Intelligent calculation center task quick response method, server, medium and product

Abstract

The invention discloses an intelligent calculation center task quick response method, a server, a medium and a product, and the method comprises the steps: extracting key information in a task request, matching the key information with an emergency keyword library, and determining whether the task request is an emergency task or not, so as to quickly recognize and classify the emergency task. The task type, the emergency degree and the potential influence of the task request are comprehensively scored, the importance index is generated, and the burst task is accurately quantified. According to the importance index, the proprietary computing resource is allocated to the primary processing object, so that the most urgent and most important task request is ensured to be processed preferentially. According to the method, the response speed and the processing efficiency of the intelligent calculation center to emergencies are remarkably improved, manual intervention is reduced, the accuracy of resource allocation is enhanced, intelligent and automatic management of resources of the intelligent calculation center is achieved, and the emergency processing capacity is greatly improved.

Description

Technical Field

[0001] The present application relates to the field of information technology, and in particular to a method, server, medium and product for rapid response to tasks in an intelligent computing center. Background Art

[0002] With the rapid development of artificial intelligence and big data technologies, intelligent computing centers are becoming increasingly important in handling complex computing tasks. Intelligent computing centers need to efficiently manage and allocate computing resources to cope with the growing number of task requests, especially in emergencies, where being able to quickly respond to task requirements is particularly critical.

[0003] In the prior art, computing resources are usually managed through preset rules and manual intervention. These methods rely on administrators to classify and sort tasks based on experience and then allocate corresponding computing resources.

[0004] However, as the complexity and urgency of task requests increase, existing technologies have shown some shortcomings in responding to emergencies. Since the analysis of task requests and resource allocation mainly rely on manual judgment, it may lead to slow response speed and inability to adjust resources in time to meet the needs of emergency tasks. Summary of the invention

[0005] The present application provides a method, server, medium and product for rapid response of intelligent computing center tasks, which are used to improve the response speed and processing efficiency of the intelligent computing center to emergencies, reduce manual intervention, enhance the accuracy of resource allocation, realize intelligent and automated management of intelligent computing center resources, and greatly improve the ability to handle emergencies.

[0006] In the first aspect, the present application provides a method for rapid response of tasks in an intelligent computing center, which is applied to a server, and the method includes: analyzing the acquired task requests based on natural language processing technology to extract key information; matching the key information with a preset emergency event keyword library to determine whether the task request is a sudden task; if so, scoring the task request according to the task type, urgency and potential impact to generate an importance index; allocating a dedicated computing node to the primary processing object according to the importance index, and reserving dedicated computing resources for the primary processing object at the dedicated computing node, and the primary processing object is used to represent the task request with the highest importance index; after determining that the dedicated computing resources match the primary processing object, transmitting the primary processing object to the dedicated computing node.

[0007] By adopting the above technical solution, the server extracts key information from the task request through natural language processing technology, and then matches the key information with the emergency keyword library to determine whether the task request is an emergency task, so as to quickly identify and classify emergency tasks. By comprehensively scoring the task type, urgency and potential impact of the task request, an importance index is generated, which realizes the accurate quantification of emergency tasks. According to the importance index, dedicated computing resources are allocated to the primary processing object to ensure that the most urgent and important task requests are given priority. This method significantly improves the response speed and processing efficiency of the intelligent computing center to emergencies, reduces manual intervention, and enhances the accuracy of resource allocation. At the same time, through the allocation of dedicated computing resources, the situation of high-priority tasks competing for resources with ordinary tasks is avoided, and the smooth execution of key tasks is guaranteed. Overall, this method realizes the intelligent and automated management of the resources of the intelligent computing center and greatly improves the emergency handling capability.

[0008] In combination with some embodiments of the first aspect, in some embodiments, the key information is matched with a preset emergency event keyword library to determine whether the task request is a sudden task, specifically including: performing semantic analysis on the key information to extract core semantic features; calculating the similarity between the core semantic features and the semantic features in the emergency event keyword library; when the similarity exceeds a preset first numerical threshold, determining that the task request is a sudden task; when the similarity does not exceed the preset first numerical threshold, performing context analysis on the key information to extract context semantic information; matching the context semantic information with the context information in the emergency event keyword library; when the matching degree exceeds a preset second numerical threshold, determining that the task request is a sudden task.

[0009] By adopting the above technical solution, the server significantly improves the accuracy and comprehensiveness of emergency task identification through multi-level semantic analysis and matching strategies. First, semantic analysis is performed on key information and core semantic features are extracted, which are preliminarily matched with the semantic features in the emergency keyword library, so that clear emergency tasks can be quickly identified. For cases where direct matching fails, further context analysis is performed to extract richer context semantic information for secondary matching, which can identify emergency tasks with implicit or complex expressions. By setting two different matching thresholds, the sensitivity of identification is taken into account while ensuring the accuracy of identification. This hierarchical and progressive matching method greatly reduces the possibility of misjudgment and missed judgment, and improves the coverage of emergency task identification. At the same time, this method is also highly scalable and adaptable, and can adapt to the emergency identification needs of different fields and levels by adjusting the emergency keyword library and matching threshold.

