A server resource configuration method based on mobile edge computing

Abstract

The application relates to the technical field of resource configuration, and particularly discloses a server resource configuration method based on mobile edge computing, which comprises the following steps: generating a server node set; establishing an edge collection relationship between a mobile monitoring terminal and a server node according to the server node set, and generating a node collection mapping table; obtaining running state data of the server node based on the node collection mapping table, and generating a node state data set; calculating the resource fluctuation degree of the server node according to the node state data set, and dynamically configuring the server node according to the resource fluctuation degree; wherein the dynamic configuration process is completed by the mobile monitoring terminal. Through the establishment of the dynamic edge collection relationship between the mobile monitoring terminal and the server node, the mobile monitoring terminal replaces the internal monitoring process of the server to complete the running state collection, so that the external dynamic perception of the server resource state and resource scheduling are realized.

Description

Technical Field

[0001] This invention relates to the field of resource allocation technology, specifically a server resource allocation method based on mobile edge computing. Background Technology

[0002] With the development of cloud computing, edge computing, and distributed task processing architectures, server clusters have been widely used in scenarios such as video analytics, industrial control, smart grids, online inference, and high-concurrency data processing. In mobile edge computing environments, a large number of server nodes collaboratively execute the same business task. Different server nodes may experience dynamic resource fluctuations due to changes in task load, network status, hardware temperature, and cache usage. Therefore, real-time resource configuration adjustments are necessary to ensure overall task processing efficiency and system stability.

[0003] In existing technologies, monitoring processes are typically deployed internally within server nodes to monitor processor utilization, memory utilization, network bandwidth, and task response status in real time, and then upload the monitoring results to a scheduling center for resource allocation decisions. While this approach enables dynamic management of server resources, the monitoring process runs continuously within the server, consuming processor, cache, and network resources. Especially in high-concurrency scenarios, the monitoring itself can interfere with task execution. Furthermore, when the server node is under high load, its internal monitoring results are susceptible to resource contention, leading to data lag or distortion.

[0004] Therefore, how to reduce the impact of resource contention caused by upload requirements during task processing is the technical problem that this invention aims to solve. Summary of the Invention

[0005] The purpose of this invention is to provide a server resource configuration method based on mobile edge computing to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A method for configuring server resources based on mobile edge computing, the method comprising:

[0008] Obtain the node topology information and node operation information of the server cluster, and generate a set of server nodes;

[0009] Establish the edge acquisition relationship between the mobile monitoring terminal and the server nodes based on the set of server nodes, and generate a node acquisition mapping table;

[0010] The node acquisition mapping table is used to obtain the running status data of the server nodes and generate a node status dataset.

[0011] The resource fluctuation level of the server node is calculated based on the node status dataset, and the server node is dynamically configured according to the resource fluctuation level; wherein, the dynamic configuration process is completed by the mobile monitoring terminal.

[0012] As a further aspect of the present invention: the step of obtaining the node topology information and node operation information of the server cluster, and generating a server node set includes:

[0013] Obtain server node identification information and node connection relationships in the server cluster;

[0014] Establish the server node topology based on the node connection relationships;

[0015] Obtain the component module parameters of the server, and statistically analyze the component module parameters based on the node topology to obtain a set of server nodes; wherein, the component module parameters include the data processing modules included in the server and their module parameters.

[0016] As a further aspect of the present invention: the step of establishing the edge acquisition relationship between the mobile monitoring terminal and the server nodes based on the server node set, and generating a node acquisition mapping table includes:

[0017] Read the component module parameters of each server node, match the tasks to be processed according to the component module parameters, and construct a task sequence;

[0018] The importance of tasks is determined based on the task sequence of each server node, and the environmental risk value is determined based on the environmental parameters of the server nodes.

[0019] Insert the task importance and environmental risk values ​​into the server node set;

[0020] The edge acquisition relationships of mobile monitoring terminals are determined based on a set of server nodes containing task importance and environmental risk values, and a node acquisition mapping table is generated.

[0021] As a further aspect of the present invention: the step of determining the edge acquisition relationship of the mobile monitoring terminal based on a set of server nodes containing task importance and environmental risk values, and generating a node acquisition mapping table includes:

[0022] The demand weight for each server node is determined based on the importance of the task and the environmental risk value.

[0023] Randomly select server nodes, accumulate demand weights, and when the demand weights reach a preset weight threshold, select server nodes to build a node group.

[0024] Record the number of remaining server nodes, repeat the process until the number of server nodes is zero, and count all node groups as the edge acquisition scheme.

