Database cache strategy adjustment method, device, and equipment

Abstract

The present invention provides a database cache strategy adjustment method, device, and equipment, which can be applied to the fields of database management technology and data cache strategy optimization technology. The database cache strategy adjustment method includes: based on the data access information of the database system in the historical time period, predicting the data access situation of the database system in the target time period, and obtaining the data access information of the target time period; based on the data access information of the target time period, the data access information of the historical time period, and the system load information of the historical time period, generating the environment state vector of the target time period; inputting the environment state vector into the reinforcement learning model, outputting the cache strategy adjustment information of the target time period; and adjusting the cache strategy of the database cache space based on the cache strategy adjustment information of the target time period.

Description

Technical Field

[0001] The present invention relates to the technical field of database management and data cache strategy optimization, and more specifically to a database cache strategy adjustment method, device and equipment. Background Art

[0002] As the size of data continues to grow and the demand for fast data access continues to increase, database system performance optimization becomes increasingly important. Setting up data cache is one of the key means to improve database system performance. By caching hot data in memory, disk input / output (I / O) operations can be reduced and data access can be accelerated.

[0003] Traditional database cache strategies mainly include algorithms such as Least Recently Used (LRU) and Least Frequently Used (LFU). These algorithms are often based on fixed rules or static strategies to manage data caches, and are difficult to adapt to the dynamic changes in database access patterns. Since the access patterns of database systems usually change over time and with changes in data volume, static cache strategies cannot flexibly adapt to such changes, resulting in a decrease in database cache performance. Summary of the invention

[0004] In view of the above problems, the present invention provides a method, apparatus and device for adjusting a database cache strategy.

[0005] According to a first aspect of the present invention, a method for adjusting a database cache strategy is provided, characterized in that the method comprises: based on the data access information of the database system in the historical time period, predicting the data access situation of the database system in the target time period, and obtaining the data access information of the target time period. Based on the data access information of the target time period, the data access information of the historical time period, and the system load information of the historical time period, an environmental state vector of the target time period is generated. The environmental state vector is input into a reinforcement learning model, and the cache strategy adjustment information of the target time period is output. Based on the cache strategy adjustment information of the target time period, the cache strategy of the database cache space is adjusted.

[0006] According to an embodiment of the present invention, the above-mentioned generating the environment state vector of the target time period based on the data access information of the target time period, the data access information of the historical time period and the system load information of the historical time period includes: extracting features from the data access information of the target time period, the data access information of the historical time period and the system load information of the historical time period to obtain multiple environment state features. Encoding the multiple environment state features to obtain the above-mentioned environment state vector.

[0007] According to an embodiment of the present invention, the data access information of the target time period, the data access information of the historical time period, and the system load information of the historical time period are subjected to feature extraction to obtain multiple environmental status features, including: determining target indicator data and historical indicator data of multiple data access indicators based on the data access information of the target time period and the data access information of the historical time period. Determining at least one target access indicator from the multiple data access indicators based on the comparison results between the target indicator data and historical indicator data of the multiple data access indicators. Determining historical indicator data of multiple system load indicators based on the system load information of the historical time period. Determining at least one target load indicator from the multiple system load indicators based on the comparison results between the historical indicator data of the multiple system load indicators. Feature extraction is performed on the target indicator data and historical indicator data of the at least one target access indicator, and the historical load data of the at least one target load indicator to obtain the multiple environmental status features.

[0008] According to an embodiment of the present invention, based on the comparison results between the target indicator data and the historical indicator data of each of the above-mentioned multiple data access indicators, at least one target access indicator is determined from the above-mentioned multiple data access indicators, including: for each of the above-mentioned data access indicators, when the comparison result of the above-mentioned data access indicators indicates that the change amplitude of the above-mentioned data access indicator between the above-mentioned target time period and the above-mentioned historical time period is greater than a preset threshold, the above-mentioned data access indicator is determined as the target access indicator.

[0009] According to an embodiment of the present invention, the data access index includes: cache hit rate, cache eviction rate, data access frequency and data activity. The cache hit rate, cache eviction rate, data access frequency and data activity are detected by a detection tool of the database system.

[0010] According to an embodiment of the present invention, encoding the multiple environmental state features to obtain an environmental state vector includes: mapping the multiple environmental state features into one or more feature vectors, and concatenating the one or more feature vectors to obtain the environmental state vector.

[0011] According to an embodiment of the present invention, the above-mentioned inputting the above-mentioned environment state vector into the reinforcement learning model and outputting the cache policy adjustment information of the target time period includes: determining one or more target policy adjustment actions from multiple policy adjustment actions based on the above-mentioned environment state vector. Determining the above-mentioned cache policy adjustment information based on the above-mentioned one or more target policy adjustment actions.

[0012] According to an embodiment of the present invention, the method further includes: determining cache performance information of the target time period based on the data access information of the historical time period and the data access information of the target time period. Determining load change information of the target time period based on the system load information of the historical time period and the system load information of the target time period. Determining multiple feedback signals based on the cache performance information of the target time period and the load change information of the target time period. Adjusting the parameters of the reinforcement learning model based on the multiple feedback signals to obtain a new reinforcement learning model.

[0013] The second aspect of the present invention provides a database cache strategy adjustment device, characterized in that the above-mentioned device includes: an information prediction module, which is used to predict the data access situation of the above-mentioned database system in a target time period based on the data access information of the database system in the historical time period, and obtain the data access information of the target time period. A vector generation module is used to generate an environmental state vector of the target time period based on the data access information of the above-mentioned target time period, the data access information of the above-mentioned historical time period and the system load information of the above-mentioned historical time period. A strategy determination module is used to input the above-mentioned environmental state vector into a reinforcement learning model and output the cache strategy adjustment information of the target time period. A strategy adjustment module is used to adjust the cache strategy of the database cache space based on the cache strategy adjustment information of the above-mentioned target time period.

