A method and system for efficient processing of cloud data

Abstract

The present invention belongs to the technical field of cloud computing. A method and system for efficient processing of cloud data are provided. The method comprises: receiving a first data behavior feature of each access terminal in a first period, using a first prediction model to predict a second data behavior feature of the corresponding access terminal in a second period, and estimating the total data processing load; when the total data processing load is higher than the data processing load threshold of the cloud, determining and grouping each access terminal according to the response timeliness requirement level of each access terminal; in the second period, in the order of data processing priority from high to low, performing data synchronization processing on each access terminal in each group. The present invention performs group processing on the access terminals, and decides on different data processing priorities according to the grouping results, thereby avoiding the collapse of the cloud, and can also minimize the risks brought to the access terminals by different processing orders.

Description

Technical Field

[0001] The present invention relates to the field of cloud computing technology, and in particular to a method and system for efficiently processing cloud data. Background Art

[0002] Cloud computing technology can transfer the computing, processing, and storage of large amounts of data from local devices to large-scale server clusters on the Internet, which can reduce the stringent requirements for local hardware performance. However, when processing large-scale data, the cloud may still have insufficient computing resources, resulting in low efficiency in processing large-scale data. In this way, there will be significant response delays on the user side, and the user experience will be reduced.

[0003] Therefore, how to process large-scale data in the cloud is a technical problem that needs to be solved urgently. Summary of the invention

[0004] In response to the above technical problems, the present invention provides a method, system, electronic device, computer storage medium and computer program product for efficient processing of cloud data.

[0005] The present invention discloses a method for efficiently processing cloud data, which is applied to the cloud. The method comprises the following steps:

[0006] receiving a first data behavior feature of each access terminal in a first period, performing prediction analysis on the first data behavior feature using a first prediction model, obtaining a second data behavior feature of the corresponding access terminal in a second period, and estimating a total data processing load according to each of the second data behavior features;

[0007] When the total data processing load is higher than the data processing load threshold of the cloud, determining the response timeliness requirement level of each access terminal, and grouping the access terminals according to the response timeliness requirement level, with different data processing priorities for each group;

[0008] In the second period, data synchronization processing is performed on each access terminal in each group in the order of the data processing priority from high to low.

[0009] Optionally, the receiving a first data behavior feature of each access terminal in a first period includes:

[0010] Receiving a second data behavior characteristic of each access terminal in a first period; wherein the first period is a preset value;

[0011] Determine terminal attribute information of each access terminal, and classify the corresponding access terminal as an associated terminal or an independent terminal according to the terminal attribute information;

[0012] If the access terminal belongs to the associated terminal, optimizing the second data behavior feature to the first data behavior feature based on the first strength; if the access terminal belongs to the independent terminal, optimizing the second data behavior feature to the first data behavior feature based on the second strength;

[0013] Wherein, the first intensity is greater than the second intensity.

[0014] Optionally, determining the response timeliness requirement level of each access terminal includes:

[0015] Acquire a third data behavior feature of each access terminal during a data group processing stage in the cloud, wherein the third data behavior feature includes the number and frequency of repeated data request clicks and the number and frequency of page refreshes;

[0016] The third data behavior characteristics are evaluated and analyzed using a second prediction model to obtain the predicted response timeliness requirement level of the corresponding access terminal; wherein the second prediction model is constructed based on Transformer.

[0017] Optionally, performing data synchronization processing on each access terminal in each group in the order of the data processing priority from high to low further includes:

[0018] The new total data processing load in the second cycle is calculated according to the third cycle. When the new total data processing load is lower than the data processing load threshold and reaches the preset data processing load, the data grouping processing is released, that is, data processing is performed in the order of data requests from each access terminal.

[0019] Optionally, the preset data processing load is determined by:

[0020] Calculate the data processing load difference between the total data processing load and the new total data processing load when it is lower than the data processing load threshold value; when the data processing load difference is larger, set the preset data processing load to be larger; when the data processing load difference is smaller, set the preset data processing load to be smaller.

[0021] The present invention also discloses a system for efficient processing of cloud data, which is applied to the cloud. The system includes a processing device and a storage device. The computer code stored in the storage device is called and executed by the processing device to implement the following steps:

[0022] receiving a first data behavior feature of each access terminal in a first period, performing prediction analysis on the first data behavior feature using a first prediction model, obtaining a second data behavior feature of the corresponding access terminal in a second period, and estimating a total data processing load according to each of the second data behavior features;

[0023] When the total data processing load is higher than the data processing load threshold of the cloud, determining the response timeliness requirement level of each access terminal, and grouping the access terminals according to the response timeliness requirement level, with different data processing priorities for each group;

[0024] In the second period, data synchronization processing is performed on each access terminal in each group in the order of the data processing priority from high to low.

[0025] Optionally, the receiving a first data behavior feature of each access terminal in a first period includes:

[0026] Receiving a second data behavior characteristic of each access terminal in a first period; wherein the first period is a preset value;

[0027] Determine terminal attribute information of each access terminal, and classify the corresponding access terminal as an associated terminal or an independent terminal according to the terminal attribute information;

[0028] If the access terminal belongs to the associated terminal, optimizing the second data behavior feature to the first data behavior feature based on the first strength; if the access terminal belongs to the independent terminal, optimizing the second data behavior feature to the first data behavior feature based on the second strength;

[0029] Wherein, the first intensity is greater than the second intensity.