[0025] The process is repeated to obtain a preset number of edge acquisition schemes. The edge acquisition schemes are then optimized to determine the final edge acquisition scheme. Based on the final edge acquisition scheme, the edge acquisition relationship between the mobile monitoring terminal and the server node is established, and a node acquisition mapping table is generated.

[0026] As a further aspect of the present invention: the step of obtaining the running status data of the server node based on the node acquisition mapping table and generating a node status dataset includes:

[0027] Establish a data acquisition channel between the mobile monitoring terminal and the server node;

[0028] Obtain processor usage data, memory usage data, processor temperature data, cache usage data, and task response time data corresponding to the server node;

[0029] Obtain network transmission rate data, bandwidth usage data, and task waiting time data corresponding to the server node;

[0030] The acquired data is sorted according to the collection time to obtain a data sequence containing type labels, and a sequence set for each server node is obtained.

[0031] Collect the sequence set of all server nodes to form the node state dataset.

[0032] As a further aspect of the present invention: the step of calculating the resource fluctuation level of the server node based on the node status dataset, and dynamically configuring the server node according to the resource fluctuation level includes:

[0033] Read the data sequence from the node state dataset, compare the data sequence with the preset standard data sequence, and calculate the sequence distance;

[0034] Query the weights of the data types corresponding to the data sequences, and sum the sequence distances based on the weights;

[0035] The sequence distance is normalized to represent the degree of resource fluctuation of the server node;

[0036] The server nodes are dynamically configured based on the degree of resource fluctuation.

[0037] As a further aspect of the present invention: the step of dynamically configuring server nodes according to the degree of resource fluctuation includes:

[0038] The floating resource parameters of the server node corresponding to each mobile monitoring terminal are queried based on the node acquisition mapping table; the floating resource parameters include resource type and resource quantity.

[0039] The floating resource ratio of each server node is determined based on the resource fluctuation of the server node, and a demand ratio array is constructed; where each element in the ratio array corresponds to a server node, and the order of the elements is preset and consistent.

[0040] Compare the demand ratio array with the current ratio array and calculate the array similarity;

[0041] When the array similarity is less than a preset threshold, the dynamic configuration instruction of the mobile monitoring terminal is determined based on the demand ratio array; the dynamic configuration instruction is used to adjust the resource amount of each corresponding server node.

[0042] As a further aspect of the present invention: the method further includes performing monitoring task drift takeover based on the communication status and location change status of the mobile monitoring terminal, and updating the node acquisition mapping table, specifically as follows:

[0043] Acquire communication delay data, battery data, and location change data corresponding to the mobile monitoring terminal;

[0044] Calculate the communication fluctuation value and location drift value corresponding to the mobile monitoring terminal;

[0045] When the communication fluctuation value and the location drift value reach the preset takeover threshold, the server node corresponding to the mobile monitoring terminal is marked as a server node to be switched.

[0046] Obtain the communication status and remaining resource status of other mobile monitoring terminals within a preset range;

[0047] Based on the communication status and the remaining resource status, the monitoring tasks corresponding to the server nodes to be switched are reassigned, and the node acquisition mapping table is updated.

[0048] As a further aspect of the present invention: the method further includes performing resource trend prediction based on resource fluctuation changes within multiple consecutive collection periods, and generating pre-configured resource adjustment results, specifically:

[0049] Record the changes in processor usage and memory usage for each server node over multiple consecutive data collection periods;

[0050] Calculate the resource fluctuation growth rate and the direction of resource fluctuation growth for each server node;

[0051] The amount of fluctuation growth within a preset time span is predicted based on the resource fluctuation growth rate and the resource fluctuation growth direction.

[0052] When the fluctuation growth reaches the preset resource growth threshold, the value of the corresponding server node in the demand ratio array is updated according to the predicted amount.

[0053] Adjust dynamic configuration instructions based on the updated demand ratio array.

[0054] As a further aspect of the present invention: the method further includes performing bypass verification based on the mobile monitoring terminal's data collection results and the server node's internal response results, and updating the node data collection mapping table, specifically as follows:

[0055] Acquire the task response time and task response results collected by the mobile monitoring terminal;

[0056] Retrieve the task execution time and task return results recorded internally by the server node;

[0057] Compare the results collected by the mobile monitoring terminal with the results recorded inside the server node to determine the task response time deviation, network transmission rate deviation, and the difference in task return results.

[0058] When the task response time deviation reaches a preset response threshold and the difference in the task return result is lower than a preset difference threshold, the corresponding server node is determined to be in a network link abnormal state.

[0059] When the task response time deviation and the task return result difference both reach the preset abnormal threshold, the corresponding server node is determined to be in an abnormal server resource state.