[0014] A third aspect of the present invention provides an electronic device, comprising: one or more processors; and a memory for storing one or more computer programs, wherein the one or more processors execute the one or more computer programs to implement the steps of the above method.

[0015] According to an embodiment of the present invention, by predicting the data access information of the database system in the target time period, an accurate decision-making basis is provided for the reinforcement learning model, thereby improving the accuracy of the cache policy adjustment information. Based on the data access information of the target time period, the data access information of the historical time period, and the system load information of the historical time period, an environmental state vector of the reinforcement learning model is generated, and the cache policy adjustment information obtained by the reinforcement learning model is used to adjust the cache policy of the database cache space. Since the reinforcement learning model can output the cache policy adjustment information corresponding to the changed environmental state vector in response to the change of the environmental state vector, it helps the database system to continuously and adaptively adjust the cache policy under the constantly changing database access model, thereby improving the cache efficiency of the database system. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The above contents and other objects, features and advantages of the present invention will become more apparent through the following description of the embodiments of the present invention with reference to the accompanying drawings.

[0017] Figure 1 An application scenario diagram of a database cache strategy adjustment method, apparatus, and device according to an embodiment of the present invention is shown.

[0018] Figure 2 A flow chart of a method for adjusting a database cache strategy according to an embodiment of the present invention is shown.

[0019] Figure 3 A flow chart for determining environmental status characteristics according to an embodiment of the present invention is shown.

[0020] Figure 4 A flow chart for determining environmental status characteristics according to another embodiment of the present invention is shown.

[0021] Figure 5 A structural block diagram of a database cache strategy adjustment device according to an embodiment of the present invention is shown.

[0022] Figure 6 A block diagram of an electronic device suitable for implementing a database cache strategy adjustment method according to an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0023] Below, embodiments of the present invention will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present invention. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of embodiments of the present invention. However, it is apparent that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of known structures and technologies are omitted to avoid unnecessary confusion of concepts of the present invention.

[0024] The terms used herein are only for describing specific embodiments and are not intended to limit the present invention. The terms "comprise", "include", etc. used herein indicate the existence of the features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0025] All terms (including technical and scientific terms) used herein have the meanings commonly understood by those skilled in the art unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0026] When using expressions such as "at least one of A, B, and C, etc.", they should generally be interpreted according to the meaning of the expression commonly understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include but is not limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.).

[0027] In order to improve the cache performance of the database, the traditional strategies for optimizing the cache space generally include using data replacement strategies for the cache space (for example, LRU, LFU, etc.), or improving cache performance by using the expiration time of the data in the appropriate cache space, where the expiration time of the data indicates the time when the data loses its validity. Specifically, after the data expires, the cache system will no longer return the data and needs to re-acquire the latest data from the data source. However, these algorithms often manage data caches based on fixed rules or static strategies, and are difficult to adapt to the dynamic changes in database access patterns. Since the access patterns of database systems usually change with time and data volume, static cache strategies cannot flexibly adapt to such changes, resulting in performance degradation.

[0028] An embodiment of the present invention provides a database cache strategy adjustment method for adjusting a database cache strategy, comprising: predicting the data access situation of the database system in a target time period based on data access information of the database system in a historical time period to obtain data access information of the target time period; generating an environment state vector for the target time period based on the data access information of the target time period, the data access information of the historical time period and the system load information of the historical time period; inputting the environment state vector into a reinforcement learning model to output cache strategy adjustment information for the target time period; and adjusting the cache strategy of the database cache space based on the cache strategy adjustment information for the target time period.

[0029] Figure 1 An application scenario diagram of a database cache strategy adjustment method, apparatus, and device according to an embodiment of the present invention is shown.

[0030] like Figure 1 As shown, the application scenario 100 according to this embodiment may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 is used to provide a medium for a communication link between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired, wireless communication links, or optical fiber cables, etc.

[0031] The user can use the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 through the network 104 to receive or send messages, etc. The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be various electronic devices with display screens and supporting web browsing, including but not limited to smart phones, tablet computers, laptop portable computers, desktop computers, etc.

[0032] The server 105 may be a server that provides various services, and the database may run on the server 105, for example, a background management server (only as an example) that provides support for websites browsed by users using the first terminal device 101, the second terminal device 102, and the third terminal device 103. The background management server may analyze and process the received data such as user requests, and feed back the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal devices.

[0033] It should be noted that the database cache strategy adjustment method provided in the embodiment of the present invention can generally be executed by the server 105. The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be used to adjust the algorithm of the database cache strategy adjustment method. Accordingly, the database cache strategy adjustment device provided in the embodiment of the present invention can generally be set in the server 105. The database cache strategy adjustment method provided in the embodiment of the present invention can also be executed by a server or server cluster that is different from the server 105 and can communicate with the first terminal device 101, the second terminal device 102, the third terminal device 103 and / or the server 105. Correspondingly, the database cache strategy adjustment device provided in the embodiment of the present invention can also be set in a server or server cluster that is different from the server 105 and can communicate with the first terminal device 101, the second terminal device 102, the third terminal device 103 and / or the server 105.

[0034] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to the implementation requirements.

[0035] The following will be based on Figure 1 The scene described by Figure 2~Figure 4 The database cache strategy adjustment method of the embodiment of the invention is described in detail.

[0036] Figure 2 A flow chart of a method for adjusting a database cache strategy according to an embodiment of the present invention is shown.

[0037] like Figure 2As shown, the database cache strategy adjustment of this embodiment includes operations S210 to S240.

[0038] In operation S210, based on the data access information of the database system in the historical time period, the data access situation of the database system in the target time period is predicted to obtain the data access information of the target time period.

[0039] In operation S220 , an environment state vector of the target time period is generated based on the data access information of the target time period, the data access information of the historical time period, and the system load information of the historical time period.

[0040] In operation S230 , the environment state vector is input into the reinforcement learning model, and cache strategy adjustment information for the target time period is output.

[0041] In operation S240, the cache strategy of the database cache space is adjusted based on the cache strategy adjustment information of the target time period.

