Network cluster data acquisition method and apparatus

Abstract

The embodiments of the present disclosure provide a network cluster data acquisition method and apparatus, the network cluster data acquisition method comprising: acquiring question text, the question text being for at least one network cluster; processing the question text on the basis of a target data processing model to generate response text, wherein the response text comprises answer text and step analysis text, the answer text is generated by the target data processing model on the basis of the step analysis text, and the step analysis text is generated by the target data processing model according to the question text and network cluster data corresponding to each network cluster. By using a target data processing model for data acquisition and processing, the present solution addresses the operational complexity and low efficiency of traditional methods, thereby achieving flexible multi-dimensional, multi-indicator responsiveness, and significantly improving the speed and accuracy of data acquisition.

Description

Network Cluster Data Acquisition Method and Device

[0001] This disclosure claims priority to Chinese Patent Application No. 202411604650.7, filed with the China Patent Office on November 11, 2024, entitled “Network Cluster Data Acquisition Method and Apparatus”, the entire contents of which are incorporated herein by reference. Technical Field

[0002] This disclosure relates to the field of data analysis, and in particular to a method for acquiring network cluster data. Background Technology

[0003] With the rapid development of cloud network technology, time-series metric data analysis plays a crucial role in network operation and maintenance. This is especially true when analyzing network cluster data, a common scenario in daily operations that helps in making rapid analytical decisions. For example, by analyzing the performance of devices in different directions and metrics within a cluster, devices that have performed exceptionally well over a recent period can be identified, providing data support for operation and optimization.

[0004] Currently, traditional methods for acquiring network cluster data mainly rely on the SQL query capabilities of databases, which can solve the problem of acquiring certain network cluster data within a specific time period. However, this method suffers from operational complexity and low efficiency, especially when acquiring and analyzing data across multiple dimensions and indicators, where the flexibility and response speed of SQL queries are limited. Therefore, to address these shortcomings, a new method for acquiring network cluster data is proposed. Summary of the Invention

[0005] In view of this, embodiments of this disclosure provide a method for acquiring network cluster data. One or more embodiments of this disclosure also relate to a network cluster data acquisition apparatus, a computing device, a computer-readable storage medium, and a computer program product, to address the technical deficiencies existing in the prior art.

[0006] According to a first aspect of the present disclosure, a method for acquiring network cluster data is provided, comprising:

[0007] Obtain the issue text, wherein the issue text pertains to at least one network cluster;

[0008] The question text is processed by the target data processing model to generate an answer text, wherein the answer text includes an answer text and a step analysis text. The answer text is generated by the target data processing model based on the step analysis text, and the step analysis text is generated by the target data processing model based on the question text and the network cluster data corresponding to each network cluster.

[0009] According to a second aspect of the present disclosure, a network cluster data acquisition apparatus is provided, comprising:

[0010] The acquisition module is configured to acquire issue text, wherein the issue text is for at least one network cluster;

[0011] The generation module is configured to process the question text based on the target data processing model and generate an answer text, wherein the answer text includes an answer text and a step analysis text. The answer text is generated by the target data processing model based on the step analysis text, and the step analysis text is generated by the target data processing model based on the question text and the network cluster data corresponding to each network cluster.

[0012] According to a third aspect of the present disclosure, a computing device is provided, comprising:

[0013] Memory and processor;

[0014] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, they implement the steps of the above-described network cluster data acquisition method.

[0015] According to a fourth aspect of the present disclosure, a computer-readable storage medium is provided that stores computer-executable instructions, which, when executed by a processor, implement the steps of the network cluster data acquisition method described above.

[0016] According to a fifth aspect of the present disclosure, a computer program product is provided, including a computer program / instructions that, when executed by a processor, implement the steps of the network cluster data acquisition method described above.

[0017] One embodiment of this disclosure implements: obtaining question text, wherein the question text is directed to at least one network cluster; processing the question text based on a target data processing model to generate answer text, wherein the target data processing model is trained based on network cluster data feature information, the network cluster data feature information including at least one network cluster indicator feature information, the network cluster indicator feature information characterizing the actual meaning corresponding to the indicator data in the network cluster data; and determining the target answer information corresponding to the question text based on the answer text.

[0018] The solution proposed in this disclosure, through a novel network cluster data acquisition process, solves the problems of operational complexity and inefficiency associated with traditional database queries. Especially when dealing with cross-analysis involving multiple dimensions and metrics, traditional manual query methods are often limited by flexibility and response speed. The new process not only simplifies operations but also significantly improves processing speed, enabling rapid response and accurate result generation. This automated data analysis approach adapts to more complex network environments, ensuring efficient data acquisition while maintaining system flexibility and adaptability. Attached Figure Description

[0019] Figure 1 is a flowchart of a network cluster data acquisition method according to an embodiment of the present disclosure;

[0020] Figure 2 is an architecture diagram of a network cluster data acquisition system provided in an embodiment of this disclosure;

[0021] Figure 3 is a flowchart of the processing procedure for a method of obtaining the top N device information of a network cluster indicator according to an embodiment of this disclosure;

[0022] Figure 4 is a schematic diagram of the structure of a network cluster data acquisition device provided in an embodiment of the present disclosure;

[0023] Figure 5 is a structural block diagram of a computing device provided in an embodiment of this disclosure. Detailed Implementation

[0024] Numerous specific details are set forth in the following description to provide a full understanding of this disclosure. However, this disclosure can be implemented in many other ways than those described herein, and those skilled in the art can make similar extensions without departing from the spirit of this disclosure. Therefore, this disclosure is not limited to the specific implementations disclosed below.

[0025] The terminology used in one or more embodiments of this disclosure is for the purpose of describing particular embodiments only and is not intended to be limiting of the one or more embodiments of this disclosure. The singular forms “a,” “the,” and “the” as used in one or more embodiments of this disclosure and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in one or more embodiments of this disclosure refers to and includes any or all possible combinations of one or more associated listed items.

[0026] It should be understood that although the terms first, second, etc., may be used to describe various information in one or more embodiments of this disclosure, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, first may also be referred to as second without departing from the scope of one or more embodiments of this disclosure, and similarly, second may also be referred to as first. Depending on the context, the word “if” as used herein may be interpreted as “when”, “in response to a determination”, or “when…”.

[0027] Furthermore, it should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in one or more embodiments of this disclosure are all information and data authorized by the user or fully authorized by all parties. Moreover, the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0028] In one or more embodiments of this disclosure, a large model refers to a deep learning model with a large number of model parameters, typically containing hundreds of millions, tens of billions, hundreds of billions, trillions, or even tens of trillions of model parameters. A large model can also be called a foundation model. It is pre-trained using large-scale unlabeled corpora to produce a pre-trained model with hundreds of millions of parameters. Such models can adapt to a wide range of downstream tasks and have good generalization ability. Examples include Large Language Models (LLMs) and multi-modal pre-training models.

[0029] In practical applications, large models only require a small number of samples to fine-tune the pre-trained model before they can be applied to different tasks. Large models can be widely used in fields such as Natural Language Processing (NLP) and Computer Vision. Specifically, they can be applied to computer vision tasks such as Visual Question Answering (VQA), Image Captioning (IC), and Image Generation, as well as NLP tasks such as text-based sentiment classification, text summarization, and machine translation. The main application scenarios for large models include digital assistants, intelligent robots, search, online education, office software, e-commerce, and intelligent design.

[0030] First, the terms and concepts involved in one or more embodiments of this disclosure will be explained.

[0031] Structured Query Language (SQL) is a standardized language used to perform operations such as querying, inserting, updating, and deleting in relational database management systems. SQL is primarily used to process structured data, i.e., datasets with a defined schema. With SQL, users can perform precise queries on data in tables according to specific syntax structures, achieving complex data filtering and classification. SQL queries can extract the required information from the database based on user-provided conditions and return a result set that meets those conditions. It is widely used in data analysis, system management, and application development.

[0032] SQL queries are a way of sending instructions to a database using specific syntax and returning results. SQL query statements typically include operations such as SELECT, WHERE, and ORDER BY, used to extract data from a table that meets specific requirements. SQL queries can handle joins between single or multiple tables, allowing data filtering by combining multiple conditions.

[0033] Cloud Data Transfer (CDT) is a service used for the efficient and secure transfer of large-scale data between different cloud environments or data centers. CDT supports cross-regional and cross-network data transfer needs, ensuring stable data exchange with low latency between different data nodes. This service is suitable for various scenarios, such as data backup, distributed data processing, and cross-regional data synchronization, improving the overall performance and reliability of data transfer by optimizing transmission paths and bandwidth usage.

[0034] Cloud Enterprise Network (CEN) is an architecture for connecting different network environments, helping enterprises establish global dedicated network connections. CEN supports seamless connectivity across different regions and network environments, and optimizes network traffic transmission paths through intelligent routing mechanisms to ensure fast and secure data transmission. This network architecture is commonly used in scenarios such as interconnecting enterprise data centers and enabling network interoperability across multiple branch offices, providing a flexible and efficient network connectivity solution.