[0060] Update the node acquisition mapping table based on abnormal status.

[0061] Compared with existing technologies, the advantages of this invention are as follows: By establishing a dynamic edge acquisition relationship between the mobile monitoring terminal and the server node, the mobile monitoring terminal replaces the internal server monitoring process to complete the acquisition of operational status, thereby achieving externalized dynamic perception and resource scheduling of server resource status. Through joint analysis of processor usage data, memory usage data, cache usage data, bandwidth usage data, and task response time data over multiple consecutive acquisition cycles, the resource fluctuation trends of server nodes can be identified in advance, and the task allocation ratio and resource usage ratio can be dynamically adjusted. Simultaneously, by using a bypass verification mechanism to cross-validate the mobile monitoring results and the server's internal response results, abnormal resource states and network link states can be distinguished, reducing the probability of erroneous scheduling. This solution reduces the resource consumption of internal server monitoring while improving the accuracy of resource configuration and the overall task stability of the server cluster. Attached Figure Description

[0062] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention.

[0063] Figure 1 This is a flowchart of a server resource configuration method based on mobile edge computing.

[0064] Figure 2 This is the first sub-flow diagram of the server resource configuration method based on mobile edge computing.

[0065] Figure 3 This is the second sub-flow diagram of the server resource configuration method based on mobile edge computing.

[0066] Figure 4 This is the third sub-flow diagram of the server resource configuration method based on mobile edge computing.

[0067] Figure 5 This is the fourth sub-flow diagram of the server resource configuration method based on mobile edge computing. Detailed Implementation

[0068] To make the technical problems to be solved, the technical solutions, and the beneficial effects of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present invention and are not intended to limit the present invention.

[0069] Figure 1 The flowchart illustrates a server resource configuration method based on mobile edge computing. In this embodiment of the invention, a server resource configuration method based on mobile edge computing includes:

[0070] Step S100: Obtain the node topology information and node operation information of the server cluster, and generate a set of server nodes;

[0071] The technical solution of this invention is applied to a server cluster, in which multiple servers are used to complete a common overall task, and each server completes a part of the task independently. The invention obtains the node topology information and node operation information of the server cluster. The node topology information is how the various server nodes cooperate and connect with each other, and the node operation information is the operating state of each server node under normal conditions. Based on the node topology information and node operation information, a set of server nodes can be generated. The set of server nodes can be understood as a graph structure. Each node in the graph contains a table (or other types of data) containing node operation information.

[0072] Step S200: Establish the edge acquisition relationship between the mobile monitoring terminal and the server nodes based on the server node set, and generate a node acquisition mapping table;

[0073] Analyzing the set of server nodes allows us to determine which mobile monitoring terminals assist the server nodes. A mobile monitoring terminal can be understood as a relay (edge) with mobile functionality, responsible for monitoring and uploading data from the server nodes. The management range of different mobile monitoring terminals is limited. If the number of server nodes is small, one mobile monitoring terminal can complete the task. However, if the number of servers is large and their distribution is high, multiple mobile monitoring terminals are needed to divide the area and complete the task. Once the correspondence between mobile monitoring terminals and server nodes is determined, it can be stored in the node acquisition mapping table.

[0074] Step S300: Obtain the running status data of the server nodes based on the node acquisition mapping table, and generate a node status dataset;

[0075] The workflow of the mobile monitoring terminal is determined based on the node acquisition mapping table to obtain the corresponding server node's operating status data. The operating status data refers to the working data of different modules during the operation of the server node. This is actually a dataset of different types and times, which needs to be statistically analyzed. The organized data after the statistics are completed is called the node status dataset.

[0076] Step S400: Calculate the resource fluctuation level of the server node based on the node status dataset, and dynamically configure the server node according to the resource fluctuation level; wherein, the dynamic configuration process is completed by the mobile monitoring terminal;

[0077] Analyzing the node status dataset reveals the resource fluctuation level of the server nodes, i.e., their fluctuation during the current operation. Based on this fluctuation, the server nodes are dynamically configured, a process completed by the mobile monitoring terminal. Specifically, analyzing the node status dataset to determine the resource fluctuation level of server nodes is essentially a simple fluctuation analysis. The causes of these fluctuations need explanation. They are related to both the server itself and the environment. During task processing, the server does not necessarily operate according to a specific metric continuously. In some cases, modules such as CPU or memory may experience problems, leading to significant data fluctuations. Furthermore, in software, different task processing may trigger a series of bugs, which also affect data fluctuations. Moreover, environmental factors such as wind speed, temperature, and heat dissipation vary at different locations, and environmental changes also affect data fluctuations. Therefore, the resource fluctuations of each server node in the entire server cluster are different. Analyzing these resource fluctuations and then configuring resources is also performed by the mobile monitoring terminal. For example, the mobile monitoring terminal carries an extension module to update the module performance at each server node.