[0042] According to an embodiment of the present invention, the data access information of the historical time period represents the information related to the data access operation in the database system in the historical time period, which may include the data accessed, the frequency of data access, etc. The prediction model can be obtained by modeling and training through the autoregressive moving average model, and the data access situation of the database system in the target time period is predicted based on the trained prediction model to obtain the data access information of the database system in the target time period.

[0043] According to an embodiment of the present invention, the system load information characterizes information used to represent the system load situation, which may include the system's memory occupancy rate, processor utilization rate, and data read and write utilization rate, wherein the data read and write utilization rate may be represented by the time occupied by the input and output operations of the database system. The environment state vector is a vector constructed from data in the system load information and the data access information. The environment state vector is generated by combining the data access information of the historical time period, the system load information of the historical time period, and the predicted data access information of the target time period. Thus, the above information is converted into a form that can be used for processing by the reinforcement learning model.

[0044] According to an embodiment of the present invention, the environment state vector is input into a reinforcement learning model, wherein the reinforcement learning model is pre-trained using data access information in historical data, system load information, and data access information for a time period in which the cache strategy is to be adjusted. The cache strategy adjustment information is used to adjust the strategy of the current database cache, and the cache strategy can be used to determine a replacement strategy for cache data, data cached in a cache space, a priority of cache data, and the amount of resources that can be stored in the cache space.

[0045] According to an embodiment of the present invention, the cache strategy characterizes the method used by the database to cache data, which may include the expiration time of cached data in the database cache space, the replacement method of cached data, the maximum storage data volume set for the cache space, etc. The cache strategy adjustment information may include multiple component identifiers and data object identifiers corresponding to each of the multiple component identifiers, wherein the component identifier is an identifier of a reusable module or component for adjusting the cache strategy, and the data object identifier characterizes the identifier of the data accessed by the data access request of the accessed database, for example, the data object identifier may be the storage address of the data. The server adjusts the cache strategy of the database cache space in the target time period based on the multiple component identifiers and the data object identifiers corresponding to each of the multiple component identifiers.

[0046] According to an embodiment of the present invention, by predicting the data access information of the database system in the target time period, an accurate decision-making basis is provided for the reinforcement learning model, thereby improving the accuracy of the cache policy adjustment information. Based on the data access information of the target time period, the data access information of the historical time period, and the system load information of the historical time period, an environmental state vector of the reinforcement learning model is generated, and the cache policy adjustment information obtained by the reinforcement learning model is used to adjust the cache policy of the database cache space. Since the reinforcement learning model can output the cache policy adjustment information corresponding to the changed environmental state vector in response to the change of the environmental state vector, it helps the database system to continuously and adaptively adjust the cache policy under the constantly changing database access model, thereby improving the cache efficiency of the database system.

[0047] According to an embodiment of the present invention, an environment state vector of a target time period is generated based on data access information of a target time period, data access information of a historical time period, and system load information of a historical time period, including: extracting features of the data access information of the target time period, the data access information of the historical time period, and the system load information of the historical time period to obtain a plurality of environment state features; and encoding the plurality of environment state features to obtain an environment state vector.

[0048] According to an embodiment of the present invention, the environmental status feature is used to characterize the features corresponding to the information related to the access request and the information related to the system load of the database system. The environmental status features may include features of the system's memory occupancy rate, processor usage rate, and data read and write usage rate, as well as features of the access frequency and access time corresponding to the data. Feature extraction is performed on the data access information of the target time period, the data access information of the historical time period, and the system load information of the historical time period. Multiple environmental status features can be obtained by extracting features from the indicator data of the data access information of the target time period, the data access information of the historical time period, and the system load information of the historical time period.

[0049] According to an embodiment of the present invention, by encoding multiple environmental state features, a specific numerical form that can be processed by a reinforcement learning model, namely, an environmental state vector, is obtained, so that the reinforcement learning model can dynamically generate cache strategy adjustment information according to different inputs to cope with the ever-changing workload and database access mode.

[0050] According to an embodiment of the present invention, feature extraction is performed on data access information of a target time period, data access information of a historical time period, and system load information of a historical time period to obtain a plurality of environmental status features, including: determining target indicator data and historical indicator data of a plurality of data access indicators based on the data access information of the target time period and the data access information of the historical time period; determining at least one target access indicator from a plurality of data access indicators based on a comparison result between the target indicator data and the historical indicator data of a plurality of data access indicators; determining historical indicator data of a plurality of system load indicators based on the system load information of the historical time period; determining at least one target load indicator from a plurality of system load indicators based on a comparison result of the historical indicator data of a plurality of system load indicators; and feature extraction is performed on the target indicator data and historical indicator data of at least one target access indicator, and the historical load data of at least one target load indicator to obtain a plurality of environmental status features.

[0051] According to an embodiment of the present invention, the data access index is used to characterize an index representing database cache performance or cache data performance. For example, the index of database cache performance may be a cache hit rate; the index of cache data performance may be the frequency of data access. By preprocessing the data access information of the target time period and the data access information of the historical time period, which may include data cleaning, denoising, data format conversion, etc., the target index data and historical index data of each of the multiple data access indexes are obtained.

[0052] According to an embodiment of the present invention, the difference between the target indicator data and the historical indicator data can be quantified, for example, the cache hit rate can be represented by a numerical value. Based on the quantified data, the difference between the target indicator data and the historical indicator data of each of the multiple data access indicators is compared, and the data access indicator with the larger difference is determined as the target access indicator. Based on the quantified difference between the target indicator data and the historical indicator data of each of the multiple data access indicators, N data access indicators with the larger difference among the multiple data access indicators are determined as the target access indicators, where N is a positive integer.