[0035] Data Transfer Plan (DTP) is a bandwidth or traffic service used for sharing among multiple network resources. It aims to help users allocate bandwidth resources across multiple network nodes or clusters. DTP can dynamically adjust bandwidth usage based on the needs of different network clusters, ensuring efficient allocation of network traffic. This service is widely used in bandwidth sharing and traffic optimization scenarios across multiple network environments, helping users flexibly manage network resources and avoid bandwidth waste or overload.

[0036] This disclosure provides a method for acquiring network cluster data, and also relates to a network cluster data acquisition device, a computing device, a computer-readable storage medium, and a computer program product, which will be described in detail in the following embodiments.

[0037] With the rapid development of cloud network technology, time-series metrics analysis has become an indispensable tool in network operation and maintenance. By continuously recording changes in various network metrics, time-series data helps administrators monitor system performance in real time and promptly identify potential problems. In network cluster management, time-series metrics can provide in-depth analysis of multiple key parameters such as network traffic, device performance, latency, and error rate, ensuring the normal operation of the system. For example, in the daily operation and maintenance of data centers, by analyzing device metrics in different directions (such as data inflow and outflow), devices with abnormal performance or excessive load can be quickly identified, thereby effectively allocating resources and optimizing overall network performance. These analytical results not only provide reliable decision-making basis for daily operation and maintenance but also predict potential network bottlenecks or failure points for early prevention.

[0038] However, many current network clusters rely on traditional SQL (Structured Query Language) queries for data acquisition and analysis. While SQL queries perform well when processing static data, their limitations become increasingly apparent when faced with complex, multi-dimensional, and multi-indicator cross-analysis. Especially in network environments, where clusters contain large amounts of real-time, dynamically changing data, SQL queries often struggle to respond quickly, leading to system processing delays and low query efficiency. Furthermore, as network clusters grow in size, manually writing SQL queries to handle the complexity of multi-dimensional analysis increases the operational burden. This approach not only lacks flexibility but also easily leads to untimely analysis results when dealing with complex scenarios, thus impacting decision-making efficiency.

[0039] Referring to Figure 1, Figure 1 shows a flowchart of a network cluster data acquisition method according to an embodiment of the present disclosure, which specifically includes the following steps.

[0040] Step 102: Obtain the issue text, wherein the issue text is for at least one network cluster.

[0041] In practical applications, the question text is the user's query or data retrieval request, which can include a specific time range, metric name, or device information. A network cluster is a collection of multiple computing or network resources that work together to process or manage data; these clusters can be systems composed of physical servers, virtual machines, or cloud computing nodes.

[0042] For example, question text can be understood as a query or request entered by a user when they want to obtain specific data from the system or solve a specific problem. It appears in natural language and involves the acquisition, analysis, or manipulation of certain data. For instance, a user might ask a question like: "What are the top two devices in network cluster A with the highest number of active sessions and the highest number of bytes input per second in the last 10 minutes?" This question text not only specifies the concrete metrics (number of active sessions and bytes input per second) but also limits the time frame of the query (last 10 minutes), aiming to identify devices that meet the criteria. Question text is a core input for analysis and processing in automated systems; by parsing this text, the system determines the range of data to be processed and the query objectives.

[0043] A network cluster, as a collection of resources comprising multiple nodes, is typically used for high-performance computing, data storage, or network management. Each node in a cluster can be a physical server, a virtual machine, or even a containerized application; they collaborate to process large amounts of data or perform tasks. For example, in cloud computing platforms, network clusters often consist of multiple distributed computing nodes to handle tasks such as high-concurrency access requests, data analysis, or machine learning model training. A typical application scenario for network clusters is a Content Delivery Network (CDN), where multiple server nodes are distributed globally to improve data access speed and reliability. Through distributed computing, network clusters can significantly enhance data processing capabilities and system stability.

[0044] It should be noted that obtaining the question text can be understood as the process of extracting or generating data query instructions that the system can process from user input. The specific methods for obtaining the question text can be through natural language questions directly input by the user, such as "What is the number of active sessions in the network cluster in the last 10 minutes?"; or through automated systems that generate question text dynamically based on the user's operation history or predefined query templates; or through speech recognition technology that converts the user's voice instructions into processable text questions for the system to process later, etc. This disclosure does not impose any restrictions on this.

[0045] In one embodiment provided in this disclosure, the question text is "What are the top two device IPs with the highest traffic in the S1 and S2 clusters in the last 10 minutes?", which refers to the cluster1 cluster corresponding to the C1 cluster and the cluster2 cluster corresponding to the C2 cluster.

[0046] By acquiring the question text, the system can accurately understand the user's query needs simply by receiving the natural language written by the user, and automatically parse the key elements in it, thereby generating efficient and targeted data query tasks. This not only simplifies the interaction process between the user and the system and reduces the difficulty of manually entering complex queries, but also improves the system's response speed and processing capabilities.

[0047] Furthermore, obtain the issue text, including:

[0048] Receive question request information input by a user, wherein the question request information includes a question text;

[0049] Input the request question text into the question judgment model and obtain the question judgment result generated by the question judgment model;

[0050] If the model determines that the issue is related to the network cluster, then the requested issue text is identified as issue text.

[0051] In practical applications, the issue request information is the complete content of the query or operation request submitted by the user to the system, including the specific question text. The request question text is the core content of the issue request information, i.e., the actual question raised by the user. The issue judgment model is a model used by the system to analyze and identify the issue type, determining whether the issue is related to the network cluster. The issue judgment result is the output of the issue judgment model after analysis, clarifying the classification or characteristics of the issue for further system processing.

[0052] A problem request can be understood as a complete information package transmitted by a user when interacting with the system. It includes the specific problem the user wants the system to solve and related auxiliary information. For example, in a network monitoring scenario, a user might submit a problem request, such as "What is the traffic situation of the S1 cluster in the last 10 minutes?" This problem request not only includes the query text but may also include parameters such as priority and time range, providing a complete background for the system's subsequent processing.

[0053] The request question text can be understood as the main content of the question request information, that is, the query target directly expressed by the user. The request question text is the core part of the question received by the system and determines the actual content that the system needs to process. For example, a user submitting "CPU utilization of the S1 cluster in the past 10 minutes" is a request question text. The system will judge and analyze this text to determine the required processing steps.

[0054] A problem identification model can be understood as a machine learning or rule-based model used by the system to analyze request text and identify its type. Based on pre-trained data, this model determines whether the request is related to a network cluster or involves other specific topics. For example, when the system receives the question text "abnormal traffic at a certain network node," the problem identification model can identify that the problem is related to network cluster monitoring and use this result for further processing and analysis.

[0055] The problem judgment result can be understood as the result generated by the problem judgment model after analyzing the request problem text. This result is used to determine the type or topic of the problem, providing guidance for subsequent steps. For example, when the problem judgment model analyzes that the problem text is related to network cluster performance, the problem judgment result will indicate "network cluster" as the category of the problem, enabling the system to use a dedicated network cluster analysis module to further process the request.

[0056] It should be noted that if the model determines that the request is unrelated to the network cluster, since the user's input request text is unrelated to the network cluster, the request text can be directly input into the large model to obtain the large model's answer.

[0057] By using a problem-judgment model to determine whether the user-inputted question text is related to the network cluster, and then deciding whether to perform network cluster-related data processing based on the judgment result, the system's processing efficiency and response accuracy can be effectively improved. This not only avoids wasting resources on irrelevant questions but also ensures that the system can focus on processing network cluster-related queries, enhancing the targeting and execution efficiency of problem analysis.

[0058] Step 104: Process the question text based on the target data processing model to generate the answer text, wherein the answer text includes the answer text and the step analysis text. The answer text is generated by the target data processing model based on the step analysis text, and the step analysis text is generated by the target data processing model based on the question text and the network cluster data corresponding to each network cluster.

[0059] In practical applications, the target data processing model is a model trained based on the network cluster data feature information, used to process input data and generate corresponding outputs. The response text is a response generated based on the input question and the model processing results, providing specific analysis steps and conclusions. The network cluster data feature information consists of key parameters used to describe the overall performance or state of the network cluster, including at least one network cluster indicator feature information. The network cluster indicator feature information consists of specific indicator data, characterizing the actual meaning of the indicator data in the network cluster and reflecting the specific performance of different nodes or devices in the network cluster.

[0060] A target data processing model can be understood as a processing tool trained on a large amount of historical data, used to parse user input and output results that meet requirements. In network operations and maintenance, a target data processing model can analyze time-series data, performance metrics, and other information in a network cluster, and provide reasonable analysis and decisions based on user problems. For example, the model can cross-process multi-dimensional data to identify bottleneck devices or high-load nodes in the cluster and automatically generate optimization suggestions or solutions to problems. This type of model has wide applications in modern big data and automated operations and maintenance scenarios.

[0061] The answer text can be understood as a detailed document generated based on the user's input question and the analysis of the target data processing model. The answer text not only provides the results but also demonstrates the entire data analysis process, enabling users to understand how the data was processed and facilitating easy access to the answer information later.

