1, wherein The monitoring of the communication network status of the financial cloud platform based on the scenario priority value described in step S2 includes: Determine the communication network status monitoring range corresponding to each scenario according to the scenario priority value, where the higher the priority value, the larger the monitoring range; On multiple nodes of the cloud platform, regularly send test signals according to the communication network status monitoring range and record the timestamps of sending and receiving; the test signal is set as a data packet of a fixed size, containing a unique timestamp, to detect the signal round-trip time, signal quantity, and actual transmission data volume of the network; Collect network status data, including network latency, packet loss rate, and bandwidth utilization; among them, network latency is evaluated by measuring the test signal round-trip time; the packet loss rate is evaluated by counting the number of test signals that do not receive responses; bandwidth utilization is evaluated by monitoring the actual transmission data volume; Upload the received network status data to the central monitoring system of the cloud platform and perform summary storage to form scenario network status data.

6. The financial data processing method based on a cloud platform according to claim 1, characterized in that The detection of the cloud platform resource requirement information from the scenario network status data and the determination of the financial task resource allocation data described in step S2 include: Extract network latency, packet loss rate, and bandwidth utilization from the scenario network status data to obtain network parameter indicators; Detect the storage resource requirements of the cloud platform according to the network parameter indicators; determine the CPU utilization rate, memory utilization rate, and disk I / O utilization rate of the financial task through the storage resource requirements to generate storage resource allocation data; Detect the computing resource requirements of the cloud platform according to the network parameter indicators; determine the number of running tasks and resource occupancy of the financial task through the computing resource requirements to generate computing resource allocation data; Detect the network resource requirements of the cloud platform according to the network parameter indicators; determine the network bandwidth of the financial task through the network resource requirements to generate network resource allocation data; Merge the storage resource allocation data, computing resource allocation data, and network resource allocation data to obtain the financial task resource allocation data.

7. The financial data processing method based on a cloud platform according to claim 1, characterized in that The fault identification of the server side of the financial cloud platform according to the financial task resource allocation data described in step S3 includes: Detect the storage hardware structure in the server side of the financial cloud platform according to the storage resource allocation data: identify the disk type of the storage hardware structure and record the disk read and write speed; perform read and write latency detection on the disk read and write speed to obtain disk read and write failure data; Detect the computing hardware structure in the server side of the financial cloud platform according to the computing resource allocation data: identify the number of CPU cores of the computing hardware structure and record the CPU cache usage; perform cache ratio failure detection on the CPU cache usage to obtain CPU cache failure data; Detect the network hardware structure in the server side of the financial cloud platform according to the network resource allocation data: identify the network interface type of the network hardware structure and record the network interface exchange volume; perform exchange volume failure detection on the network interface exchange volume to obtain exchange volume failure data; Integrate the disk read and write failure data, CPU cache failure data, and exchange volume failure data to generate server-side failure data.

8. The financial data processing method based on a cloud platform according to claim 7, wherein The failure risk optimization of the server side of the financial cloud platform based on the server-side failure data described in step S3 includes: Locate the read and write failure disk blocks on the server side of the financial cloud platform according to the disk read and write failure data, and migrate the financial data of the read and write failure disk blocks to generate disk failure optimization measures; Determine the cache kernel control group according to the CPU cache failure data, limit the cache usage of non-critical tasks based on the cache kernel control group, and set critical tasks to real-time scheduling to generate cache failure optimization measures; Detect the exchange path on the server side of the financial cloud platform according to the exchange volume failure data, and dynamically adjust the exchange path volume to generate exchange volume failure optimization measures; Perform failure optimization adjustment on the server side of the financial cloud platform based on the disk failure optimization measures, cache failure optimization measures, and exchange volume failure optimization measures to obtain the optimized financial cloud platform.

9. The financial data processing method based on a cloud platform according to claim 1, wherein Step S4 includes the following steps: Step S41: Split the standard multi-source financial data into multiple financial data blocks, each data block not exceeding 1MB in size; upload the financial data blocks one by one to the financial cloud platform through the network interface to obtain financial data upload information; Step S42: Define the database table structure based on the financial data upload information, including table name, field name, and field type, and create a database table, initialize the table structure and index; import the financial data blocks into the database table one by one to build a financial database; Step S43: Execute an SQL query on the financial database and extract the data that meets the query conditions; Step S44: Set the data that meets the query conditions with the title, header, and footer structures of the report to obtain report data, and export the report data as a financial data report.

10. A financial data processing system based on a cloud platform, characterized in that, For executing the financial data processing method based on the cloud platform as described in claim 1, the financial data processing system based on the cloud platform includes: A financial data collection module, used for distributedly collecting multi-source original financial data; performing data standardization on the multi-source original financial data to obtain standard multi-source financial data; matching the standard multi-source financial data with financial scenarios and generating financial scenario identifiers; A communication network status detection module, which is used to determine the priority ranking of financial scenarios according to financial scenario identifiers to obtain scenario priority degree values; monitor the communication network status of the financial cloud platform based on the scenario priority degree values to obtain scenario network status data; detect the cloud platform resource demand information from the scenario network status data, and determine the financial task resource allocation data according to the cloud platform resource demand information; A server fault optimization module, which is used to identify faults on the server side of the financial cloud platform according to the financial task resource allocation data to generate server fault data; optimize the fault risk of the server side of the financial cloud platform based on the server side fault data to obtain an optimized financial cloud platform; A financial database management module, which is used to upload standard multi-source financial data to the optimized financial cloud platform and construct a financial database; present financial data reports according to the financial database.

Images