[0010] In combination with some embodiments of the first aspect, in some embodiments, if so, a score is given according to the task type, urgency and potential impact of the task request to generate an importance index, specifically including: determining the task type weight according to a preset task type weight table and the task type of the task request; extracting time words and degree words indicating the urgency from the task request, and determining the urgency score according to preset urgency scoring rules; determining the potential impact score by analyzing the scope of the field and affected objects involved in the task request; and performing weighted summation of the task type weight, the urgency score and the potential impact score to obtain the importance index.

[0011] By adopting the above technical solution, the server achieves accurate quantification of the importance of burst tasks through multi-dimensional scoring and weighted summation. First, the server determines the task type weight according to the preset task type weight table, which reflects the inherent importance differences of different types of task requests. Secondly, the server evaluates the urgency by extracting time words and degree words, which can accurately capture the time sensitivity of task requests. Thirdly, the server evaluates the potential impact by analyzing the scope of the field and the affected objects involved in the task request, which reflects the potential influence of the task request. Finally, the scores of these three dimensions are weighted and summed to obtain a comprehensive importance index. This method not only takes into account multiple important features of the task request, but also realizes the flexible adjustment of each factor through the weighting mechanism, and can adjust the weight of each factor according to actual needs. This refined scoring mechanism provides a reliable basis for subsequent resource allocation, ensures the rationality and accuracy of resource allocation, and thus greatly improves the efficiency and quality of the intelligent computing center in processing burst tasks.

[0012] In combination with some embodiments of the first aspect, in some embodiments, a dedicated computing node is allocated to the primary processing object according to the importance index, and dedicated computing resources are reserved for the primary processing object on the dedicated computing node, and the primary processing object is used to represent the task request with the highest importance index, specifically including: sorting multiple burst tasks from high to low according to the importance index, and selecting the burst task with the highest importance index as the primary processing object; determining the computing resource demand characteristics of the primary processing object based on a machine learning model; selecting, from a pool of available computing nodes, a computing node with the highest match to the computing resource demand characteristics as a dedicated computing node; on the dedicated computing node, dynamically adjusting and reserving corresponding computing resources as dedicated computing resources according to the computing resource demand characteristics.

[0013] By adopting the above technical solution, efficient processing of burst tasks is achieved through a multi-step intelligent resource allocation process. First, the server sorts burst tasks according to the importance index to ensure that the most critical burst tasks are given priority. Secondly, the server uses a machine learning model to analyze the computing resource demand characteristics of the primary processing object. This data-driven method can more accurately predict resource requirements than traditional fixed rules. Then, the server selects the most suitable dedicated computing node through matching analysis to improve resource utilization efficiency. Finally, dynamically adjusting and reserving resources on the selected dedicated computing node ensures the smooth execution of burst tasks and avoids resource waste. This method not only improves the accuracy of resource allocation, but also enhances the flexibility and adaptability of the method. Through intelligent resource scheduling, the ability of the intelligent computing center to handle burst tasks has been greatly improved, while optimizing the overall resource utilization, providing strong technical support for handling complex and changeable emergencies.

[0014] In combination with some embodiments of the first aspect, in some embodiments, after the step of transmitting the primary processing object to the dedicated computing node after determining that the dedicated computing resource matches the primary processing object, the method also includes: monitoring the resource utilization of the dedicated computing node; when it is detected that the resource utilization is lower than a preset third numerical threshold, selecting the next burst task from the importance index sorting queue, and allocating the remaining dedicated computing resources for parallel processing, the importance index sorting queue is used to represent the result of sorting multiple burst tasks from high to low according to the importance index.

[0015] By adopting the above technical solutions, the resource utilization efficiency of the intelligent computing center is significantly improved through real-time monitoring and dynamic adjustment mechanisms. By continuously monitoring the resource utilization of dedicated computing nodes, the server can promptly detect idle resources. When it is detected that the resource utilization is lower than the preset third numerical threshold, the server will automatically select the next burst task from the importance index sorting queue and allocate the remaining resources for parallel processing. This mechanism effectively solves the problem of resource waste caused by the traditional fixed resource allocation method, improves the overall computing efficiency, and ensures that while processing the highest priority burst tasks, other important burst tasks can also be processed in a timely manner, thereby improving throughput. This dynamic scheduling strategy not only maximizes resource utilization, but also shortens the average waiting time of burst tasks, thereby improving the overall performance and response speed of the intelligent computing center.