[0053] According to an embodiment of the present disclosure, the comparison results between the target indicator data and the historical indicator data of each of the multiple data access indicators can characterize the difference between the target indicator data and the historical indicator data of each of the multiple data access indicators after quantization. For example, for the data access indicator of cache hit rate, the cache hit rate in the target time period is increased by 5% compared with the cache hit rate in the historical time period. The cache hit rate is increased by 5%, which is the comparison result of the target indicator data and the historical indicator data of the cache hit rate. According to an embodiment of the present invention, based on the comparison results between the target indicator data and the historical indicator data of each of the multiple data access indicators, at least one target access indicator is determined from the multiple data access indicators, including: for each data access indicator, when the comparison result of the data access indicator characterizes that the change amplitude of the data access indicator between the target time period and the historical time period is greater than a preset threshold, the data access indicator is determined as the target access indicator.

[0054] According to an embodiment of the present invention, different preset thresholds can be set for different data access indicators, wherein the preset threshold can be set based on expert experience, or the average of the change amount of the data access indicator in multiple historical time periods is determined as the preset threshold. By determining the data access indicator whose change range between the target time period and the historical time period is greater than the preset threshold as the target access indicator, the accuracy of the determined target access indicator is improved.

[0055] According to an embodiment of the present invention, the system load index may include memory occupancy, processor occupancy and data read / write usage rate, wherein the data read / write usage rate represents the time taken by the database system to input or output data from the disk. The historical load data of multiple system load indicators are constantly changing during the historical time period, and the range of change of the historical load data of the multiple system load indicators is different. Therefore, by comparing the range of change of the historical load data of the multiple data load indicators, the system load indicator with the larger range of change is determined as the target load indicator. The comparison result of the historical indicator data of the system load indicator represents the range of change of the historical load data of the multiple data load indicators. For example, the range of change of the processor usage rate during the historical time period.

[0056] According to an embodiment of the present invention, by presetting a preset threshold, when the variation range among historical load data of multiple system load indicators is greater than the preset threshold, the historical load data corresponding to the system load indicator greater than the preset threshold is used as the target load data. Alternatively, the historical load data of the first M data load indicators with the larger variation range are determined as the target load indicator.

[0057] According to an embodiment of the present invention, in the process of determining the variation range of the system load index, the variation range of the system load index is determined by calculating the standard deviation or variance in multiple time windows in the historical time period, and based on the change rate between the standard deviation or variance obtained in the multiple time windows. Feature extraction is performed on the target index data and the historical load data of each target access index respectively to obtain the environmental state feature corresponding to the target index data of each target access index, and the environmental state feature corresponding to the historical access data of each target access index. Feature extraction is performed on the historical load data of each target load index to obtain the environmental state feature corresponding to each target load index.

[0058] According to an embodiment of the present invention, feature extraction is performed on the target indicator data and historical load data of each target access indicator to obtain environmental state features, and the method further includes: determining a change rate threshold of the target access indicator in the historical time period and the target time period based on the target indicator data of the target access indicator and the historical load data of the target access indicator. Feature extraction is performed on the change rate threshold of the target access indicator in the historical time period and the target time period to obtain environmental state features corresponding to each target access indicator.

[0059] According to an embodiment of the present invention, the data access information of the target time period and the data access information of the historical time period, as well as the system load information of the historical time period, are huge in quantity. Through the threshold of the rate of change of the data access information in the historical time period and the target time period, and the amplitude of the change of the system load index in the historical time period, the historical load data corresponding to the system load index with large changes, and the historical index data and target index data corresponding to the access data index with large changes are used as the objects for extracting features, and multiple environmental state features are obtained, thereby reducing the data dimension and the amount of calculation of the subsequent reinforcement learning model. In addition, representative data is used as the data for extracting environmental state features, which reduces the influence of unimportant data on the reinforcement learning model and improves the accuracy of the cache strategy adjustment information.

[0060] Figure 3 A flow chart for determining environmental status characteristics according to an embodiment of the present invention is shown.

[0061] like Figure 3 As shown, the operation of determining the environmental state characteristics in this embodiment includes operation S301 to operation S305.

[0062] In operation S301 , target indicator data and historical indicator data of each of a plurality of data access indicators are determined based on data access information of a target time period and data access information of a historical time period.

[0063] In operation S302, it is determined whether the variation range of the data access indicator between the target time period and the historical time period is greater than a preset threshold corresponding to the data access indicator.

[0064] When the variation range of the data access indicator between the target time period and the historical time period is greater than a preset threshold corresponding to the data access indicator, operation S303 is performed; otherwise, operation S304 is performed.

[0065] In operation S303 , the data access indicator is determined as a target access indicator.

[0066] In operation S304, it is determined whether all data access indicators are traversed.

[0067] In the case of traversing all data access indicators, operation S305 is performed. Otherwise, operation S302 is performed.

[0068] In operation S305 , feature extraction is performed on the target indicator data and the historical indicator data of each target access indicator to obtain an environmental state feature corresponding to each target access indicator.

[0069] Figure 4 A flow chart for determining environmental status characteristics according to another embodiment of the present invention is shown.

[0070] like Figure 4 As shown, the operation of determining the environmental state characteristics in this embodiment includes operation S401 to operation S405.

[0071] In operation S401, historical indicator data of each of a plurality of system load indicators is determined based on system load information of a historical time period.

[0072] In operation S402, it is determined whether the variation range of the system load indicator between the target time period and the historical time period is greater than a preset threshold corresponding to the system load indicator.

[0073] When the variation range of the system load indicator between the target time period and the historical time period is greater than a preset threshold corresponding to the system load indicator, operation S403 is performed; otherwise, operation S404 is performed.

[0074] In operation S403, the system load index is determined as a target load index.

[0075] In operation S404, it is determined whether all system load indicators are traversed.

[0076] In the case of traversing all target load indicators, operation S405 is performed. Otherwise, operation S402 is performed.

[0077] In operation S405 , feature extraction is performed on the historical load data of each target load indicator to obtain an environmental state feature corresponding to each target load indicator.

[0078] According to an embodiment of the present invention, data access indicators include: cache hit rate, cache eviction rate, data access frequency and data activity; wherein, cache hit rate, cache eviction rate, data access frequency and data activity are detected by a detection tool of the database system.