[0078] Figure 2 The first sub-flow diagram of the server resource configuration method based on mobile edge computing includes the step of obtaining the node topology information and node operation information of the server cluster, and generating a set of server nodes:

[0079] Step S101: Obtain server node identification information and node connection relationships in the server cluster;

[0080] Step S102: Establish the server node topology based on the node connection relationship;

[0081] Step S103: Obtain the component module parameters of the server, and statistically analyze the component module parameters based on the node topology to obtain a set of server nodes; wherein, the component module parameters include the data processing module and its module parameters included in the server.

[0082] A server cluster contains many server nodes. Once determined, the node identification information and node connection relationships are fixed. A server node topology is established based on these connections. If the actual distance parameter is introduced, the distances between different server nodes can be calculated at the same scale when establishing the topology. In this case, the server node topology is essentially a simplified map. Then, the component module parameters of the server are obtained, and these parameters are statistically analyzed based on the node topology to obtain the server node set. These component module parameters may include processor usage data, memory usage data, processor temperature data, cache usage data, and task response time data corresponding to each server node. They may also include network transmission rate data, bandwidth usage data, and task waiting time data corresponding to each server node, depending on the administrator's needs.

[0083] Figure 3 The second sub-flow diagram of the server resource configuration method based on mobile edge computing includes the step of establishing the edge acquisition relationship between the mobile monitoring terminal and the server nodes according to the server node set, and generating a node acquisition mapping table.

[0084] Step S201: Read the component module parameters of each server node, match the tasks to be processed according to the component module parameters, and construct a task sequence;

[0085] Step S202: Determine the task importance based on the task sequence of each server node, and determine the environmental risk value based on the environmental parameters of the server node;

[0086] Step S203: Insert the task importance and environmental risk values ​​into the server node set;

[0087] Step S204: Determine the edge acquisition relationship of the mobile monitoring terminal based on the set of server nodes containing task importance and environmental risk values, and generate a node acquisition mapping table.

[0088] The above describes the process of generating the node acquisition mapping table. It reads the component module parameters of each server node and matches the tasks to be processed based on these parameters. This process involves determining what tasks a server node can perform based on its component module parameters, and then matching the tasks to be processed by that server node. This process is actually a task decomposition process, which is also related to the total number of tasks to be processed. It will not be elaborated further here. For the execution body of this method, it is assumed that each server node has already received the tasks to be processed.

[0089] It should be noted that the number of tasks is generally not unique; a large task may have many smaller tasks, which are called pending tasks. These tasks are statistically analyzed to obtain a task sequence. The task importance is determined based on the task sequence of each server node, and the environmental risk value is determined based on the environmental parameters of the server node. The task importance and environmental risk value are then inserted into the set of server nodes. This process involves using the task importance and environmental risk value as attribute data for each server node. Based on the set of server nodes containing task importance and environmental risk values, the edge acquisition relationship of the mobile monitoring terminal is determined, and a node acquisition mapping table is generated.

[0090] Specifically, regarding task importance and environmental risk value, task importance refers to the importance of each subtask, which is a parameter of the task itself and can be read and accumulated. Environmental risk value is used to characterize the environmental state, including temperature and wind speed. By subtracting it from the preset standard state and then converting the difference, we can obtain a parameter representing environmental risk, called environmental risk value. The larger the difference, the greater the deviation from the standard state, and the greater the environmental risk value.

[0091] As a preferred embodiment of the technical solution of the present invention, the step of determining the edge acquisition relationship of the mobile monitoring terminal based on the server node set containing task importance and environmental risk value, and generating a node acquisition mapping table includes:

[0092] The demand weight for each server node is determined based on the importance of the task and the environmental risk value.

[0093] Randomly select server nodes, accumulate demand weights, and when the demand weights reach a preset weight threshold, select server nodes to build a node group.

[0094] Record the number of remaining server nodes, repeat the process until the number of server nodes is zero, and count all node groups as the edge acquisition scheme.

[0095] The process is repeated to obtain a preset number of edge acquisition schemes. The edge acquisition schemes are then optimized to determine the final edge acquisition scheme. Based on the final edge acquisition scheme, the edge acquisition relationship between the mobile monitoring terminal and the server node is established, and a node acquisition mapping table is generated.