[0016] In combination with some embodiments of the first aspect, in some embodiments, after the step of transmitting the primary processing object to the dedicated computing node after determining that the dedicated computing resource matches the primary processing object, the method also includes: monitoring the task execution progress of the dedicated computing node in real time; when it is detected that the task execution progress lags behind the preset progress, migrating the primary processing object to a spare high-performance computing node.

[0017] By adopting the above technical solutions, the reliability and efficiency of the intelligent computing center in handling sudden tasks are greatly improved through real-time monitoring and task migration mechanisms. By continuously monitoring the task execution progress of the dedicated computing node, the server can promptly detect abnormalities or delays in task execution. When it is detected that the task execution progress lags behind the preset progress, the server will automatically migrate the primary processing object to the backup high-performance computing node. This mechanism effectively prevents task execution delays or failures due to insufficient performance or failure of a single node, ensuring the timely completion of critical tasks. At the same time, the method also provides a task execution guarantee mechanism, which enhances fault tolerance and stability. This dynamic adjustment strategy not only ensures the execution quality of high-priority sudden tasks, but also improves the reliability and response speed of the entire intelligent computing center, especially when dealing with complex and time-consuming sudden tasks. It shows significant advantages.

[0018] In combination with some embodiments of the first aspect, in some embodiments, after the step of transmitting the primary processing object to the dedicated computing node after determining that the dedicated computing resource matches the primary processing object, the method also includes: recording the dedicated computing node and dedicated computing resources of the primary processing object, the computing resources consumed by the primary processing object, the task execution time of the primary processing object, and the task execution status of the primary processing object.

[0019] By adopting the above technical solutions and recording various indicators of task execution in detail, valuable data support is provided for the long-term optimization and management of the intelligent computing center. Recording the dedicated computing nodes and dedicated computing resources of the primary processing object helps to analyze the rationality and efficiency of resource allocation. Recording the computing resources consumed by the primary processing object and the task execution time can be used to evaluate the actual resource requirements and execution efficiency of the primary processing object, and provide a basis for resource estimation for similar tasks in the future. Recording the task execution status can be used to analyze the quality of task completion. This method can significantly improve the overall operating efficiency of the intelligent computing center, optimize the resource allocation strategy, and provide data support for the continuous improvement of emergency response capabilities.

[0020] In a second aspect, an embodiment of the present application provides a server, comprising: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the server to execute the method described in the first aspect and any possible implementation method of the first aspect.

[0021] In a third aspect, an embodiment of the present application provides a computer program product comprising instructions, which, when executed on a server, enables the server to execute the method described in the first aspect and any possible implementation of the first aspect.

[0022] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, comprising instructions, which, when executed on a server, enable the server to execute the method described in the first aspect and any possible implementation of the first aspect.

[0023] It is understandable that the server provided in the second aspect, the computer program product provided in the third aspect, and the computer storage medium provided in the fourth aspect are all used to execute the method provided in the embodiment of the present application. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the corresponding method, which will not be repeated here.

[0024] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. Due to the adoption of an intelligent task identification and resource allocation mechanism based on natural language processing and machine learning, the present invention can quickly and accurately identify emergency tasks and prioritize them, effectively solving the problems of low efficiency and easy errors in manual judgment in the prior art, thereby realizing the rapid response and efficient processing of emergencies by the intelligent computing center.

[0025] 2. Due to the adoption of multi-dimensional scoring and dynamic resource scheduling strategies, the present invention can accurately quantify the importance of tasks and achieve optimal allocation of resources, effectively solving the problems of rigid resource scheduling and inability to flexibly respond to emergencies in the prior art, thereby achieving a significant improvement in the resource utilization of the intelligent computing center and a substantial improvement in task processing efficiency.

[0026] 3. Due to the adoption of comprehensive task execution records and continuous optimization mechanisms, the present invention can continuously accumulate experience data and optimize server performance, effectively solving the problems of lack of long-term optimization capabilities and difficulty in responding to new emergencies in the prior art, thereby achieving continuous improvement of the performance of the intelligent computing center and continuous enhancement of its response capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1This is a flowchart of a method for rapid response of intelligent computing center tasks in an embodiment of the present application; Figure 2 This is another flowchart of the method for rapid response of intelligent computing center tasks in an embodiment of the present application; Figure 3 It is a schematic diagram of a physical device structure of a server in an embodiment of the present application. DETAILED DESCRIPTION

[0028] The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to be used as limitations to the present application. As used in the specification of the present application, the singular expressions "one", "a kind of", "above", "the" and "this" are intended to also include plural expressions, unless there is a clear indication to the contrary in the context. It should also be understood that the term "and / or" used in the present application refers to any or all possible combinations comprising one or more of the listed items.