[0079] According to an embodiment of the present invention, the cache eviction rate represents the ratio of the number of requests to evict cached data when the cache space is insufficient. The data access frequency represents the frequency with which data objects in the database are accessed. There may be multiple data objects accessed by data access requests in the historical time period and the target time period. The data activity represents the immediacy requirement of the data object. The greater the data activity, the more frequently the data is queried, and the higher the immediacy requirement for the data object. The data activity can be determined by the ratio of the data access frequency of the data object to the average access interval. The detection tool of the database system is a tool provided by the database management system, which is used to obtain data such as the database cache hit rate, cache eviction rate, data access frequency and data activity.

[0080] According to an embodiment of the present invention, the cache hit rate, cache eviction rate, data access frequency and data activity of the database system are used as data access indicators. Since the cache hit rate, cache eviction rate, data access frequency and data activity can comprehensively evaluate the cache performance of the database, the accuracy of the input data of the reinforcement learning model is improved, thereby improving the accuracy of the cache strategy adjustment information and the cache performance of the database.

[0081] According to an embodiment of the present invention, encoding multiple environmental state features to obtain an environmental state vector includes: mapping the multiple environmental state features into one or more feature vectors; and concatenating the one or more feature vectors to obtain the environmental state vector.

[0082] According to an embodiment of the present invention, multiple environmental state features are converted into digital format, i.e., feature vectors. One or more feature vectors obtained by mapping are simply spliced ​​to obtain a feature vector. In the case of obtaining multiple feature vectors, the multiple feature vectors can be set in the same dimension to facilitate splicing of multiple environmental state vectors to ensure that the environmental state vector can fully express important information about the environment and can effectively support the learning and decision-making of the reinforcement learning model.

[0083] According to an embodiment of the present invention, an environment state vector is input into a reinforcement learning model, and cache policy adjustment information for a target time period is output, including: determining one or more target policy adjustment actions from multiple policy adjustment actions based on the environment state vector; and determining cache policy adjustment information based on the one or more target policy adjustment actions.

[0084] According to an embodiment of the present invention, a policy adjustment action is used to adjust the cache policy. The policy adjustment action can be set based on expert experience and can include all actions previously performed to improve database cache performance. The policy adjustment action can include policy adjustment actions for data objects and policy adjustment actions for non-data objects, wherein the policy adjustment action for data objects can include increasing the priority of cached data, reducing the priority of cached data, replacing data in cache space, increasing the expiration time of data objects, etc. The policy adjustment action for non-data objects can include increasing the amount of storable resources in the cache space, reducing the amount of storable resources in the cache space, etc.

[0085] According to an embodiment of the present invention, the reinforcement learning model determines one or more target policy adjustment actions for non-data objects and one or more target policy adjustment actions for data objects based on the input environment state vector. The cache policy adjustment information is determined based on the identification of the target policy adjustment action for the data object, the identification of the data object, and the identification of the target policy adjustment action for the non-data object.

[0086] For example, when the environment state vector represents that multiple data access requests frequently access the same data set in a short period of time, the target policy adjustment action obtained by the reinforcement learning model may include increasing the priority of the data in the data set and increasing the amount of storage resources in the cache space.

[0087] For example, when the environment state vector represents that the access mode of some data objects changes from random access to sequential access, the target policy adjustment action obtained by the reinforcement learning model may include increasing the expiration time of the data objects.

[0088] For example, when the environment state vector represents that the data access frequency of some data objects is reduced, the target policy adjustment action obtained by the reinforcement learning model may include removing the part of data objects from the cache space and shortening the expiration time of the data objects, wherein increasing or shortening the expiration time of the data objects may be for all data objects stored in the cache space.

[0089] According to an embodiment of the present invention, the reinforcement learning model determines one or more target policy adjustment actions from multiple policy adjustment actions based on the environment state vector, determines cache policy adjustment information based on the one or more target policy adjustment actions, strictly regulates the output of the reinforcement learning model, and thus reduces the error rate of the cache policy adjustment information.

[0090] According to an embodiment of the present invention, the database cache strategy adjustment method also includes: determining cache performance information of the target time period based on data access information of the historical time period and data access information of the target time period; determining load change information of the target time period based on system load information of the historical time period and system load information of the target time period; determining multiple feedback signals based on the cache performance information of the target time period and the load change information of the target time period; and adjusting parameters of the reinforcement learning model based on the multiple feedback signals to obtain a new reinforcement learning model.

[0091] According to an embodiment of the present invention, cache performance information is used to characterize the change threshold of cache performance indicators of the database cache space between the historical time period and the target time period, and the cache performance indicators may include cache hit rate, cache eviction rate, etc. Load change information is used to characterize the difference between the system load performance indicators between the historical time period and the target time period, wherein the system load performance indicators may include memory occupancy rate, processor usage rate, and data read and write usage rate, etc. For each cache performance indicator, the data access information of the historical time period corresponding to the cache performance indicator is compared with the data access information of the target time period to determine the cache performance information of the target time period. For each system load performance indicator, the system load information of the historical time period is compared with the system load information of the target time period to determine the load change information of the target time period.

[0092] According to an embodiment of the present invention, the feedback signal is used to provide feedback to the reinforcement learning model based on the effect of the cache strategy adjustment information output by the reinforcement learning model to adjust the parameters of the reinforcement learning model. The feedback signal may include a positive feedback signal and a negative feedback signal. The positive feedback signal is used to strengthen the reinforcement learning model based on the environment state vector of the target state, and output the cache strategy adjustment information for the target time period.

[0093] According to an embodiment of the present invention, when the cache performance information of the target time period indicates that the database cache performance has decreased, the feedback signal is a negative feedback signal. When the cache performance information of the target time period indicates that the database cache performance has improved, the feedback signal is a positive feedback signal. When the load change information of the target time period indicates that the system load performance of the database has decreased, the feedback signal is a negative feedback signal. When the load change information of the target time period indicates that the system load performance of the database has improved, the feedback signal is a positive feedback signal.