[0062] Network cluster data characteristics can be understood as a dataset describing the overall operational status of a network cluster, containing various key metrics for monitoring cluster performance. These characteristics include network traffic, device latency, and error rates, playing a fundamental role in system analysis and optimization. For example, in a cloud computing environment, network cluster data characteristics can help operations personnel understand the real-time load of each node in the cluster, ensuring stable system operation and providing data support for optimizing resource allocation.

[0063] Network cluster performance metrics can be understood as performance data of specific devices or nodes within a network cluster, used to evaluate the cluster's overall performance. Typical examples include network inbound / outbound bandwidth, bytes per second (bps), and packet loss rate. These metrics provide quantitative data on the behavior of each device in the cluster, enabling the system to identify abnormal nodes or devices and make appropriate adjustments. This type of information is widely used in network monitoring, traffic optimization, and troubleshooting to ensure the overall efficient operation of the cluster.

[0064] In one embodiment provided in this description, the user-input question text follows the example above, and will not be repeated here. The IPs of the two machines with high traffic in network clusters S1 and S2 are identified, and the execution status of each step is listed in detail. The response text output by the target data processing model is:

[0065] "Based on the data you provided, we need to identify the two IP addresses in each cluster with the largest sum of OutBps and InBps. Since we cannot directly obtain the traffic data, we will calculate and identify the target devices using the following steps:"

[0066] First, we extract the OutBps and InBps data for each device from the S1 and S2 clusters respectively:

[0067] S1 cluster device data:

[0068] Device 1: OutBps = 1000, InBps = 500;

[0069] Device 2: OutBps = 2000, InBps = 1000;

[0070] Device 3: OutBps = 1500, InBps = 800.

[0071] S2 cluster device data:

[0072] Device 4: OutBps = 3000, InBps = 1500;

[0073] Device 5: OutBps = 4000, InBps = 2500;

[0074] Device 6: OutBps=1000, InBps=300.

[0075] Next, we calculate the sum of OutBps and InBps for each device:

[0076] S1 cluster traffic calculation results:

[0077] Device 1 flow rate = 1000 + 500 = 1500;

[0078] Device 2 flow rate = 2000 + 1000 = 3000;

[0079] Device 3 flow rate = 1500 + 800 = 1300.

[0080] S2 cluster traffic calculation results:

[0081] Device 4 flow rate = 3000 + 1500 = 4500;

[0082] Device 5 flow rate = 4000 + 2500 = 6500;

[0083] Device 6 flow rate = 1000 + 300 = 1300.

[0084] Then, we sort the traffic in each cluster and identify the two devices with the highest traffic:

[0085] S1 cluster sorting results:

[0086] Device 2: Flow rate = 3000;

[0087] Device 1: Flow rate = 1500.

[0088] S2 cluster sorting results:

[0089] Device 4: Flow rate = 4500;

[0090] Device 5: Flow rate = 6500.

[0091] ##result:

[0092] The two machine IPs with the highest traffic in the current S1 cluster are:

[0093] Device 2 IP:IP2;

[0094] Device 1 IP:IP1.

[0095] The two machine IPs with the highest traffic in the current S2 cluster are:

[0096] Device 4 IP: IP4;

[0097] Device 5 IP: IP5.

[0098] I hope this analysis is helpful! Please let me know if you need further information or other analysis.

[0099] Because the target data processing model is trained based on network cluster data feature information, including network cluster indicator feature information, it can recognize natural language descriptions corresponding to the IDs or names of actual indicators to meet user needs. Compared to using SQL statements, this approach provides a more intuitive way to parse complex queries, reducing the burden on users to manually write query statements and significantly improving query flexibility and efficiency. By directly using natural language input, users can quickly obtain multi-dimensional data analysis results. Especially when dealing with multi-indicator and multi-dimensional cross-analysis, the system can automatically identify relevant data and generate detailed analysis steps and results, avoiding the cumbersome and inefficient operation of SQL statements in complex data queries and providing a more intelligent and convenient user experience.

[0100] Furthermore, the target data processing model is trained using the following steps:

[0101] Obtain network cluster data feature information and an initial data processing model, wherein the network cluster data feature information includes at least one network cluster indicator feature information and network cluster question and answer information;

[0102] The initial data processing model is trained based on the network cluster data feature information to obtain the target data processing model.

[0103] In practical applications, the initial data processing model is a model that has been pre-trained or configured without a specific task; the network cluster question-answering information is a relevant data structure of queries and responses in the network cluster in the past, which helps the system understand and generate answers that meet the user's needs.

[0104] For example, the initial data processing model can be understood as a basic model that has not undergone targeted training. It could be a large model capable of processing general data but not yet optimized for a specific domain or task, or it could be a randomly initialized model, etc. This disclosure does not impose any limitations on this. Network cluster question-and-answer information can be understood as user interaction data generated in the network cluster direction, including the user's input query text and the response information generated by the system based on these queries. For example, when a user asks about the number of active devices in a network cluster, the network cluster question-and-answer information will include the user's natural language query and relevant information such as device status and performance data extracted by the system from the cluster. This question-and-answer information is not merely a record of query results but can also be used by the system to further optimize answers to similar questions. By accumulating a large amount of question-and-answer information, a better target data processing model can be trained subsequently, enabling the target data processing model to better understand user needs, thereby improving the model's responsiveness and accuracy.

[0105] It should be noted that training an initial data processing model based on network cluster data feature information to obtain a target data processing model can be understood as further optimizing the initial model using network cluster data feature information to better adapt it to specific tasks related to network clusters. Specifically, this can be achieved by setting a fine-tuning layer and adjusting its parameters using network cluster data feature information to obtain the target data processing model; alternatively, the initial model's parameters can be adjusted through cross-validation using network cluster data feature information to improve its performance in different scenarios; or by weighting specific network cluster metrics from the network cluster data feature information during training to enable the model to make more accurate predictions and analyses of key performance indicators, etc. This disclosure does not impose any limitations in these areas.

[0106] In one embodiment provided in this disclosure, the network cluster data characteristic information includes:

[0107] Network cluster: [

[0108] cluster1: {

[0109] name: S1 cluster

[0110] Abstract: Cloud Data Transfer (CDT) is used for efficient and secure data transfer between different cloud environments or data centers. S1 clusters are suitable for cross-regional data synchronization, data backup, and distributed data processing scenarios, achieving stable, low-latency data exchange by optimizing transmission paths and bandwidth usage.

[0111] }, cluster2: {

[0112] name: S2 cluster

[0113] Abstract: Cloud Enterprise Network (CEN) is used to connect different network environments, enabling dedicated network interconnection across regions and networks. S2 clusters can optimize data transmission paths through intelligent routing, making them suitable for enterprise data center interconnection and network interoperability needs between multiple branch offices.

[0114] }, cluster3: {

[0115] name: S3 cluster

[0116] Abstract: Data Transfer Plan (DTP) is used to dynamically allocate bandwidth resources among multiple network nodes or clusters. S3 clusters are suitable for scenarios requiring flexible management of network bandwidth, ensuring efficient allocation of network traffic and avoiding bandwidth waste or overload.

[0117] },…];Network cluster metrics: [

[0118] in_bps: Inbound bandwidth (bps), the number of bits transmitted per second in the inbound direction, often used to express the inbound bandwidth of a cloud network.

[0119] out_bps: outbound bandwidth (bps), the number of bits transmitted per second in the outbound direction, often used to express the outbound bandwidth of a cloud network.

[0120] `out_drop_bps`: Outbound packet drop rate (bps). This refers to the number of bits dropped per second in the outbound direction. It's commonly used to indicate packet loss in the outbound direction of a cloud network and is considered an abnormal situation.

[0121] in_slowpath_bps: Inbound slow path bps, the number of bits transmitted per second on the slow path, used to express the bandwidth of cloud network inbound transmission on the slow path, with lower performance than the fast path.

[0122] out_slowpath_bps: Inbound slow [s1] bps, outbound, bits per second transmitted on the slow path, used to express the bandwidth of cloud network outbound transmission on the slow path, with lower performance than the fast path.

[0123] …]; Q&A: [

[0124] Question and Answer 1: {

[0125] Q: What are the top two device IPs with the highest OutBps and ActiveSessionCount in network cluster XX-cluster-2 within the last 15 minutes?

[0126] A: We first obtained the OutBps (bytes per second) and ActiveSessionCount (active sessions) data for the XX-cluster-2 cluster over the past 15 minutes, sorted them, and identified the top two devices. The sorting results showed that the device IPs with higher OutBps were 123.456.789.0 and 123.456.789.1, while the device IPs with higher ActiveSessionCount were 123.456.789.2 and 123.456.789.3. By comparing these two metrics, we found that OutBps and ActiveSessionCount are not simultaneously high, meaning that some devices may have a high number of sessions, but their data output may not be high.

[0127] Q: So, which device IPs have both high OutBps and High ActiveSessionCount?

[0128] A: Next, we analyze devices with high OutBps and ActiveSessionCount. By filtering for devices ranking high in both, we identified the IPs 123.456.789.0 and 123.456.789.2 as having high OutBps and ActiveSessionCount. These two devices not only have a large number of sessions but also a high data output volume, suggesting they may be critical nodes in the system.