[0029] In the following, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as suggesting or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features, and in the description of the embodiments of the present application, unless otherwise specified, "plurality" means two or more.

[0030] The following is a description of the process of the method provided by this implementation. Figure 1 , which is a flow chart of the method for rapid response of intelligent computing center tasks in an embodiment of the present application.

[0031] S101. Analyze the acquired task request based on natural language processing technology to extract key information; Among them, natural language processing technology refers to computer technology used to process and analyze human language, including semantic analysis, grammatical analysis, entity recognition, etc. Task request refers to the specific work content submitted by the user or system that needs to be processed by the intelligent computing center. Key information refers to the core content extracted from the task request, including important features such as task type, urgency, and scope of impact, which are not limited here.

[0032] Specifically, when the server receives a task request from a user or system, it immediately performs word segmentation and part-of-speech tagging on the task request through natural language processing technology, and then uses a pre-trained semantic model to deeply understand the text of the task request. Next, the server extracts key entities such as task type, time requirements, and objects involved through entity recognition technology. At the same time, the server uses dependency syntactic analysis technology to capture the relationship between entities in order to more accurately understand the overall context of the task request. Finally, the server structures all the extracted key information to form a data structure containing the core features of the task, in preparation for subsequent task classification and priority evaluation.

[0033] S102, matching the key information with a preset emergency keyword library to determine whether the task request is an emergency task; The preset emergency keyword library refers to a collection of predefined and stored words, phrases or semantic patterns related to various emergency events. Matching refers to comparing key information with the content in the keyword library to find similar or related items. Emergency tasks refer to tasks that need to be handled urgently, are time-sensitive or have important impacts.

[0034] Specifically, after the server completes the extraction of key information, it will immediately start the matching process. First, the server will convert the extracted key information into a standardized representation, such as a vector or a semantic graph. Then, it will use efficient index structures (such as inverted indexes or semantic hashes) to quickly find relevant items in the emergency keyword library. The matching process not only considers exact matching, but also uses fuzzy matching algorithms to handle synonyms, near-synonyms, and so on. At the same time, the server will consider the context and semantic relationship of the words and use deep learning models (such as BERT or Word2Vec) to calculate semantic similarity. If the semantic similarity exceeds the preset threshold, the task request is marked as a burst task. In addition, the server will preliminarily classify the burst tasks based on the matching results to provide a basis for subsequent priority evaluation. The entire matching process is usually completed in milliseconds, ensuring a rapid response to emergencies.

[0035] Optionally, under normal circumstances, matching the key information with a preset emergency keyword library to determine whether the task request is a sudden task can be achieved in the following way: performing semantic analysis on the key information and extracting core semantic features; calculating the similarity between the core semantic features and the semantic features in the emergency keyword library; when the similarity exceeds a preset first numerical threshold, determining that the task request is a sudden task; when the similarity does not exceed the preset first numerical threshold, performing context analysis on the key information and extracting context semantic information; matching the context semantic information with the context information in the emergency keyword library; when the matching degree exceeds a preset second numerical threshold, determining that the task request is a sudden task.

[0036] S103, if yes, then score the task request according to its task type, urgency and potential impact to generate an importance index; Among them, the task type indicates the nature and field of the task, such as natural disasters, public health incidents, social security incidents, etc. The urgency refers to the time urgency of the task to be processed. The potential impact refers to the degree of social, economic or environmental effects that the task may have. The importance index is a value obtained by comprehensively considering multiple factors, which is used to quantify the overall importance of the task request.

[0037] Specifically, when the server determines that a task request is an emergency task, it will immediately start the importance index evaluation process. The server can assign different scores to the task request according to the task type, urgency and potential impact, and finally perform weighted summation to obtain the importance index. For example, a task request involving life safety may receive a higher basic score. Then, the server will analyze the time words and degree words in the task description, such as "immediately" and "right away", and evaluate the urgency score accordingly. For potential impact, the server will consider factors such as the geographical scope, population size, and economic losses involved in the task, and calculate the impact degree score through a pre-trained regression model. Finally, the server combines the scores of these three dimensions through weighted summation to generate the final importance index. In this process, the weights of each dimension can be dynamically adjusted according to actual needs. The importance index is usually standardized to a range of 0-100 to facilitate subsequent task priority sorting and resource allocation. The entire scoring process is highly automated and can usually be completed within seconds, providing key support for rapid response to emergency tasks.

[0038] Optionally, in general, if yes, then the task request is scored according to its task type, urgency and potential impact, and the importance index can be generated in the following ways: determine the task type weight according to a preset task type weight table and the task type of the task request; extract time words and degree words indicating the urgency from the task request, and determine the urgency score according to preset urgency scoring rules; determine the potential impact score by analyzing the scope of the field and affected objects involved in the task request; and perform weighted summation of the task type weight, urgency score and potential impact score to obtain the importance index.