[0094] According to an embodiment of the present invention, for different cache performance information or load change information, the feedback signal can be represented by a numerical value. When the cache hit rate increases in the target time period, the value of the feedback signal can be determined based on one or more of the data object size, access frequency, and data object priority. For example, when the resource utilization rate in the target time period decreases by 3% to 5%, the feedback signal is 2; correspondingly, when the resource utilization rate in the target time period decreases by 3% to 5%, the feedback signal is -2. For example, when the average response time in the target time period decreases by more than 10%, the feedback signal is 2.

[0095] According to an embodiment of the present invention, the detection tool of the database system can also be used to detect the system load of the database system. For example, when it is detected that the processor usage rate suddenly increases, that is, the increase in the processor usage rate is greater than the preset increase threshold, the reinforcement learning model is negatively feedback adjusted according to the percentage of the processor usage rate increase exceeding the preset increase threshold; or when it is detected that the memory usage rate increases by more than the preset increase threshold, the reinforcement learning model is negatively feedback adjusted according to the percentage of the memory usage rate increase exceeding the preset increase threshold, wherein the increase can be calculated based on the ratio of the increased usage rate to the time. For another example, when the percentage of the preset increase threshold exceeds 10%, the feedback signal is -20.

[0096] According to an embodiment of the present invention, determining the feedback signal of the reinforcement learning model through the cache performance information and load change information of the target time period helps to help the reinforcement learning model to adjust the model parameters in time to optimize the model performance, and improve the accuracy of the cache strategy adjustment information, thereby improving the cache performance of the database.

[0097] According to an embodiment of the present invention, the adjustment of the database cache strategy can be triggered by a specific event or by a fixed time node in each time period. The form of periodic triggering helps to keep the cache data up to date and reduce the probability of data inconsistency caused by data expiration or invalidation.

[0098] According to an embodiment of the present invention, by setting up a performance evaluation module, the detected system load information and data access information are statistically analyzed. For example, curve graphs of multiple performance indicators such as memory usage are generated. Database administrators or operation and maintenance personnel can improve the database based on the performance analysis results to further improve the database's cache performance.

[0099] Based on the above database cache strategy adjustment method, the present invention also provides a database cache strategy adjustment device. Figure 5 The device is described in detail.

[0100] Figure 5 A structural block diagram of a database cache strategy adjustment device according to an embodiment of the present invention is shown.

[0101] like Figure 5 As shown, the database cache strategy adjustment device 500 of this embodiment includes an information prediction module 510 , a vector generation module 520 , a strategy determination module 530 and a strategy adjustment module 540 .

[0102] The information prediction module 510 is used to predict the data access situation of the database system in the target time period based on the data access information of the database system in the historical time period, and obtain the data access information of the target time period. In one embodiment, the information prediction module 510 can be used to perform the operation S210 described above, which will not be repeated here.

[0103] The vector generation module 520 is used to generate an environment state vector of the target time period based on the data access information of the target time period, the data access information of the historical time period, and the system load information of the historical time period. In one embodiment, the vector generation module 520 can be used to perform the operation S220 described above, which will not be repeated here.

[0104] The strategy determination module 530 is used to input the environment state vector into the reinforcement learning model and output the cache strategy adjustment information of the target time period. In one embodiment, the strategy determination module 530 can be used to perform the operation S230 described above, which will not be repeated here.

[0105] The policy adjustment module 540 is used to adjust the cache policy of the database cache space based on the cache policy adjustment information of the target time period. In one embodiment, the policy adjustment module 540 can be used to perform the operation S240 described above, which will not be described in detail here.

[0106] According to an embodiment of the present invention, the vector generation module 520 includes: a feature determination submodule and a vector determination submodule.

[0107] The feature determination submodule is used to extract features from the data access information of the target time period, the data access information of the historical time period, and the system load information of the historical time period to obtain multiple environmental state features.

[0108] The vector determination submodule is used to encode multiple environmental state features to obtain an environmental state vector.

[0109] According to an embodiment of the present invention, the feature determination submodule includes: a first determination unit, a second determination unit, a third determination unit, a fourth determination unit, and a feature determination unit.

[0110] The first determining unit is used to determine target indicator data and historical indicator data of each of the plurality of data access indicators based on the data access information of the target time period and the data access information of the historical time period.

[0111] The second determining unit is used to determine at least one target access indicator from the multiple data access indicators based on the comparison result between the target indicator data and the historical indicator data of each of the multiple data access indicators.

[0112] The third determining unit is used to determine the historical indicator data of each of the plurality of system load indicators based on the system load information of the historical time period.

[0113] The fourth determining unit is used to determine at least one target load indicator from the multiple system load indicators based on the comparison result of the historical indicator data of each of the multiple system load indicators.

[0114] The feature determination unit is used to extract features from target indicator data and historical indicator data of at least one target access indicator and historical load data of at least one target load indicator to obtain multiple environmental state features.

[0115] According to an embodiment of the present invention, the second determination unit includes: an indicator determination subunit.

[0116] The indicator determination subunit is used to determine, for each data access indicator, the data access indicator as a target access indicator when the comparison result of the data access indicator indicates that the change range of the data access indicator between the target time period and the historical time period is greater than a preset threshold.

[0117] According to an embodiment of the present invention, the data access indicators include: cache hit rate, cache eviction rate, data access frequency and data activity.

[0118] Among them, the cache hit rate, cache eviction rate, data access frequency and data activity are detected by the detection tool of the database system.

[0119] According to an embodiment of the present invention, the vector determination submodule includes: a feature mapping unit and a vector splicing unit.

[0120] The feature mapping unit is used to map multiple environmental state features into one or more feature vectors.

[0121] The vector concatenation unit is used to concatenate one or more feature vectors to obtain an environment state vector.

[0122] According to the embodiment of the present invention, the policy determination module 530 includes: an action determination submodule and an information determination submodule.