[0129] }, Question and Answer 2{

[0130] Q: What are the IPs of the two devices with the highest traffic in network cluster XX-cluster-3 over the past 10 minutes?

[0131] A: We obtained the InBps and OutBps data for the most recent 10 minutes from XX-cluster-3, and then calculated the traffic (OutBps + InBps) for each device. The results show that the device IPs with higher traffic are 123.456.789.4 and 123.456.789.5, indicating that these devices have higher data throughput.

[0132] Q: Do these devices with high traffic also have a large number of active sessions?

[0133] A: Further analysis of the ActiveSessionCount of these devices revealed that device 123.456.789.4 also had a high number of active sessions, indicating that this device not only handles a large amount of data transmission but also has a high number of concurrent sessions. Device 123.456.789.5, on the other hand, had a relatively low number of active sessions, suggesting that its data transmission volume mainly comes from fewer sessions, possibly related to its task type.

[0134] },…], where the data in Questions and Answers: [...] is the network cluster question and answer information, and the data in Network Cluster Indicators: [...] is the network cluster indicator feature information. In this embodiment, it also includes information corresponding to the description of each network cluster, which is the data in Network Cluster: [...].

[0135] Training the initial data processing model with network cluster data feature information, including network cluster question-and-answer information and network cluster indicator features, enables the resulting target data processing model to more accurately understand and handle complex network cluster data query requirements. Because the training process incorporates actual question-and-answer information and various indicators within the network cluster, the model can better capture the relationship between network operation status and user needs, thereby quickly generating accurate answers when faced with queries in different scenarios.

[0136] Furthermore, training the initial data processing model based on the network cluster data feature information to obtain the target data processing model includes:

[0137] Obtain an initial model training layer and an initial model training layer, wherein the initial model training layer is connected to the self-attention layer in the initial data processing model;

[0138] Adjust the parameters of the initial model training layer based on the network cluster data feature information to obtain the target model training layer;

[0139] By combining the target model training layer and the initial data processing model, the target data processing model is obtained.

[0140] In practical applications, the initial model training layer is a structured layer used for preliminary training of the data processing model, responsible for improving the basic performance of the model; the self-attention layer is the neural network layer in the initial data processing model; the target model training layer is a model layer that is further optimized after the initial training is completed, and can accurately handle complex task requirements.

[0141] For example, the initial model training layer can be understood as a structured layer used to capture the overall characteristics of the data in the early stages of model training, often represented by a matrix. The main function of this layer is to help the model learn the basic patterns of the data, such as performance metrics of devices in a network cluster and user behavior data. In network cluster data processing, the initial model training layer identifies and processes a wide range of data features.

[0142] Since the initial data processing model's self-attention layer focuses on extracting relevant information from the input data rather than processing all data dimensions, it is preferable to have a self-attention layer with a rank lower than that of the initial data processing model. This allows the self-attention layer to capture important features more efficiently and reduces computational complexity. This structure not only reduces the model's resource consumption but also improves its performance in specific tasks, especially when processing high-dimensional, noisy data. A lower rank means the model can focus on key data dimensions, reducing interference from redundant information. This allows the model to analyze core indicators more accurately in network cluster data processing, improving the overall efficiency and accuracy of the analysis.

[0143] The target model training layer can be understood as a high-performance layer formed through further optimization after the initial model training. Unlike the initial model training layer, the target model training layer undergoes more detailed training and is capable of handling more complex and specific tasks. For example, in a network cluster, the target model training layer can not only identify basic performance issues of devices but also perform cross-analysis of multi-dimensional performance indicators to identify key factors affecting the overall efficiency of the network. This layer enables the model to have higher accuracy and adaptability, providing more insightful analytical results in complex network environments.

[0144] It should be noted that the initial data processing model is a neural network model that includes a self-attention layer, and the target data processing model is a neural network model that connects the self-attention layer to the target model's training layer.

[0145] Obtaining the target model training layer by adjusting the parameters of the initial model training layer based on network cluster data feature information can be understood as making the model more adaptable to specific network cluster data processing tasks by adjusting various parameters of the initial training model. Specific methods include freezing the parameters of the initial model and adjusting only the parameters of the initial model training layer to improve the model's generalization ability without increasing computational resources, etc. This disclosure does not impose any limitations on this approach.

[0146] By adjusting the parameters of the initial model training layer to obtain the target model training layer and generating the target data processing model, the model can be made more adaptable to specific network cluster data processing tasks while reducing model training costs.

[0147] Furthermore, the question text is processed based on the target data processing model to generate the answer text, including:

[0148] Based on the question text, obtain network cluster data corresponding to at least one network cluster;

[0149] The question text and data from each network cluster are processed based on the target data processing model to generate the answer text.

[0150] In practical applications, obtaining network cluster data corresponding to at least one network cluster based on the question text can be understood as extracting network cluster data related to the question by parsing the user-inputted question. Specific methods include using a target data processing model to analyze the question text, automatically identifying relevant network clusters and extracting their data; manually or automatically mapping keywords in the question text to network clusters using preset keyword matching rules to extract the corresponding data; or directly retrieving data from a database or API interface using a user-specified network cluster ID or name, etc. This disclosure does not impose any restrictions in these areas.

[0151] By automatically acquiring network cluster data corresponding to the problem text, the complexity and error rate of manual operations can be reduced, improving the efficiency of data query and analysis. The system can quickly parse the text entered by the user, automatically identify the relevant network clusters, and extract the corresponding real-time data for processing, ensuring that the analysis results are more timely and accurate. This not only speeds up problem resolution but also provides maintenance personnel with a more intelligent tool, enabling them to focus on high-value decisions and optimizations without having to manually search for and input tedious network data.

[0152] Furthermore, based on the question text, at least one network cluster data corresponding to a network cluster is obtained, including:

[0153] Obtain a cluster determination prompt word template, and based on the cluster determination prompt word template and the question text, obtain the cluster determination prompt word corresponding to the question text, wherein the cluster determination prompt word template is a formatted template used to enable the data processing model to generate interface call information;

[0154] The cluster determination prompt is input into the target data processing model, and the data acquisition interface call information output by the target data processing model is obtained, wherein the data acquisition interface call information corresponds to at least one network cluster;

[0155] Based on the data acquisition interface call information, obtain the network cluster data corresponding to each network cluster.

[0156] In practical applications, the data acquisition interface call information is an instruction obtained from the target data processing model, which includes the interface parameters required to call the relevant data of the network cluster; the cluster determination prompt word template is a predefined template for generating specific cluster query requests, which helps the target data processing model generate effective prompt words based on the question text; the cluster determination prompt word is a specific request generated based on the question text and the prompt word template, which is used to obtain data from a specific network cluster.

[0157] For example, data acquisition interface call information can be understood as a call instruction generated by the target data processing model for extracting network cluster data. This information includes the required interface, query parameters, and the corresponding cluster ID, ensuring accurate retrieval of real-time data from the corresponding cluster. For instance, when a user queries "the number of active sessions in a certain cluster," interface call information is generated based on the model, specifying the cluster ID, query metric, and time range, and the relevant API is called to retrieve the data. This information is a crucial step in the data acquisition process, ensuring the accuracy and efficiency of the query.

[0158] The cluster query prompt template can be understood as a formatted template used by the system to generate cluster data query requests. This template contains a fixed set of query formats and variable positions. When the system receives a user's question, it fills the question text into the corresponding positions in the template, generating a specific cluster query prompt. For example, when a user asks "What is the traffic situation of cluster B in the last 10 minutes?", the template will embed "What is the traffic situation of cluster B in the last 10 minutes?" as a variable into the template, thus forming a complete query prompt for subsequent system processing.

[0159] Cluster identification prompts can be understood as specific query requests generated based on the question text and prompt templates, used to guide the system to retrieve relevant cluster data from the target data processing model. For example, when a user asks the question "What devices in cluster A have a large number of active sessions in the last 10 minutes?", the system will generate a prompt such as "Please tell me how to obtain the data of the cluster described in 'Cluster A has a large number of active sessions in the last 10 minutes' within the corresponding time period by calling the 'Metric API' and 'Device Information API'." This prompt accurately guides the data processing model to execute the query, thereby efficiently obtaining the required cluster data.

[0160] In one embodiment provided in this disclosure, the cluster identification prompt template is: Please call the interface "Metric API and Device Information API" to obtain the data call information of the cluster described in the "<Problem Text>" within the corresponding time period. Following the example above, the generated cluster identification prompt is: Please call the interface "Metric API and Device Information API" to obtain the data call information of the cluster described in "What are the top two device IPs with the highest traffic in the S1 and S2 clusters in the last 10 minutes?" within the corresponding time period.