[0096] In one example of the technical solution of this invention, the process of determining the edge acquisition relationship is described. The demand weight of each server node is determined based on the task importance and environmental risk value. The demand weight can also be considered a demand score, representing how much service resources a server node needs from the mobile monitoring terminal. The greater the task importance and the greater the environmental risk value, the greater the demand weight. Server nodes are randomly selected, and the demand weights are accumulated. When the demand weight reaches a preset weight threshold, server nodes are selected to form a node group. The weight threshold is the total amount of service resources that a mobile monitoring terminal can provide. For each server node, server nodes are continuously selected to form a node group. Once a node group reaches the weight threshold, a mobile monitoring terminal is assigned to it. This process is repeated until all server nodes are paired with mobile monitoring terminals. Since the process of selecting server nodes is random, many pairing methods are obtained, and each pairing method is called an edge acquisition scheme. This process is repeated to obtain a preset number of edge acquisition schemes. The edge acquisition schemes are optimized to determine the final edge acquisition scheme. Based on the final edge acquisition scheme, the edge acquisition relationship between the mobile monitoring terminal and the server node is established, and a node acquisition mapping table is generated.

[0097] Regarding the optimization process, a feasible approach is to calculate the maximum distance between internal nodes in each node group and the variance of the required weights of internal nodes. The balance of the node group is evaluated based on the maximum distance and the mean (which is inversely proportional to the maximum distance and the variance). Finally, the mean of the balance of all node groups is calculated. The larger the mean, the more balanced the entire edge acquisition scheme. The edge acquisition scheme with the largest mean is selected as the final edge acquisition scheme.

[0098] Figure 4 The third sub-process flowchart of the server resource configuration method based on mobile edge computing includes the step of obtaining the running status data of server nodes based on the node acquisition mapping table and generating a node status dataset.

[0099] Step S301: Establish a data acquisition channel between the mobile monitoring terminal and the server node;

[0100] Step S302: Obtain processor usage data, memory usage data, processor temperature data, cache usage data, and task response time data corresponding to the server node;

[0101] Step S303: Obtain network transmission rate data, bandwidth usage data, and task waiting time data corresponding to the server node;

[0102] Step S304: Sort the acquired data according to the acquisition time to obtain a data sequence containing type labels, and obtain the sequence set for each server node;

[0103] Step S305: Calculate the sequence set of all server nodes as the node state dataset.

[0104] In one example of the technical solution of this invention, the construction process of the node status dataset is described. A data acquisition channel is established between the mobile monitoring terminal and the server node. Processor usage data, memory usage data, processor temperature data, cache usage data, and task response time data corresponding to the server node are acquired. Network transmission rate data, bandwidth usage data, and task waiting time data corresponding to the server node are also acquired. The acquired data are sorted according to the acquisition time to obtain a data sequence containing type labels. A sequence set for each server node is obtained. Each server node has a sequence set. The sequence sets are statistically analyzed to obtain the node status dataset.

[0105] Figure 5 This is the fourth sub-flow diagram of the server resource configuration method based on mobile edge computing. The step of calculating the resource fluctuation degree of the server node based on the node status dataset and dynamically configuring the server node according to the resource fluctuation degree includes:

[0106] Step S401: Read the data sequence from the node state dataset, compare the data sequence with the preset standard data sequence, and calculate the sequence distance;

[0107] Step S402: Query the weight of the data type corresponding to the data sequence, and sum the sequence distance according to the weight;

[0108] Step S403: Normalize the sequence distance to reflect the resource fluctuation of the server node;

[0109] Step S404: Dynamically configure the server nodes according to the degree of resource fluctuation.

[0110] The above describes the dynamic configuration process. Data sequences are read from the node status dataset and compared with preset standard data sequences to calculate the sequence distance. The sequence distance reflects the difference between each data sequence and the standard state. Generally, a time period needs to be preset when calculating the sequence distance, and data sequences within that time period are compared and the sequence distance is calculated. The weights of the data types corresponding to the data sequences are queried. These weights represent the importance of the corresponding data and are also preset values. The sequence distances are summed based on these weights to obtain the comprehensive difference between the server node and the standard state within the time period. The summed sequence distances are then normalized to represent the resource fluctuation level of the server node. Normalization can be understood as converting it to a dimensionless value within the target range. Finally, the server node is dynamically configured based on the resource fluctuation level.

[0111] As a preferred embodiment of the technical solution of the present invention, the step of dynamically configuring the server node according to the degree of resource fluctuation includes:

[0112] The floating resource parameters of the server node corresponding to each mobile monitoring terminal are queried based on the node acquisition mapping table; the floating resource parameters include resource type and resource quantity.