[0039] S104, allocating a dedicated computing node to the primary processing object according to the importance index, and reserving dedicated computing resources for the primary processing object on the dedicated computing node, the primary processing object being used to represent the task request with the highest importance index; Among them, the primary processing object refers to the task request with the highest importance index, which represents the most urgent task that needs to be processed first. Dedicated computing nodes refer to computing servers or virtual machine instances that are specially allocated for specific tasks. Dedicated computing resources refer to computing resources such as CPU, memory, and storage reserved for the primary processing object. Reservation means reserving a certain amount of resources on the dedicated computing node to ensure that these resources are dedicated to the primary processing task.

[0040] Specifically, after the server completes the calculation of the importance index, it immediately enters the resource allocation stage. First, the server selects the burst task with the highest importance index from all burst tasks as the primary processing object. Then, the server scans the list of currently available computing nodes, considers factors such as node performance, load status, network connection, etc., and selects the most suitable node as the dedicated computing node. A complex scoring algorithm is used in the selection process, which comprehensively considers indicators such as the node's CPU performance, memory capacity, I / O speed, and the degree of matching with the characteristics of the primary processing object. After determining the dedicated computing node, the server immediately starts the resource reservation process on the dedicated computing node. The number of reserved resources is dynamically calculated based on the estimated demand of the primary processing object, and usually a slightly higher amount of resources than the estimate is reserved to cope with possible demand fluctuations. The reservation process includes operations such as setting CPU priority, memory locking, and disk space allocation. At the same time, the server updates the resource management database and records the reservation status to ensure that these resources are not occupied by other tasks. The entire allocation and reservation process can usually be completed within a few seconds, providing an exclusive high-performance computing environment for the primary processing object.

[0041] Optionally, in general, a dedicated computing node is allocated to the primary processing object according to the importance index, and dedicated computing resources are reserved for the primary processing object on the dedicated computing node. The primary processing object is used to represent the task request with the highest importance index, which can be achieved in the following manner: multiple burst tasks are sorted from high to low according to the importance index, and the burst task with the highest importance index is selected as the primary processing object; Based on the machine learning model, the computing resource demand characteristics of the primary processing object are determined; from the available computing node pool, the computing node with the highest match with the computing resource demand characteristics is selected as the dedicated computing node; on the dedicated computing node, according to the computing resource demand characteristics, the corresponding computing resources are dynamically adjusted and reserved as dedicated computing resources.

[0042] S105: After determining that the dedicated computing resource matches the primary processing object, transmit the primary processing object to the dedicated computing node.

[0043] Among them, determining whether the dedicated computing resources match the primary processing object means verifying whether the reserved computing resources meet the requirements of the primary processing object. Specifically, after completing the resource reservation, the server will immediately start the resource matching verification process. First, the server will analyze the resource demand characteristics of the primary processing object in detail, including CPU usage mode, memory access characteristics, I / O intensity, etc. Then, these resource demand characteristics are compared with the reserved dedicated computing resources to ensure that each resource can meet or exceed the resource requirements of the primary processing object of the task. If a resource is found to be insufficient, the server will immediately adjust the reservation strategy, which may increase the resource reservation amount or switch to a more suitable computing node. After confirming the resource match, the server starts the task transmission process. The transmission process will first establish a secure data channel, and SSL / TLS encryption may be used to ensure the security of data transmission. Then, the server will package all data related to the primary processing object, including input data, processing scripts, configuration files, etc., and use efficient transmission protocols (such as RDMA) to quickly transmit the data to the dedicated computing node. At the same time, the server will prepare the execution environment on the dedicated computing node, including setting the necessary system parameters, loading the required software libraries, etc. After the transmission is complete, the server will perform an integrity check to ensure that all necessary components have been correctly transmitted. Finally, the server will send a task start instruction to the dedicated computing node to start the formal task processing process. The entire transmission and preparation process can usually be completed within minutes, ensuring that burst tasks can quickly enter the execution phase.

[0044] By adopting the above technical solution, the server extracts key information from the task request through natural language processing technology, and then matches the key information with the emergency keyword library to determine whether the task request is an emergency task, so as to quickly identify and classify emergency tasks. By comprehensively scoring the task type, urgency and potential impact of the task request, an importance index is generated, which realizes the accurate quantification of emergency tasks. According to the importance index, dedicated computing resources are allocated to the primary processing object to ensure that the most urgent and important task requests are given priority. This method significantly improves the response speed and processing efficiency of the intelligent computing center to emergencies, reduces manual intervention, and enhances the accuracy of resource allocation. At the same time, through the allocation of dedicated computing resources, the situation of high-priority tasks competing for resources with ordinary tasks is avoided, and the smooth execution of key tasks is guaranteed. Overall, this method realizes the intelligent and automated management of the resources of the intelligent computing center and greatly improves the emergency handling capability.