[0123] The action determination submodule is used to determine one or more target policy adjustment actions from multiple policy adjustment actions based on the environment state vector.

[0124] The information determination submodule is used to determine cache policy adjustment information based on one or more target policy adjustment actions.

[0125] According to an embodiment of the present invention, the database cache strategy adjustment device further includes: a performance determination module, a load determination module, a signal determination module, and a model determination module.

[0126] The performance determination module is used to determine the cache performance information of the target time period based on the data access information of the historical time period and the data access information of the target time period.

[0127] The load determination module is used to determine the load change information of the target time period based on the system load information of the historical time period and the system load information of the target time period.

[0128] The signal determination module is used to determine a plurality of feedback signals based on the cache performance information of the target time period and the load change information of the target time period.

[0129] The model determination module is used to adjust the parameters of the reinforcement learning model based on multiple feedback signals to obtain a new reinforcement learning model.

[0130] According to an embodiment of the present invention, any multiple modules of the information prediction module 510, the vector generation module 520, the strategy determination module 530 and the strategy adjustment module 540 can be combined into one module for implementation, or any one of the modules can be split into multiple modules. Alternatively, at least part of the functions of one or more of these modules can be combined with at least part of the functions of other modules and implemented in one module. According to an embodiment of the present invention, at least one of the information prediction module 510, the vector generation module 520, the strategy determination module 530 and the strategy adjustment module 540 can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application specific integrated circuit (ASIC), or can be implemented by hardware or firmware such as any other reasonable way of integrating or packaging the circuit, or implemented in any one of the three implementation methods of software, hardware and firmware or in a suitable combination of any of them. Alternatively, at least one of the information prediction module 510 , the vector generation module 520 , the strategy determination module 530 , and the strategy adjustment module 540 may be at least partially implemented as a computer program module, and when the computer program module is executed, a corresponding function may be performed.

[0131] Figure 6A block diagram of an electronic device suitable for implementing a database cache strategy adjustment method according to an embodiment of the present invention is shown.

[0132] like Figure 6 As shown, the electronic device 600 according to an embodiment of the present invention includes a processor 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage part 608 to a random access memory (RAM) 603. The processor 601 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or a related chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 601 may also include an onboard memory for caching purposes. The processor 601 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present invention.

[0133] In RAM 603, various programs and data required for the operation of electronic device 600 are stored. Processor 601, ROM 602 and RAM 603 are connected to each other via bus 604. Processor 601 performs various operations of the method flow according to the embodiment of the present invention by executing the program in ROM 602 and / or RAM 603. It should be noted that the program can also be stored in one or more memories other than ROM 602 and RAM 603. Processor 601 can also perform various operations of the method flow according to the embodiment of the present invention by executing the program stored in the one or more memories.

[0134] According to an embodiment of the present invention, the electronic device 600 may further include an input / output (I / O) interface 605, which is also connected to the bus 604. The electronic device 600 may further include one or more of the following components connected to the input / output (I / O) interface 605: an input portion 606 including a keyboard, a mouse, etc.; an output portion 607 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage portion 608 including a hard disk, etc.; and a communication portion 609 including a network interface card such as a LAN card, a modem, etc. The communication portion 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to the input / output (I / O) interface 605 as needed. A removable medium 611, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 610 as needed, so that a computer program read therefrom is installed into the storage portion 608 as needed.

[0135] The present invention also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiment; or may exist independently without being assembled into the device / apparatus / system. The above computer-readable storage medium carries one or more programs, and when the above one or more programs are executed, the method according to the embodiment of the present invention is implemented.

[0136] According to an embodiment of the present invention, the computer-readable storage medium may be a non-volatile computer-readable storage medium, for example, it may include but is not limited to: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), 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, the computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, an apparatus or a device. For example, according to an embodiment of the present invention, the computer-readable storage medium may include the ROM602 and / or RAM603 described above and / or one or more memories other than ROM602 and RAM603.

[0137] The embodiment of the present invention also includes a computer program product, which includes a computer program, and the computer program contains program code for executing the method shown in the flowchart. When the computer program product is run in a computer system, the program code is used to enable the computer system to implement the database cache strategy adjustment method provided by the embodiment of the present invention.

[0138] The computer program executes the above functions defined in the system / device of the embodiment of the present invention when it is executed by the processor 601. According to the embodiment of the present invention, the system, device, module, unit, etc. described above can be implemented by a computer program module.

[0139] In one embodiment, the computer program may rely on tangible storage media such as optical storage devices, magnetic storage devices, etc. In another embodiment, the computer program may also be transmitted and distributed in the form of signals on a network medium, and downloaded and installed through the communication part 609, and / or installed from a removable medium 611. The program code contained in the computer program may be transmitted using any appropriate network medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0140] In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 609, and / or installed from the removable medium 611. When the computer program is executed by the processor 601, the above functions defined in the system of the embodiment of the present invention are performed. According to the embodiment of the present invention, the system, device, means, module, unit, etc. described above can be implemented by a computer program module.

[0141] According to an embodiment of the present invention, the program code for executing the computer program provided by the embodiment of the present invention can be written in any combination of one or more programming languages. Specifically, these computing programs can be implemented using high-level process and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, Java, C++, python, "C" language or similar programming languages. The program code can be executed entirely on the user computing device, partially on the user device, partially on the remote computing device, or entirely on the remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (e.g., using an Internet service provider to connect through the Internet).

[0142] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each box in the flow chart or block diagram can 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 realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0143] It will be appreciated by those skilled in the art that the features described in the various embodiments of the present invention may be combined and / or combined in various ways, even if such combinations and / or combinations are not explicitly described in the present invention. In particular, without departing from the spirit and teachings of the present invention, the features described in the various embodiments of the present invention may be combined and / or combined in various ways. All of these combinations and / or combinations fall within the scope of the present invention.