[0113] The floating resource ratio of each server node is determined based on the resource fluctuation of the server node, and a demand ratio array is constructed; where each element in the ratio array corresponds to a server node, and the order of the elements is preset and consistent.

[0114] Compare the demand ratio array with the current ratio array and calculate the array similarity;

[0115] When the array similarity is less than a preset threshold, the dynamic configuration instruction of the mobile monitoring terminal is determined based on the demand ratio array; the dynamic configuration instruction is used to adjust the resource amount of each corresponding server node.

[0116] In one example of the technical solution of this invention, the specific configuration process is described. Based on the node acquisition mapping table, the floating resource parameters of the server node corresponding to each mobile monitoring terminal are queried. Floating resource parameters are resources that can be transferred between different server nodes, including resource type and quantity. In practical scenarios, these are generally transferable split modules. The floating resource ratio of each server node is determined based on the resource fluctuation level of the server nodes, constructing a demand ratio array. This process is very simple: accumulate the resource fluctuation levels of each server node, calculate the ratio of each server node's resource fluctuation level to the total, and then statistically analyze them into an array according to a preset order to obtain the demand ratio array. Because the floating resources are already in a fixed state in the current state, there exists a current ratio array (the initial ratio array is preset). The demand ratio array is compared with the current ratio array, and the array similarity is calculated. When the array similarity is less than a preset threshold, it indicates that the current demand cannot be met. At this time, the dynamic configuration process is triggered, and the dynamic configuration instructions for the mobile monitoring terminal are determined based on the demand ratio array. The dynamic configuration instructions are used to adjust the resource quantities of the corresponding server nodes, such as CPU, memory, and heat sink resources.

[0117] As a preferred embodiment of the technical solution of the present invention, the method further includes performing monitoring task drift takeover based on the communication status and location change status of the mobile monitoring terminal, and updating the node acquisition mapping table, specifically:

[0118] Acquire communication delay data, battery data, and location change data corresponding to the mobile monitoring terminal;

[0119] Calculate the communication fluctuation value and location drift value corresponding to the mobile monitoring terminal;

[0120] When the communication fluctuation value and the location drift value reach the preset takeover threshold, the server node corresponding to the mobile monitoring terminal is marked as a server node to be switched.

[0121] Obtain the communication status and remaining resource status of other mobile monitoring terminals within a preset range;

[0122] Based on the communication status and the remaining resource status, the monitoring tasks corresponding to the server nodes to be switched are reassigned, and the node acquisition mapping table is updated.

[0123] In one example of the technical solution of this invention, the pairing process between a mobile monitoring terminal and a server node is described. This involves acquiring communication delay data, battery power data, and location change data corresponding to the mobile monitoring terminal. Based on these data, communication fluctuation values ​​and location drift values ​​are calculated. When the communication fluctuation values ​​and location drift values ​​reach a preset takeover threshold, the server node corresponding to the mobile monitoring terminal is marked as a server node to be switched. The communication status and remaining resource status of other mobile monitoring terminals within a preset range are acquired. Based on these communication status and remaining resource status, monitoring tasks corresponding to the server node to be switched are reallocated, and the node acquisition mapping table is updated. In simpler terms, when a mobile monitoring terminal cannot function properly, a switching and pairing process exists between different mobile monitoring terminals.

[0124] As a preferred embodiment of the technical solution of the present invention, the method further includes performing resource trend prediction based on resource fluctuation changes within multiple consecutive collection periods, and generating pre-configured resource adjustment results, specifically:

[0125] Record the changes in processor usage and memory usage for each server node over multiple consecutive data collection periods;

[0126] Calculate the resource fluctuation growth rate and the direction of resource fluctuation growth for each server node;

[0127] The amount of fluctuation growth within a preset time span is predicted based on the resource fluctuation growth rate and the resource fluctuation growth direction.

[0128] When the fluctuation growth reaches the preset resource growth threshold, the value of the corresponding server node in the demand ratio array is updated according to the predicted amount.

[0129] Adjust dynamic configuration instructions based on the updated demand ratio array.

[0130] In one embodiment of the technical solution of this invention, a prediction process is introduced. The processor usage change value and memory usage change value of each server node are recorded in multiple consecutive collection cycles. Based on the processor usage change value and memory usage change value, the resource fluctuation growth rate and resource fluctuation growth direction of each server node are calculated. Based on the resource fluctuation growth rate and resource fluctuation growth direction, the fluctuation growth amount within a preset time span is predicted. When the fluctuation growth amount reaches a preset resource growth threshold, the value of the corresponding server node in the demand ratio array is updated according to the prediction amount. At this time, based on the analysis of the current demand, the prediction content is added. After the demand ratio array is updated, the dynamic configuration instruction is updated accordingly.