[0161] The cluster identification prompt is then input into the target data processing model. Since the data processing model is trained based on cluster data feature information, it can identify the target network cluster in the question text and output API call information. The API call information output by the target data processing model is as follows:

[0162] By automatically acquiring network cluster data corresponding to the question text through the target data processing model, the efficiency and accuracy of data processing can be significantly improved. The system can quickly parse the questions raised by users, automatically identify relevant clusters and corresponding indicators, and directly retrieve the required data from the interface, avoiding the errors and delays that may be caused by manual queries. This not only simplifies the operation process but also ensures the real-time nature and accuracy of query results, providing users with more intuitive and rapid feedback, and helping to make more efficient monitoring and optimization decisions in complex network environments.

[0163] Furthermore, based on the target data processing model, the question text and data from each network cluster are processed to generate the answer text, including:

[0164] Obtain a question answer prompt template, wherein the question answer prompt template is used to enable the target data processing model to output answer text based on the question text;

[0165] Based on the question answer prompt template, the question text, and data from each network cluster, generate question answer prompts;

[0166] Input the question answer prompts into the target data processing model to obtain the answer text generated by the target data processing model.

[0167] In practical applications, the question-answer prompt template is a predefined framework that guides the system to process user-submitted questions and perform detailed analysis of network cluster data. The question-answer prompts are detailed analysis instructions generated based on the question-answer prompt template and the specific question text, used to guide the system in performing multi-step data processing and outputting results.

[0168] For example, a question-and-answer prompt template can be understood as a structured template used by the system to generate detailed data analysis steps. This template provides clear operational steps for the target data processing model, ensuring that the system can fully process and analyze the network cluster data based on the question text. The template typically includes steps such as "extract data," "sort analysis," and "interpret results," helping the system process complex problems in stages. For instance, the question-and-answer prompt template might require the system to first extract specific indicators from the network cluster, then calculate these indicators, and output analytical results with explanations. The question-and-answer prompts can be understood as specific analysis instructions generated based on the question text and the question-and-answer prompt template. The system fills the relevant parts of the template with the question text to form a complete set of analysis steps.

[0169] In one embodiment provided in this disclosure, the question-answer prompt template is:

[0170] Based on the network cluster performance data and the input question '<question text>', please conduct a comprehensive analysis following these steps:

[0171] 1. Extract performance-related data from the network cluster, such as CPU utilization and memory consumption.

[0172] 2. Perform statistical and analytical operations on these performance data according to the problem requirements.

[0173] 3. Classify or sort the analyzed data to identify the devices or nodes that meet the requirements.

[0174] 4. Describe in detail the analysis process and related intermediate results for each step.

[0175] 5. Output the analysis results and explain the relationship between these results and cluster performance.

[0176] Network Cluster: "<Network Cluster Data>", following the example above, then fill in the corresponding part with the question text from the previous example and the network cluster data to obtain the question answer prompts, which will not be elaborated here.

[0177] By using question-and-answer prompt templates, the target data processing model can directly process the question text and obtain the answer text, thereby improving the system's automation capabilities and response speed. The templates provide the model with a structured analysis path, enabling it to systematically parse the question text and perform multi-step data processing, making the entire process from extracting relevant data to generating results more accurate and efficient.

[0178] Furthermore, based on the target data processing model, the question text and data from each network cluster are processed to generate the answer text, including:

[0179] Based on the question text, obtain the answer step information corresponding to the question text;

[0180] The question text, the answer step information, and the data of each network cluster are input into the target data processing model to obtain the answer text generated by the target data processing model.

[0181] In practical applications, the answer step information is a structured set of steps generated by the system when processing the question text to guide the model in completing the answer step by step.

[0182] For example, the answer step information can be understood as a set of operations and analysis steps that the model needs to perform when generating the answer text. It provides a clear path for the target data processing model, and the entire process from question parsing, data extraction, processing to result generation relies on the guidance of this information. For instance, when a user asks "the devices with the highest traffic in the network cluster in the last 10 minutes," the answer step information includes operations such as extracting relevant cluster data, calculating traffic peaks, and sorting devices. This information ensures that the system can execute the necessary analysis steps in sequence to generate an accurate answer that meets the user's needs.

[0183] It should be noted that obtaining answer step information from the question text can be understood as generating the steps required for the system to perform data processing and analysis by analyzing the question text. Specifically, this can be achieved by using a target data processing model to analyze the question text and determine its corresponding answer step information; by matching predefined step templates based on the question's keywords to quickly generate a processing path; or by extracting commonly used steps from answers to similar historical questions and adjusting them for the current question, etc. This disclosure does not impose any limitations on these methods.

[0184] Furthermore, inputting the question text, answer step information, and data from each network cluster into the target data processing model to obtain the answer text generated by the target data processing model can be understood as automatically generating analysis results and answers by combining the user-input text, system-generated step information, and real-time network cluster data. Specifically, this can be achieved by simply concatenating the question text, answer step information, and data from each network cluster into the target data processing model using a simple prompt template, allowing the target data processing model to generate the answer text, etc. This disclosure does not impose any limitations on this approach.

[0185] In one embodiment provided in this disclosure, the question text, answer step information and data of each network cluster are linked together by a question answer prompt template that includes a reminder of the answer steps. The answer prompt template here is: Please process the following network cluster data according to the steps described below to answer the <question text>.

[0186] The steps to follow are: <Answer the step information>;

[0187] Network Cluster: <Network Cluster Data>, then the question text, the obtained answer step information, and the network cluster data are filled into the corresponding parts to obtain question answer prompts containing answer step reminders, so that the target data processing model can output answer text, which will not be elaborated here.

[0188] By acquiring the answer step information and inputting the prompts for the answer steps into the target data processing model to obtain the answer text, the model's processing efficiency and accuracy can be improved. In this way, the system can systematically process the question text and related data according to preset analysis steps, ensuring the execution order and logic of each step and reducing confusion and uncertainty in the data processing process. This not only ensures that the system-generated answer text better meets user needs but also provides a more transparent and traceable analysis path, facilitating the generation of more targeted results in complex query tasks while reducing the need for manual intervention.

[0189] Furthermore, based on the question text, the answer step information corresponding to the question text is obtained, including:

[0190] Obtain the step-by-step prompt template corresponding to the question text;

[0191] Based on the step-by-step parsing prompt template and the question text, step-by-step parsing prompts are generated;

[0192] The step-by-step prompts are input into the target data processing model to obtain the answer step information generated by the target data processing model.

[0193] In practical applications, the step-by-step parsing prompt template serves as a framework guiding the system to generate processing steps. It is used to parse user questions and generate analysis steps. The step-by-step parsing prompts are specific instructions generated based on the template, helping the system to progressively execute the problem processing and analysis until a result is generated.

[0194] For example, a step-by-step parsing prompt template can be understood as a preset template used to guide the system in parsing the question and generating execution steps. The structured content in the template ensures that the system can break down the question step by step, instructing the target data processing model on how to extract key information from the question text and generate appropriate analysis steps. For instance, when a user inquires about network performance data, the template requires the system to first extract traffic metrics, then sort and analyze them, and finally generate the results. Through such a template, the system can better transform the question into an executable task, ensuring a clear and explicit processing path.

[0195] Step-by-step prompts can be understood as instructions generated based on a specific question and a step-by-step prompt template, clearly defining the steps the system should perform. The system fills in the template based on the question text, generating prompts for operations including data extraction, sorting, and calculation. For example, if a user asks "What is the network traffic situation of cluster S1 in the last 10 minutes?", the step-by-step prompts might generate something like: "Sort the network traffic data from the cluster data, identify devices with high traffic, and generate analysis results." This prompt provides the system with a step-by-step execution path, ensuring the model can accurately and efficiently answer complex questions.

[0196] It should be noted that obtaining the step-by-step prompt template corresponding to the question text can be understood as generating a preset template suitable for question parsing by analyzing the content of the question text, thereby helping the system to gradually solve the user's problem. Specifically, this can be achieved by determining the preset step-by-step prompt template as the one corresponding to the question text; it can also be achieved by analyzing the keywords and structure in the question text using a target data processing model and selecting the most matching question parsing template from an existing template library; or it can be achieved by combining historical data and the answer steps of similar questions to automatically match and adjust the template content to optimize the question processing process, etc. This disclosure does not impose any limitations on these methods.

[0197] In one embodiment provided in this disclosure, the problem text follows the example above: What are the IPs of the top two devices with the highest traffic in the S1 and S2 clusters in the last 10 minutes? The step-by-step prompt template is: Please follow these steps to determine the steps required to solve the <problem text>: 1. Analyze the problem text and extract the main analysis tasks. 2. Determine the key data processing steps that need to be performed. 3. List the various operation steps required to solve the problem, and ensure that each step is clear and logical; then fill the problem text into the above step-by-step prompt template to obtain the step-by-step prompts, and then input the prompts into the target data processing model to obtain the answer step information output by the target data processing model: 1. Extract the OutBps and InBps data for the last 10 minutes from the S1 and S2 clusters. 2. For each device in each cluster, calculate the sum of OutBps and InBps (traffic = OutBps + InBps). 3. Sort the devices in each cluster by traffic and find the top two devices with the highest traffic. 4. Record the IPs of these devices and output the results, displaying the IP addresses of the devices with the highest traffic in the S1 and S2 clusters.