[0144] The embodiments of the present invention are described above. However, these embodiments are only for the purpose of illustration, and are not intended to limit the scope of the present invention. Although each embodiment is described above, it does not mean that the measures in each embodiment cannot be used in combination advantageously. Without departing from the scope of the present invention, those skilled in the art may make various substitutions and modifications, which should all fall within the scope of the present invention.

Claims

1. A database cache strategy adjustment method, characterized in that: The method comprises: Based on the data access information of the database system in the historical time period, predict the data access situation of the database system in the target time period to obtain the data access information of the target time period; Generate an environment state vector for the target time period based on the data access information for the target time period, the data access information for the historical time period, and the system load information for the historical time period; Inputting the environment state vector into a reinforcement learning model, and outputting cache strategy adjustment information for a target time period; Adjusting the cache strategy of the database cache space based on the cache strategy adjustment information of the target time period; The step of generating the environment state vector of the target time period based on the data access information of the target time period, the data access information of the historical time period, and the system load information of the historical time period includes: Extracting features from the data access information of the target time period, the data access information of the historical time period, and the system load information of the historical time period to obtain a plurality of environmental state features; Encoding the multiple environmental state features to obtain the environmental state vector; The feature extraction is performed on the data access information of the target time period, the data access information of the historical time period, and the system load information of the historical time period to obtain a plurality of environmental status features, including: Determining target indicator data and historical indicator data of each of a plurality of data access indicators based on the data access information of the target time period and the data access information of the historical time period; Determine at least one target access indicator from the plurality of data access indicators based on a comparison result between the target indicator data and the historical indicator data of each of the plurality of data access indicators; Determining historical indicator data of each of the plurality of system load indicators based on the system load information of the historical time period; Determining at least one target load indicator from the plurality of system load indicators based on a comparison result of the respective historical indicator data of the plurality of system load indicators; Performing feature extraction on the target indicator data and the historical indicator data of each of the at least one target access indicator and the historical load data of each of the at least one target load indicator to obtain the plurality of environmental state features; The determining at least one target access indicator from the plurality of data access indicators based on the comparison result between the target indicator data and the historical indicator data of each of the plurality of data access indicators comprises: For each of the data access indicators, when the comparison result of the data access indicators indicates that the change range of the data access indicator between the target time period and the historical time period is greater than a preset threshold, the data access indicator is determined as a target access indicator; The determining at least one target load indicator from the multiple system load indicators based on the comparison result of the historical indicator data of each of the multiple system load indicators comprises: By comparing the change ranges of the historical load data of each of the plurality of system load indicators, the system load indicator with a larger change range is determined as the target load indicator.

2. The method according to claim 1, characterized in that The data access indicators include: cache hit rate, cache eviction rate, data access frequency and data activity; The cache hit rate, the cache eviction rate, the data access frequency and the data activity are detected by a detection tool of the database system.

3. The method according to claim 1, characterized in that The step of encoding the plurality of environmental state features to obtain an environmental state vector comprises: Mapping the plurality of environmental state features into one or more feature vectors; The one or more feature vectors are concatenated to obtain the environment state vector.

4. The method according to claim 1, characterized in that: The step of inputting the environment state vector into a reinforcement learning model and outputting cache strategy adjustment information for a target time period includes: Determine one or more target policy adjustment actions from a plurality of policy adjustment actions based on the environment state vector; The cache policy adjustment information is determined based on the one or more target policy adjustment actions.

5. The method according to claim 1, characterized in that The method further comprises: Determining cache performance information of the target time period based on the data access information of the historical time period and the data access information of the target time period; Determining load change information for the target time period based on the system load information for the historical time period and the system load information for the target time period; Determining a plurality of feedback signals based on the cache performance information of the target time period and the load change information of the target time period; The parameters of the reinforcement learning model are adjusted based on the multiple feedback signals to obtain a new reinforcement learning model.

6. A database cache strategy adjustment device, characterized in that: The device comprises: An information prediction module, used to predict the data access situation of the database system in a target time period based on the data access information of the database system in a historical time period, and obtain the data access information of the target time period; A vector generation module, configured to generate an environment state vector of a target time period based on the data access information of the target time period, the data access information of the historical time period, and the system load information of the historical time period; A strategy determination module, used to input the environment state vector into a reinforcement learning model and output cache strategy adjustment information for a target time period; A policy adjustment module, used to adjust the cache policy of the database cache space based on the cache policy adjustment information of the target time period; Wherein, the vector generation module includes: A feature determination submodule, used for extracting features from the data access information of the target time period, the data access information of the historical time period, and the system load information of the historical time period to obtain a plurality of environmental state features; A vector determination submodule, used for encoding the plurality of environmental state features to obtain the environmental state vector; The feature determination submodule comprises: A first determining unit, configured to determine target indicator data and historical indicator data of each of a plurality of data access indicators based on the data access information of the target time period and the data access information of the historical time period; a second determining unit, configured to determine at least one target access indicator from the plurality of data access indicators based on a comparison result between the target indicator data and the historical indicator data of each of the plurality of data access indicators; A third determining unit, configured to determine historical indicator data of each of the plurality of system load indicators based on the system load information of the historical time period; a fourth determining unit, configured to determine at least one target load indicator from the plurality of system load indicators based on a comparison result of the respective historical indicator data of the plurality of system load indicators; A feature determination unit, configured to extract features from the target indicator data and the historical indicator data of each of the at least one target access indicator and the historical load data of each of the at least one target load indicator to obtain the plurality of environmental state features; The second determining unit includes: an indicator determination subunit, for each of the data access indicators, when a comparison result of the data access indicators indicates that a change range of the data access indicator between the target time period and the historical time period is greater than a preset threshold, determining the data access indicator as a target access indicator; The fourth determination unit is further configured to determine the system load indicator with a larger change range as the target load indicator by comparing the change ranges of the historical load data of each of the plurality of system load indicators.

7. An electronic device, characterized in that: include: one or more processors; a memory for storing one or more computer programs, It is characterized in that the one or more processors execute the one or more computer programs to implement the steps of the method according to any one of claims 1 to 5.

Images