[0131] As a preferred embodiment of the technical solution of the present invention, the method further includes performing bypass verification based on the mobile monitoring terminal's collection results and the server node's internal response results, and updating the node collection mapping table, specifically:

[0132] Acquire the task response time and task response results collected by the mobile monitoring terminal;

[0133] Retrieve the task execution time and task return results recorded internally by the server node;

[0134] Compare the results collected by the mobile monitoring terminal with the results recorded inside the server node to determine the task response time deviation, network transmission rate deviation, and the difference in task return results.

[0135] When the task response time deviation reaches a preset response threshold and the difference in the task return result is lower than a preset difference threshold, the corresponding server node is determined to be in a network link abnormal state.

[0136] When the task response time deviation and the task return result difference both reach the preset abnormal threshold, the corresponding server node is determined to be in an abnormal server resource state.

[0137] Update the node acquisition mapping table based on abnormal status.

[0138] In one embodiment of the technical solution of this invention, another pairing process update process is introduced, which involves acquiring the task response time and task response result collected by the mobile monitoring terminal; acquiring the task execution time and task return result recorded internally by the server node, wherein the task response time and task response result collected by the mobile monitoring terminal; and the task execution time and task return result recorded internally by the server node are essentially the same data, only the acquisition subject is different, and the different names are for the purpose of differentiation; comparing the results collected by the mobile monitoring terminal with the results recorded internally by the server node, the task response time deviation value, network transmission rate deviation value, and task return result difference degree are determined. When the task response time deviation value reaches a preset response threshold and the task return result difference degree is lower than a preset difference threshold, the corresponding server node is determined to be in a network link abnormal state; when the task response time deviation value and the task return result difference degree both reach a preset abnormal threshold, the corresponding server node is determined to be in a server resource abnormal state; and updating the node acquisition mapping table according to the abnormal state, the update process may be that for server nodes in such abnormal states, an independent data acquisition process is directly set up, and the mobile monitoring terminal is no longer used, and they are not included in the selection range during pairing.

[0139] It is worth mentioning that the data process of a server node is not actually uploaded in real time, but acquired periodically. Therefore, for example, staff members periodically go to the server node to collect local data. This also means that the cycle of the above-mentioned anomaly identification process is actually the maintenance cycle of the server node.

[0140] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A server resource allocation method based on mobile edge computing, characterized in that, The method includes: Obtain the node topology information and node operation information of the server cluster, and generate a set of server nodes; Establish the edge acquisition relationship between the mobile monitoring terminal and the server nodes based on the set of server nodes, and generate a node acquisition mapping table; The node acquisition mapping table is used to obtain the running status data of the server nodes and generate a node status dataset. The resource fluctuation level of the server node is calculated based on the node status dataset, and the server node is dynamically configured according to the resource fluctuation level; wherein, the dynamic configuration process is completed by the mobile monitoring terminal.

2. The server resource allocation method based on mobile edge computing according to claim 1, characterized in that, The steps of obtaining the node topology information and node operation information of the server cluster, and generating a set of server nodes include: Obtain server node identification information and node connection relationships in the server cluster; Establish the server node topology based on the node connection relationships; Obtain the component module parameters of the server, and statistically analyze the component module parameters based on the node topology to obtain a set of server nodes; wherein, the component module parameters include the data processing modules included in the server and their module parameters.

3. The server resource allocation method based on mobile edge computing according to claim 1, characterized in that, The step of establishing the edge acquisition relationship between the mobile monitoring terminal and the server nodes based on the server node set, and generating a node acquisition mapping table includes: Read the component module parameters of each server node, match the tasks to be processed according to the component module parameters, and construct a task sequence; The importance of tasks is determined based on the task sequence of each server node, and the environmental risk value is determined based on the environmental parameters of the server nodes. Insert the task importance and environmental risk values ​​into the server node set; The edge acquisition relationships of mobile monitoring terminals are determined based on a set of server nodes containing task importance and environmental risk values, and a node acquisition mapping table is generated.

4. The server resource allocation method based on mobile edge computing according to claim 3, characterized in that, The step of determining the edge acquisition relationship of the mobile monitoring terminal based on a set of server nodes containing task importance and environmental risk values, and generating a node acquisition mapping table, includes: The demand weight of each server node is determined based on the importance of the task and the environmental risk value. Randomly select server nodes, accumulate demand weights, and when the demand weights reach a preset weight threshold, select server nodes to build a node group. Record the number of remaining server nodes, repeat the process until the number of server nodes is zero, and count all node groups as the edge acquisition scheme. The process is repeated to obtain a preset number of edge acquisition schemes. The edge acquisition schemes are then optimized to determine the final edge acquisition scheme. Based on the final edge acquisition scheme, the edge acquisition relationship between the mobile monitoring terminal and the server node is established, and a node acquisition mapping table is generated.