[0198] By parsing prompts to extract the steps required to resolve a problem, the problem-solving process can be effectively simplified, ensuring that the system executes each step logically and systematically. This approach helps the system quickly identify the key steps needed, avoiding omissions or duplication, thereby improving the efficiency and accuracy of data processing. Furthermore, it makes problem analysis more transparent and controllable, providing users with a clear execution path and result explanations, facilitating more precise analysis and problem-solving in complex tasks.

[0199] To improve efficiency during subsequent problem analysis, after processing the question text based on the target data processing model and generating the answer text, the method further includes:

[0200] Based on the response text, determine the target answer information corresponding to the question text.

[0201] In practical applications, the target answer information is the information actually displayed to the user. It can be understood as the actual output generated by the system after processing the question text, for the user's reference or use. Depending on the user's query, the target answer information may include direct numerical results, charts, or further analysis and explanations. For example, in network cluster management, a user might query the performance data of a cluster over the past 10 minutes; the target answer information might display the traffic, latency, and load of the relevant devices. If the user prefers to view the data in chart form, the target answer information might generate a visual chart to show data trends. In simpler query scenarios, the system might only return a numerical value or a brief result. This flexibility ensures that the target answer information can meet the needs of users in different scenarios.

[0202] It's important to note that the display method of the target answer information can be diverse, depending on the usage scenario and user needs. In other words, the target answer information is determined based on the answer display information sent along with the question text. For example, when receiving a question text from technical operations personnel, the target answer information might be displayed through the interface of an operations and maintenance dialogue assistant, providing real-time feedback on the problem. For question texts sent by administrators, the target answer information might be displayed in the form of a data dashboard, providing a summary of the overall network cluster's operational status. In specific scenarios, the target answer information may also only display the numerical values ​​or status of the analysis results, avoiding excessive information interference. This flexible display method allows the target answer information to not only adapt to various usage scenarios but also provide different users with feedback that aligns with their usage habits.

[0203] Optionally, based on the answer text, the target answer information corresponding to the question text is determined, including:

[0204] Based on the answer text, the target answer information corresponding to the question text is generated.

[0205] In practical applications, the analysis step text is a detailed description of each step the system performs when generating the final answer, demonstrating the specific operations the system takes during problem processing. The answer text is the final response generated by the system through these steps, directly addressing the user's query.

[0206] For example, analysis step text can be understood as a record of the system's execution process when processing a question text. This text demonstrates how the system solves the problem step by step and is typically used to explain the logic behind the solution. For instance, when a user inquires about the performance of a network cluster, the system might first extract the relevant data, then perform calculations and sorting, and finally draw a conclusion. Analysis step text records these operations in detail, allowing the user to understand the problem's resolution path. This is especially important for complex technical problems, helping users understand how the system arrives at its conclusions and ensuring process transparency.

[0207] Answer text can be understood as the final result obtained by the system based on the executed steps, used to directly respond to the user's query. For example, when a user asks, "What are the active devices in the network cluster in the last 10 minutes?", the answer text will include the specific IP addresses and performance metrics of these active devices. Answer text is usually concise and direct, aiming to provide users with a clear and unambiguous answer, reducing unnecessary additional information. It is the core result that the user needs, which can be a numerical value, status, or execution summary, etc.

[0208] It should be noted that generating target answer information based on answer text can be understood as transforming the answer text generated by the target data processing model into a display format that is easy for users to understand. Specific methods include visualizing the data in the answer text, such as generating charts or trend analyses, to help users intuitively understand complex data relationships; presenting key information in the answer text in a concise summary format, suitable for scenarios requiring rapid decision-making; or creating interactive data dashboards, allowing users to view and analyze different dimensions of data in the results in real time, etc. This disclosure does not impose any limitations in this regard.

[0209] By transforming answer text into diverse target answer information, it can not only adapt to various usage scenarios, but also allow different users to obtain feedback that is more in line with their usage habits according to their needs.

[0210] The solution proposed in this disclosure addresses the efficiency issues of traditional methods in multi-dimensional data cross-processing, particularly when dealing with complex problems. It automatically generates analysis steps and answer text, simplifying user interaction. Unlike previous methods of manually writing queries, the new data processing model automatically calls relevant data interfaces to obtain real-time data from the network cluster based on the question text and corresponding prompts. It then quickly parses the data based on the generated answer prompts, improving both the accuracy and speed of data acquisition. Furthermore, it automates the solution process for complex data analysis, reducing human intervention and making the system more suitable for diverse application scenarios. This approach provides users with an efficient and flexible solution, enhancing the ability to acquire and analyze network cluster data.

[0211] Referring to Figure 2, Figure 2 shows an architecture diagram of a network cluster data acquisition system provided in an embodiment of the present disclosure. The network cluster data acquisition system may include a client 100 and a server 200.

[0212] Client 100 is used to send a problem text to server 200 for at least one network cluster;

[0213] Server 200 is used to obtain the question text, process the question text based on the target data processing model, and generate the answer text. The answer text includes the answer text and the step analysis text. The answer text is generated by the target data processing model based on the step analysis text. The step analysis text is generated by the target data processing model based on the question text and the network cluster data corresponding to each network cluster. The server 200 sends the answer text to client 100.

[0214] Client 100 is also used to receive the response text sent by server 200.

[0215] The solution implemented in this disclosure significantly optimizes the network cluster data processing workflow through efficient client-server collaboration. The client can send question text to the server, which then automatically processes and generates corresponding response text or target answers based on a trained target data processing model. This design eliminates the burden of complex manual queries and data processing found in traditional methods. The server can not only quickly obtain key information based on the multi-dimensional data characteristics of the network cluster but also automatically parse the data meaning and generate detailed answers. The client, by receiving the response or answer text from the server, further simplifies the data interaction process and improves the efficiency of users obtaining accurate data. The entire solution, through efficient client-server interaction, significantly enhances the processing capability for complex data, making the system applicable to a wider range of network cluster application scenarios.

[0216] A network cluster data acquisition system may include multiple clients 100 and a server 200. Clients 100 can be referred to as end-side devices, and server 200 can be referred to as cloud-side devices. Multiple clients 100 can establish communication connections through server 200. In a network cluster data acquisition scenario, server 200 is used to provide network cluster data acquisition services between multiple clients 100. Each client 100 can act as a sender or receiver, communicating through server 200.

[0217] Users can interact with server 200 through client 100 to receive data sent by other clients 100, or send data to other clients 100, etc. In a network cluster data acquisition scenario, users can publish data streams to server 200 through client 100, and server 200 can generate response text based on the data stream and push the response text to other clients that have established communication.

[0218] In this system, client 100 and server 200 establish a connection via a network. The network provides the medium for communication between client 100 and server 200. The network can include various connection types, such as wired or wireless communication links or fiber optic cables. Data transmitted by client 100 may need to undergo encoding, transcoding, compression, or other processing before being published to server 200.

[0219] Client 100 can be a browser, an app (application), a web application such as an H5 (HyperText Markup Language 5) application, a lightweight application (also known as a mini-program), or a cloud application. Client 100 can be developed based on the software development kit (SDK) of the corresponding service provided by server 200, such as a real-time communication (RTC) SDK. Client 100 can be deployed on electronic devices and depends on the device or certain apps on the device to run. Electronic devices may have displays and support information browsing, such as personal mobile terminals like mobile phones, tablets, and personal computers. Various other types of applications can also be configured on electronic devices, such as human-computer interaction applications, model training applications, text processing applications, web browser applications, shopping applications, search applications, instant messaging tools, email clients, and social media platform software.

[0220] Server 200 may include servers providing various services, such as servers providing communication services to multiple clients, servers supporting backend training of models used on clients, and servers processing data sent by clients. It should be noted that server 200 can be implemented as a distributed server cluster composed of multiple servers, or as a single server. The server can also be a server in a distributed system, or a server integrated with blockchain. The server can also be a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms, or an intelligent cloud computing server or intelligent cloud host with artificial intelligence technology.

[0221] It is worth noting that the network cluster data acquisition method provided in this disclosure is generally executed by the server. However, in other embodiments of this disclosure, the client may also have similar functions to the server, thereby executing the network cluster data acquisition method provided in this disclosure. In other embodiments, the network cluster data acquisition method provided in this disclosure may also be executed jointly by the client and the server.

[0222] The following description, in conjunction with Figure 3, uses the application of the network cluster data acquisition method provided in this disclosure in acquiring the top N device information items of network cluster indicators as an example to further illustrate the network cluster data acquisition method. Figure 3 shows a flowchart of the processing procedure of a method for acquiring the top N device information items of network cluster indicators according to an embodiment of this disclosure, specifically including the following steps.

[0223] Step 302: Obtain the question text: "What are the IPs of the two machines with the largest number of active sessions and the largest number of bytes input per second in network cluster A in the last 10 minutes?"

[0224] Step 304: Combine the cluster determination prompt word template with the above question text set to generate cluster determination prompt words. Input these prompt words into the target data processing model and obtain the data acquisition interface call information output by the data processing model.