[0045] The following is a more detailed description of the process of the method provided by this implementation. Figure 2 , which is another flow chart of the method for rapid response of intelligent computing center tasks in an embodiment of the present application.

[0046] S201, analyzing the acquired task request based on natural language processing technology to extract key information; For details, please refer to step S101, which will not be described in detail here.

[0047] S202, matching the key information with a preset emergency keyword library to determine whether the task request is an emergency task; For details, please refer to step S102, which will not be described in detail here.

[0048] S203, if yes, then score the task request according to its task type, urgency and potential impact to generate an importance index; For details, please refer to step S103, which will not be described in detail here.

[0049] S204, allocating a dedicated computing node to the primary processing object according to the importance index, and reserving dedicated computing resources for the primary processing object on the dedicated computing node, the primary processing object being used to represent the task request with the highest importance index; For details, please refer to step S104, which will not be described in detail here.

[0050] S205: After determining that the dedicated computing resource matches the primary processing object, transmit the primary processing object to the dedicated computing node.

[0051] For details, please refer to step S105, which will not be described in detail here.

[0052] S206, monitoring the resource utilization of the dedicated computing node; Resource utilization refers to the degree of use of computing resources (such as CPU, memory, storage, network bandwidth, etc.), usually expressed as a percentage. A dedicated computing node refers to an independent computing unit assigned to a specific task, which can be a physical server or a virtual machine instance.

[0053] Specifically, after the server transfers the primary processing object to the dedicated computing node and starts task execution, it immediately monitors resource utilization. The server collects various resource usage indicators of the dedicated computing node at a high frequency (usually once every second or every few seconds). The monitored resource usage indicators include but are not limited to: CPU usage, memory occupancy, disk I / O rate, network traffic, GPU utilization (if applicable), etc. The server uses a lightweight monitoring agent to minimize the impact on the performance of the monitored nodes. The server uses statistical methods to smooth the data to eliminate the impact of short-term fluctuations and obtain a more stable resource utilization trend. At the same time, the server also calculates the moving average and standard deviation of each resource to detect anomalies. These processed data will be presented in a visual form, which is convenient for administrators to grasp resource usage in real time.

[0054] S207, when it is detected that the resource utilization is lower than a preset third numerical threshold, selecting the next burst task from the importance index sorting queue, and allocating the remaining dedicated computing resources for parallel processing, the importance index sorting queue being used to represent the result of sorting the multiple burst tasks from high to low according to the importance index; The preset third numerical threshold refers to a predefined resource utilization standard used to determine whether resources are fully utilized. The importance index sorting queue represents a list of burst tasks to be processed arranged from high to low by importance index. Remaining resources refer to dedicated computing resources that are currently not fully utilized. Parallel processing refers to processing multiple burst tasks simultaneously to improve overall resource utilization efficiency.

[0055] Specifically, in the process of continuously monitoring resource utilization, the server will continuously compare the current resource utilization with the preset third numerical threshold. The preset third numerical threshold is usually set between 60% and 80%, and can also be set according to actual conditions, which is not limited here. When the server detects that one or more resource utilizations are continuously lower than the preset third numerical threshold (for example, for more than 5 consecutive minutes), the resource optimization process will be triggered. First, the server will accurately calculate the amount of available remaining resources, including the number of idle CPU cores, the size of available memory, the remaining storage space, etc. Then, the server will access the importance index sorting queue, which is a dynamically updated data structure that contains all pending burst tasks in descending order of importance index. The server will start from the top of the importance index sorting queue and select the first burst task whose resource demand can be met by the remaining resources. After selecting the burst task, the server will immediately start the resource allocation process and allocate the remaining resources to the burst task. This process includes creating a new virtual machine instance or container, setting resource limits, preparing the execution environment, etc. At the same time, the server will update the resource management database to record the new resource allocation status. If there are still significant remaining resources after allocation, the server may continue to select the next appropriate burst task to achieve parallel processing of multiple burst tasks. The whole process is automated and can usually be completed within a few seconds to a few minutes, greatly improving resource utilization efficiency.

[0056] S208, monitoring the task execution progress of the dedicated computing node in real time; Specifically, after the task starts executing, the server will periodically (usually every minute or less) check the completion status of the task. Progress checking may be implemented in a variety of ways, such as parsing task output logs, checking the completion status of key checkpoints, analyzing intermediate result files, etc.