5. The server resource allocation method based on mobile edge computing according to claim 1, characterized in that, The step of obtaining the server node's running status data based on the node acquisition mapping table and generating a node status dataset includes: Establish a data acquisition channel between the mobile monitoring terminal and the server node; Obtain processor usage data, memory usage data, processor temperature data, cache usage data, and task response time data corresponding to the server node; Obtain network transmission rate data, bandwidth usage data, and task waiting time data corresponding to the server node; The acquired data is sorted according to the collection time to obtain a data sequence containing type labels, and a sequence set for each server node is obtained. Collect the sequence set of all server nodes to form the node state dataset.

6. The server resource allocation method based on mobile edge computing according to claim 1, characterized in that, The steps of calculating the resource fluctuation level of the server node based on the node status dataset and dynamically configuring the server node according to the resource fluctuation level include: Read the data sequence from the node state dataset, compare the data sequence with the preset standard data sequence, and calculate the sequence distance; Query the weights of the data types corresponding to the data sequences, and sum the sequence distances based on the weights; The sequence distance is normalized to represent the degree of resource fluctuation of the server node; The server nodes are dynamically configured based on the degree of resource fluctuation.

7. The server resource allocation method based on mobile edge computing according to claim 6, characterized in that, The steps for dynamically configuring server nodes based on resource fluctuation levels include: The floating resource parameters of the server node corresponding to each mobile monitoring terminal are queried based on the node acquisition mapping table; the floating resource parameters include resource type and resource quantity. The floating resource ratio of each server node is determined based on the resource fluctuation of the server node, and a demand ratio array is constructed; where each element in the ratio array corresponds to a server node, and the order of the elements is preset and consistent. Compare the demand ratio array with the current ratio array and calculate the array similarity; When the array similarity is less than a preset threshold, the dynamic configuration instruction of the mobile monitoring terminal is determined based on the demand ratio array; the dynamic configuration instruction is used to adjust the resource amount of each corresponding server node.

8. The server resource allocation method based on mobile edge computing according to claim 1, characterized in that, The method further includes performing monitoring task drift takeover based on the communication status and location change status of the mobile monitoring terminal, and updating the node acquisition mapping table, specifically: Acquire communication delay data, battery data, and location change data corresponding to the mobile monitoring terminal; Calculate the communication fluctuation value and location drift value corresponding to the mobile monitoring terminal; When the communication fluctuation value and the location drift value reach the preset takeover threshold, the server node corresponding to the mobile monitoring terminal is marked as a server node to be switched. Obtain the communication status and remaining resource status of other mobile monitoring terminals within a preset range; Based on the communication status and the remaining resource status, the monitoring tasks corresponding to the server nodes to be switched are reassigned, and the node acquisition mapping table is updated.

9. The server resource allocation method based on mobile edge computing according to claim 7, characterized in that, The method further includes performing resource trend prediction based on resource fluctuation changes over multiple consecutive collection periods, and generating pre-configured resource adjustment results, specifically: Record the changes in processor usage and memory usage for each server node over multiple consecutive data collection periods; Calculate the resource fluctuation growth rate and the direction of resource fluctuation growth for each server node; The amount of fluctuation growth within a preset time span is predicted based on the resource fluctuation growth rate and the resource fluctuation growth direction. When the fluctuation growth reaches the preset resource growth threshold, the value of the corresponding server node in the demand ratio array is updated according to the predicted amount. Adjust dynamic configuration instructions based on the updated demand ratio array.

10. The server resource allocation method based on mobile edge computing according to claim 1, characterized in that, The method further includes performing bypass verification based on the mobile monitoring terminal's data collection results and the server node's internal response results, and updating the node data collection mapping table, specifically: Acquire the task response time and task response results collected by the mobile monitoring terminal; Retrieve the task execution time and task return results recorded internally by the server node; Compare the results collected by the mobile monitoring terminal with the results recorded inside the server node to determine the deviation values ​​of task response time, network transmission rate, and task return results. When the task response time deviation reaches a preset response threshold and the difference in the task return result is lower than a preset difference threshold, the corresponding server node is determined to be in a network link abnormal state. When the deviation of the task response time and the difference of the task return result both reach the preset abnormal threshold, the corresponding server node is determined to be in an abnormal server resource state. Update the node acquisition mapping table based on abnormal status.

Images