[0225] In this embodiment, the prompt template is: Please tell me how to obtain the data of the cluster described in the "<problem text>" within the corresponding time period by calling the interface "Metric API and Device Information API"; the obtained prompt is: Please tell me how to obtain the data of the cluster described in "What are the two machine IPs with the largest number of active sessions and the largest number of input bytes per second in network cluster A in the last 10 minutes?" by calling the interface "Metric API and Device Information API" within the corresponding time period.

[0226] Based on the above, the data processing model outputs the following data retrieval interface call information:

[0227] Step 306: Use the above interface call information to call the cluster information retrieval interface to obtain network cluster data of network cluster A within 10 minutes.

[0228] In this embodiment, the obtained network cluster data is exemplified as follows:

[0229] Step 308: Combine the question answer prompt template, network cluster data, and the question text mentioned above to generate question answer prompts.

[0230] In this embodiment, the answer prompt template is:

[0231] "Based on the network cluster data and the question '<Question Text>', please conduct a detailed analysis following these steps:"

[0232] 1. Extract relevant indicator information from network cluster data.

[0233] 2. Based on the problem requirements, perform necessary data processing and calculations on the extracted indicators.

[0234] 3. Sort or perform other statistical operations on the processed data to identify devices or values ​​that meet the criteria.

[0235] 4. Provide a detailed explanation of the execution process and intermediate results for each step.

[0236] 5. Output the analysis results and explain the meaning of the results and related context.

[0237] Network cluster data: "<Network cluster data>" Then, fill in the above question text and network cluster data into the corresponding part to obtain the question answer prompt, which will not be elaborated here.

[0238] Step 310: Input the above question answer prompts into the target data processing model so that the target data processing model can analyze the question text and generate an answer text including the analysis steps.

[0239] In this embodiment, the obtained answer text is:

[0240] "Based on the data you provided, we need to identify the two IP addresses with the highest ActiveSessionCount and InBps for each machine. This can be done through the following steps:"

[0241] ## Find the two machine IPs with the larger ActiveSessionCount:

[0242] First, we sort all machines by ActiveSessionCount from largest to smallest, and then take the first two:

[0243] IP3-ActiveSessionCount:3.4833333;

[0244] IP4-ActiveSessionCount:0.8108108.

[0245] Find the two machine IPs with higher InBps:

[0246] Next, we will sort all machines from largest to smallest in InBps and take the first two:

[0247] IP1-InBps: 78912.18;

[0248] IP2-InBps:53636.246.

[0249] ##result:

[0250] The two machine IPs with the largest ActiveSessionCount are IP3 and IP4.

[0251] The two largest machine IPs in InBps are IP1 and IP2.

[0252] I hope this analysis is helpful! Please let me know if you need further information or other analysis.

[0253] Where ActiveSessionCount is the ID of the number of active sessions in the network cluster data, and InBps is the ID of the number of bytes input per second in the network cluster data.

[0254] Step 312: Extract the answer data from the answer text, and generate the IP addresses of the first two devices based on the answer data.

[0255] The solution applied in this disclosure combines question text and network cluster data to automatically generate question answer prompts. The target data processing model can then perform detailed data processing, sorting, and calculations based on different indicator requirements, thereby accurately obtaining device information that meets the criteria. This method not only improves the efficiency of multi-dimensional, multi-indicator data processing but also ensures the transparency of the analysis steps. Users can clearly understand each step of the analysis process and intermediate results, obtaining accurate device IP addresses and related indicator data. The automated analysis and answer generation process significantly reduces the complexity of manual operations and adapts to dynamic network environments.

[0256] Corresponding to the above method embodiments, this disclosure also provides an embodiment of a network cluster data acquisition device. Figure 4 shows a schematic diagram of the structure of a network cluster data acquisition device provided in one embodiment of this disclosure. As shown in Figure 4, the device includes:

[0257] The acquisition module 402 is configured to acquire problem text, wherein the problem text is for at least one network cluster;

[0258] The generation module 404 is configured to process the question text based on the target data processing model to generate an answer text, wherein the answer text includes an answer text and a step analysis text. The answer text is generated by the target data processing model based on the step analysis text, and the step analysis text is generated by the target data processing model based on the question text and the network cluster data corresponding to each network cluster.

[0259] Optionally, the acquisition module 402 is further configured to:

[0260] Receive question request information input by a user, wherein the question request information includes a question text;

[0261] Input the request question text into the question judgment model and obtain the question judgment result generated by the question judgment model;

[0262] If the model determines that the issue is related to the network cluster, then the requested issue text is identified as issue text.

[0263] Optionally, the generation module 404 is further configured to:

[0264] Based on the question text, obtain network cluster data corresponding to at least one network cluster;

[0265] The question text and data from each network cluster are processed based on the target data processing model to generate the answer text.

[0266] Optionally, the generation module 404 is further configured to:

[0267] Obtain a cluster determination prompt word template, and based on the cluster determination prompt word template and the question text, obtain the cluster determination prompt word corresponding to the question text, wherein the cluster determination prompt word template is a formatted template used to enable the data processing model to generate interface call information;

[0268] The cluster determination prompt is input into the target data processing model, and the data acquisition interface call information output by the target data processing model is obtained, wherein the data acquisition interface call information corresponds to at least one network cluster;

[0269] Based on the data acquisition interface call information, obtain the network cluster data corresponding to each network cluster.

[0270] Optionally, the generation module 404 is further configured to:

[0271] Obtain a question answer prompt template, wherein the question answer prompt template is used to enable the target data processing model to output answer text based on the question text;

[0272] Based on the question answer prompt template, the question text, and data from each network cluster, generate question answer prompts;

[0273] Input the question answer prompts into the target data processing model to obtain the answer text generated by the target data processing model.

[0274] Optionally, the generation module 404 is further configured to:

[0275] Based on the question text, obtain the answer step information corresponding to the question text;

[0276] The question text, the answer step information, and the data of each network cluster are input into the target data processing model to obtain the answer text generated by the target data processing model.

[0277] Optionally, the generation module 404 is further configured to:

[0278] Obtain the step-by-step prompt template corresponding to the question text;

[0279] Based on the step-by-step parsing prompt template and the question text, step-by-step parsing prompts are generated;

[0280] The step-by-step prompts are input into the target data processing model to obtain the answer step information generated by the target data processing model.

[0281] Optionally, the network cluster data acquisition device further includes an answer determination module, configured as follows:

[0282] Based on the response text, determine the target answer information corresponding to the question text.

[0283] Optionally, the network cluster data acquisition device further includes a model training module, configured as follows:

[0284] Obtain network cluster data feature information and an initial data processing model, wherein the network cluster data feature information includes at least one network cluster indicator feature information and network cluster question and answer information;

[0285] The initial data processing model is trained based on the network cluster data feature information to obtain the target data processing model.

[0286] Optionally, the model training module is further configured as follows:

[0287] Obtain an initial model training layer and an initial model training layer, wherein the initial model training layer is connected to the self-attention layer in the initial data processing model;

[0288] Adjust the parameters of the initial model training layer based on the network cluster data feature information to obtain the target model training layer;

[0289] By combining the target model training layer and the initial data processing model, the target data processing model is obtained.

[0290] The solution presented in this embodiment significantly improves the automation level of data processing through the design of this network cluster data acquisition device, particularly demonstrating good efficiency in complex multi-dimensional data analysis. The acquisition module accurately acquires the question text and, in conjunction with the generation module, automatically generates the answer text based on a pre-trained target data processing model, avoiding the tedious manual query writing required in traditional methods. Furthermore, by generating cluster-determined prompt word templates and question answer prompt words, the device can efficiently call data interfaces to acquire real-time data from the network cluster and generate detailed analysis steps and results based on this data, ensuring the accuracy and timeliness of the answer text. The confirmation module further reduces user intervention requirements and optimizes the user experience by generating answer information based on the analyzed text. The entire device not only improves the speed of data acquisition and processing but also enhances the generalization ability of the data processing model through adaptive model adjustment, providing support for rapid analysis and automated answering of network cluster data.

[0291] The above is an illustrative scheme of a network cluster data acquisition device according to this embodiment. It should be noted that the technical solution of this network cluster data acquisition device and the technical solution of the network cluster data acquisition method described above belong to the same concept. For details not described in detail in the technical solution of the network cluster data acquisition device, please refer to the description of the technical solution of the network cluster data acquisition method described above.

[0292] Figure 5 shows a structural block diagram of a computing device 500 according to an embodiment of the present disclosure. The components of the computing device 500 include, but are not limited to, a memory 510 and a processor 520. The processor 520 is connected to the memory 510 via a bus 530, and a database 550 is used to store data.

[0293] The computing device 500 also includes an access device 540, which enables the computing device 500 to communicate via one or more networks 560. Examples of these networks include Public Switched Telephone Network (PSTN), Local Area Network (LAN), Wide Area Network (WAN), Personal Area Network (PAN), or combinations of communication networks such as the Internet. The access device 540 may include one or more of any type of wired or wireless network interface (e.g., a network interface controller (NIC)), such as an IEEE 802.11 Wireless Local Area Network (WLAN) wireless interface, a Wi-MAX (Worldwide Interoperability for Microwave Access) interface, an Ethernet interface, a Universal Serial Bus (USB) interface, a cellular network interface, a Bluetooth interface, or a Near Field Communication (NFC) interface.