[0057] S209, when it is detected that the execution progress of the task lags behind the preset progress, migrating the primary processing object to the standby high-performance computing node; The server compares the actual task execution progress with the preset progress. The preset progress is usually pre-determined based on historical data and task characteristics. If the server finds that the actual progress continues to lag behind the task execution progress and the preset progress (for example, it lags behind by more than 20% for three consecutive checks), the task migration process will be triggered. First, the server selects a new node that is most suitable from the pre-prepared high-performance node pool as the migration target. The selection criteria include the current load of the node, hardware specifications, network connection quality, etc. Then, the server starts the migration preparation process, including preparing the execution environment on the new node and reserving necessary resources. Next, the server uses real-time migration technology (such as memory page transfer) to transfer the execution status of the task from the original dedicated computing node to the new node. This process minimizes the task interruption time and can usually be completed within a few seconds to a few minutes. After the migration is completed, the server resumes the task execution on the new node and updates the relevant information in the task management database. At the same time, the resources on the original dedicated computing node will be released and can be used to process other tasks. The entire migration process is automated, aiming to minimize task delays caused by performance issues and ensure that high-priority tasks can be completed on time.

[0058] S210, recording the dedicated computing nodes and dedicated computing resources of the primary processing object, the computing resources consumed by the primary processing object, the task execution time of the primary processing object, and the task execution status of the primary processing object.

[0059] The server provides valuable data support for the long-term optimization and management of the intelligent computing center by recording various indicators of task execution in detail. Recording the dedicated computing nodes and dedicated computing resources of the primary processing object helps to analyze the rationality and efficiency of resource allocation. Recording the computing resources consumed by the primary processing object and the task execution time can be used to evaluate the actual resource requirements and execution efficiency of the primary processing object, and provide a basis for resource estimation for similar tasks in the future. Recording the task execution status can be used to analyze the quality of task completion. This method can significantly improve the overall operating efficiency of the intelligent computing center, optimize resource allocation strategies, and provide data support for the continuous improvement of emergency response capabilities.

[0060] In implementing the present invention, since an intelligent task identification and resource allocation mechanism based on natural language processing and machine learning is adopted, the present invention can quickly and accurately identify emergency tasks and prioritize them, effectively solving the problem of low efficiency and error-prone manual judgment in the prior art, thereby realizing the rapid response and efficient processing of emergencies by the intelligent computing center. Since a multi-dimensional scoring and dynamic resource scheduling strategy are adopted, the present invention can accurately quantify the importance of tasks and realize the optimal allocation of resources, effectively solving the problem of rigid resource scheduling and inability to flexibly respond to emergencies in the prior art, thereby realizing a significant improvement in the resource utilization rate of the intelligent computing center and a substantial improvement in task processing efficiency. Since a comprehensive task execution record and continuous optimization mechanism are adopted, the present invention can continuously accumulate empirical data and optimize server performance, effectively solving the problem of lack of long-term optimization capabilities and difficulty in responding to new emergencies in the prior art, thereby realizing the continuous improvement of the performance of the intelligent computing center and the continuous enhancement of its response capabilities.

[0061] The following describes the server in the embodiment of the present invention from the perspective of hardware processing. Figure 3 , is a schematic diagram of a physical device structure of a server in an embodiment of the present application.

[0062] It should be noted that Figure 3 The structure of the server shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0063] like Figure 3 As shown, the server includes a central processing unit (CPU) 301, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 302 or the program loaded from the storage part 308 to the random access memory (RAM) 303, such as executing the method described in the above embodiment. In RAM 303, various programs and data required for system operation are also stored. CPU 301, ROM 302 and RAM 303 are connected to each other through bus 304. Input / output (I / O) interface 305 is also connected to bus 304.

[0064] The following components are connected to the I / O interface 305: an input section 306 including an audio input device, a button switch, etc.; an output section 307 including a liquid crystal display (LCD) and an audio output device, an indicator light, etc.; a storage section 308 including a hard disk, etc.; and a communication section 309 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to the I / O interface 305 as needed. A removable medium 311, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 310 as needed so that a computer program read therefrom is installed into the storage section 308 as needed.

[0065] In particular, according to an embodiment of the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through the communication part 309, and / or installed from a removable medium 311. When the computer program is executed by the central processing unit (CPU) 301, various functions defined in the present invention are performed.

[0066] It should be noted that specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, apparatus, or device.

[0067] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. Each box in the flowchart or block diagram may represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box may also occur in an order different from that marked in the accompanying drawings.

[0068] Specifically, the server of this embodiment includes a processor and a memory. The memory stores a computer program. When the computer program is executed by the processor, the method for rapid response to tasks of the intelligent computing center provided in the above embodiment is implemented.

[0069] As another aspect, the present invention further provides a computer-readable storage medium, which may be included in the server described in the above embodiment; or may exist independently without being assembled into the server. The above storage medium carries one or more computer programs, and when the above one or more computer programs are executed by a processor of the server, the server implements the intelligent computing center task rapid response method provided in the above embodiment.

[0070] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

[0071] As used in the above embodiments, the term "when..." may be interpreted to mean "if..." or "after..." or "in response to determining..." or "in response to detecting...", depending on the context. Similarly, the phrases "upon determining..." or "if (the stated condition or event) is detected" may be interpreted to mean "if determining..." or "in response to determining..." or "upon detecting (the stated condition or event)" or "in response to detecting (the stated condition or event)", depending on the context.

[0072] Those skilled in the art can understand that to implement all or part of the processes in the above-mentioned embodiments, the processes can be completed by computer programs to instruct related hardware, and the programs can be stored in computer-readable storage media. When the programs are executed, they can include the processes of the above-mentioned method embodiments. The aforementioned storage media include: ROM or random access memory RAM, magnetic disk or optical disk and other media that can store program codes.

Claims

1. A method for rapid response of intelligent computing center tasks, characterized in that: Applied to a server, the method comprises: Analyze the obtained task requests based on natural language processing technology to extract key information; Matching the key information with a preset emergency keyword library to determine whether the task request is an emergency task; If yes, then score the task request according to its task type, urgency and potential impact to generate an importance index; Allocating a dedicated computing node to a primary processing object according to the importance index, and reserving dedicated computing resources for the primary processing object on the dedicated computing node, wherein the primary processing object is used to represent the task request with the highest importance index; After determining that the dedicated computing resource matches the primary processing object, the primary processing object is transmitted to the dedicated computing node.

2. The method according to claim 1, characterized in that The step of matching the key information with a preset emergency keyword library to determine whether the task request is an emergency task specifically includes: Performing semantic analysis on the key information to extract core semantic features; Calculate the similarity between the core semantic features and the semantic features in the emergency keyword library; When the similarity exceeds a preset first numerical threshold, determining that the task request is a burst task; When the similarity does not exceed the preset first numerical threshold, performing context analysis on the key information to extract context semantic information; Matching the context semantic information with the context information in the emergency keyword library; When the matching degree exceeds a preset second numerical threshold, the task request is determined to be a burst task.

3. The method according to claim 1, characterized in that If so, the task request is scored according to its task type, urgency and potential impact to generate an importance index, including: Determining a task type weight according to a preset task type weight table and the task type of the task request; Extracting time words and degree words indicating urgency from the task request, and determining an urgency score according to a preset urgency score rule; Determine the potential impact score based on the analysis of the scope of the task request and the objects affected; The task type weight, the urgency score and the potential impact score are weighted and summed to obtain the importance index.

4. The method according to claim 1, characterized in that: The allocating a dedicated computing node to the primary processing object according to the importance index, and reserving dedicated computing resources for the primary processing object on the dedicated computing node, wherein the primary processing object is used to represent the task request with the highest importance index, specifically includes: sorting the multiple emergency tasks from high to low according to the importance index, and selecting the emergency task with the highest importance index as the primary processing object; Determining the computing resource demand characteristics of the primary processing object based on the machine learning model; Selecting a computing node with the highest matching degree with the computing resource requirement characteristics from a pool of available computing nodes as a dedicated computing node; On the dedicated computing node, according to the computing resource demand characteristics, the corresponding computing resources are dynamically adjusted and reserved as dedicated computing resources.

5. The method according to claim 1, characterized in that After the step of transmitting the primary processing object to the dedicated computing node after determining that the dedicated computing resource matches the primary processing object, the method further includes: Monitoring resource utilization of the dedicated computing nodes; When it is detected that the resource utilization is lower than a preset third numerical threshold, the next burst task is selected from the importance index sorting queue and the remaining computing dedicated resources are allocated for parallel processing. The importance index sorting queue is used to represent the result of sorting multiple burst tasks from high to low according to the importance index.

6. The method according to claim 1, characterized in that After the step of transmitting the primary processing object to the dedicated computing node after determining that the dedicated computing resource matches the primary processing object, the method further includes: Monitoring the task execution progress of the dedicated computing node in real time; When it is detected that the task execution progress lags behind the preset progress, the primary processing object is migrated to the standby high-performance computing node.

7. The method according to claim 1, characterized in that After the step of transmitting the primary processing object to the dedicated computing node after determining that the dedicated computing resource matches the primary processing object, the method further includes: The dedicated computing nodes and dedicated computing resources of the primary processing object, the computing resources consumed by the primary processing object, the task execution time of the primary processing object, and the task execution status of the primary processing object are recorded.

8. A server, characterized in that: The server includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the server to execute the method described in any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that: When the instructions are executed on a server, the server is caused to execute the method according to any one of claims 1 to 7.

10. A computer program product, characterized in that When the computer program product is run on a server, the server is caused to execute the method according to any one of claims 1 to 7.

Images