[0294] In one embodiment of this disclosure, the aforementioned components of the computing device 500, as well as other components not shown in FIG. 5, may also be connected to each other, for example, via a bus. It should be understood that the computing device structural block diagram shown in FIG. 5 is merely for illustrative purposes and is not intended to limit the scope of this disclosure. Those skilled in the art can add or replace other components as needed.

[0295] Computing device 500 can be any type of stationary or mobile computing device, including mobile computers or mobile computing devices (e.g., tablet computers, personal digital assistants, laptop computers, notebook computers, netbooks, etc.), mobile phones (e.g., smartphones), wearable computing devices (e.g., smartwatches, smart glasses, etc.) or other types of mobile devices, or stationary computing devices such as desktop computers or personal computers (PCs). Computing device 500 can also be a mobile or stationary server.

[0296] The processor 520 is configured to execute the following computer-executable instructions, which, when executed by the processor, implement the steps of the above-described network cluster data acquisition method.

[0297] The above is an illustrative scheme of a computing device according to this embodiment. It should be noted that the technical solution of this computing device and the technical solution of the network cluster data acquisition method described above belong to the same concept. For details not described in detail in the technical solution of the computing device, please refer to the description of the technical solution of the network cluster data acquisition method described above.

[0298] An embodiment of this disclosure also provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the above-described network cluster data acquisition method.

[0299] The above is an illustrative scheme of a computer-readable storage medium according to this embodiment. It should be noted that the technical solution of this storage medium belongs to the same concept as the technical solution of the network cluster data acquisition method described above. Details not described in detail in the technical solution of the storage medium can be found in the description of the technical solution of the network cluster data acquisition method described above.

[0300] An embodiment of this disclosure also provides a computer program product, including a computer program / instructions that, when executed by a processor, implement the steps of the above-described network cluster data acquisition method.

[0301] The above is an illustrative example of a computer program according to this embodiment. It should be noted that the technical solution of this computer program and the technical solution of the network cluster data acquisition method described above belong to the same concept. Details not described in detail in the computer program's technical solution can be found in the description of the technical solution of the network cluster data acquisition method described above.

[0302] The foregoing has described specific embodiments of this disclosure. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired results. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0303] The computer instructions include computer program code, which may be in the form of source code, object code, executable file, or certain intermediate forms. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium may be appropriately added or removed according to the requirements of patent practice. For example, in some regions, according to patent practice, computer-readable media may not include electrical carrier signals and telecommunication signals.

[0304] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the embodiments of this disclosure are not limited to the described order of actions, because according to the embodiments of this disclosure, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the embodiments of this disclosure.

[0305] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0306] The preferred embodiments disclosed above are merely illustrative of this disclosure. The optional embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the embodiments of this disclosure. These embodiments are selected and specifically described in this disclosure to better explain the principles and practical applications of the embodiments of this disclosure, thereby enabling those skilled in the art to better understand and utilize this disclosure. This disclosure is limited only by the claims and their full scope and equivalents.

Claims

1. A method for acquiring network cluster data, comprising: Obtain the issue text, wherein the issue text pertains to at least one network cluster; The question text is processed by the target data processing model to generate an answer text, wherein the answer text includes an answer text and a step analysis text. The answer text is generated by the target data processing model based on the step analysis text, and the step analysis text is generated by the target data processing model based on the question text and the network cluster data corresponding to each network cluster.

2. The method as described in claim 1, wherein the question text is processed based on the target data processing model to generate the answer text, includes: Based on the question text, obtain network cluster data corresponding to at least one network cluster; The question text and data from each network cluster are processed based on the target data processing model to generate the answer text.

3. The method as described in claim 2, wherein obtaining network cluster data corresponding to at least one network cluster based on the question text, includes: Obtain a cluster determination prompt word template, and based on the cluster determination prompt word template and the question text, obtain the cluster determination prompt word corresponding to the question text, wherein the cluster determination prompt word template is a formatted template used to enable the data processing model to generate interface call information; The cluster determination prompt is input into the target data processing model, and the data acquisition interface call information output by the target data processing model is obtained, wherein the data acquisition interface call information corresponds to at least one network cluster; Based on the data acquisition interface call information, obtain the network cluster data corresponding to each network cluster.

4. The method of claim 3, wherein, The data acquisition interface call information includes network cluster identifier, time range parameters, and indicator query parameters. The process of obtaining network cluster data corresponding to each network cluster based on the data acquisition interface call information includes: The system calls the indicator data interface and the device information interface to obtain the indicator data of each device within the time range based on the network cluster identifier, and associates the device IP address information to form a structured network cluster data set.

5. The method as described in claim 2, wherein the question text and data from each network cluster are processed based on the target data processing model to generate the answer text, includes: Obtain a question answer prompt template, wherein the question answer prompt template is used to enable the target data processing model to output answer text based on the question text; Based on the question answer prompt template, the question text, and data from each network cluster, generate question answer prompts; Input the question answer prompts into the target data processing model to obtain the answer text generated by the target data processing model.

6. The method of claim 5, wherein, The question-answering prompt template includes step-by-step analysis instructions to guide the target data processing model to perform a multi-stage analysis process, including data extraction, computation, sorting and filtering, and result interpretation.

7. The method as described in claim 2, wherein the question text and data from each network cluster are processed based on the target data processing model to generate the answer text, includes: Based on the question text, obtain the answer step information corresponding to the question text; The question text, the answer step information, and the data of each network cluster are input into the target data processing model to obtain the answer text generated by the target data processing model.

8. The method as described in claim 7, wherein obtaining the answer step information corresponding to the question text based on the question text includes: Obtain the step-by-step prompt template corresponding to the question text; Based on the step-by-step parsing prompt template and the question text, step-by-step parsing prompts are generated; The step-by-step prompts are input into the target data processing model to obtain the answer step information generated by the target data processing model.

9. The method of claim 8, wherein, The step-by-step prompt template includes task decomposition instructions, which guide the target data processing model to parse the question text into structured answer step information containing data source identification, indicator extraction, computational logic, and output format.

10. The method according to any one of claims 1 to 9, wherein after processing the question text based on the target data processing model to generate the answer text, the method further comprises: Based on the response text, determine the target answer information corresponding to the question text.

11. The method of claim 10, wherein, The step of determining the target answer information corresponding to the question text based on the answer text includes: Extract the answer text from the answer text, and convert the answer text into an appropriate display format according to the user role or client type. The display format includes text summary, data table, visualization chart or interactive data dashboard.

12. The method of any one of claims 1 to 11, wherein the target data processing model is trained by the following steps: Obtain network cluster data feature information and an initial data processing model, wherein the network cluster data feature information includes at least one network cluster indicator feature information and network cluster question and answer information; The initial data processing model is trained based on the network cluster data feature information to obtain the target data processing model.

13. The method of claim 12, wherein training the initial data processing model based on the network cluster data feature information to obtain the target data processing model includes: Obtain an initial model training layer, wherein the initial model training layer is connected to the self-attention layer in the initial data processing model; Adjust the parameters of the initial model training layer based on the network cluster data feature information to obtain the target model training layer; By combining the target model training layer and the initial data processing model, the target data processing model is obtained.

14. The method according to any one of claims 1 to 13, wherein obtaining the question text comprises: Receive question request information input by a user, wherein the question request information includes a question text; Input the request question text into the question judgment model and obtain the question judgment result generated by the question judgment model; If the model determines that the issue is related to the network cluster, then the requested issue text is identified as issue text.

15. The method as claimed in any one of claims 1 to 14, wherein, The analysis text of the steps includes the data extraction process, intermediate calculation results, and sorting analysis logic, which is used to provide users with a traceable and verifiable analysis path.

16. The method as claimed in any one of claims 1 to 15, wherein, The target data processing model is a fine-tuned large language model. Its training process includes adjusting the parameters of the model's attention mechanism using network cluster indicator feature information to enhance the understanding of the semantics of network performance indicators.

17. A network cluster data acquisition device, comprising: The acquisition module is configured to acquire issue text, wherein the issue text is for at least one network cluster; The generation module is configured to process the question text based on the target data processing model and generate an answer text, wherein the answer text includes an answer text and a step analysis text. The answer text is generated by the target data processing model based on the step analysis text, and the step analysis text is generated by the target data processing model based on the question text and the network cluster data corresponding to each network cluster.

18. A computing device, comprising: Memory and processor; The memory is used to store computer programs / instructions, and the processor is used to execute the computer programs / instructions, which, when executed by the processor, implement the steps of the method according to any one of claims 1-16.

19. A computer-readable storage medium storing a computer program / instructions that, when executed by a processor, implement the steps of the method according to any one of claims 1-16.

20. A computer program product comprising a computer program / instructions that, when executed by a processor, implement the steps of the method according to any one of claims 1-16.

Images