[0030] Optionally, determining the response timeliness requirement level of each access terminal includes:

[0031] Acquire a third data behavior feature of each access terminal during a data group processing stage in the cloud, wherein the third data behavior feature includes the number and frequency of repeated data request clicks and the number and frequency of page refreshes;

[0032] The third data behavior characteristics are evaluated and analyzed using a second prediction model to obtain the predicted response timeliness requirement level of the corresponding access terminal; wherein the second prediction model is constructed based on Transformer.

[0033] Optionally, performing data synchronization processing on each access terminal in each group in the order of the data processing priority from high to low further includes:

[0034] The new total data processing load in the second cycle is calculated according to the third cycle. When the new total data processing load is lower than the data processing load threshold and reaches the preset data processing load, the data grouping processing is released, that is, data processing is performed in the order of data requests from each access terminal.

[0035] Optionally, the preset data processing load is determined by:

[0036] Calculate the data processing load difference between the total data processing load and the new total data processing load when it is lower than the data processing load threshold value; when the data processing load difference is larger, set the preset data processing load to be larger; when the data processing load difference is smaller, set the preset data processing load to be smaller.

[0037] The present invention also discloses an electronic device, comprising: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor executes the computer program to implement any of the above methods.

[0038] The present invention also discloses a computer storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program is executed by a processor to implement any of the above methods.

[0039] The present invention also discloses a computer program product, which includes computer codes. When the computer codes are executed by a processor of an electronic device, any of the above methods is implemented.

[0040] The beneficial effects of the present invention are at least:

[0041] The present invention can predict in advance whether the data processing load in the future period exceeds the processing capacity of the cloud, and adopts group processing for the access end, and decides different data processing priorities according to the grouping results, so as to avoid the cloud crash, and can also minimize the risks brought to the access end by different processing orders. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0043] Figure 1 It is a flowchart of a method for efficient processing of cloud data disclosed in an embodiment of the present invention;

[0044] Figure 2 It is a structural schematic diagram of a cloud data efficient processing system disclosed in an embodiment of the present invention;

[0045] Figure 3It is a schematic diagram of the internal structure of the processing device disclosed in the embodiment of the present invention. DETAILED DESCRIPTION

[0046] The following is a description of the implementation of the present application by specific specific embodiments. People familiar with the technology can easily understand other advantages and effects of the present application from the contents disclosed in this specification. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0047] In addition, the technical features involved in the different embodiments of the present application described below can be combined with each other as long as they do not conflict with each other.

[0048] like Figure 1 As shown, in view of the above technical problems, an embodiment of the present invention discloses a method for efficient processing of cloud data, which is applied to the cloud. The method comprises the following steps:

[0049] Receive the first data behavior characteristics of each access terminal in the first period, use the first prediction model to predict and analyze the first data behavior characteristics, obtain the second data behavior characteristics of the corresponding access terminal in the second period, and estimate the total data processing load based on each of the second data behavior characteristics.

[0050] In this step, the access terminal can be various networked smart devices, such as smart phones, smart sensors, etc. In a specific first period, such as the past hour, the cloud receives operation records, data transmission frequency, etc. sent by these access terminals and uses them as the first data behavior feature. The first prediction model is a machine learning model trained based on a large amount of historical data, such as a model based on a recurrent neural network (RNN).

[0051] Taking the access end of the e-commerce platform as an example, the frequency of users browsing product pages and the number of times they added products to the shopping cart in the past hour are collected and input into the first prediction model mentioned above. After complex internal calculations, the model predicts the user's possible behavior in the next half hour (i.e., the second cycle), such as whether the user will frequently modify the shopping cart items, whether the user will initiate payment, and other second data behavior characteristics. By combining the prediction results of all access ends, the total data processing load that the entire cloud will bear in the next hour can be estimated, for example, it is expected to process 100,000 data interactions.

[0052] When the total data processing load is higher than the data processing load threshold of the cloud, the response timeliness requirement level of each access terminal is determined, and the access terminals are grouped according to the response timeliness requirement level, and the data processing priority of each group is different.

[0053] In this step, when the estimated total load exceeds the preset threshold, for example, in the above example, if the maximum processing capacity of the cloud is 80,000 data interactions (excluding some reserved data processing capacity), 100,000 data interactions exceed the processing capacity of the cloud, and the conventional load balancing method cannot effectively cope with it. In view of the occurrence of this situation, the present invention is set to classify and group each access terminal, and different groups have different data processing priorities. Taking the Internet of Vehicles as an example, the sensor of the automatic / assisted driving vehicle is used as the access terminal. It transmits road condition data in real time. A slight delay may cause an accident. The response timeliness requirement level of this type of access terminal is extremely high; while the access terminal of the in-car entertainment system can be accepted by users for a slight delay of a few seconds in playing music, and its response timeliness requirement level is low. These access terminals are grouped into different levels, such as high, medium, and low. The access terminals related to automatic / assisted driving are in the high priority group, and the access terminals of the entertainment system are in the low priority group.

[0054] In the second period, data synchronization processing is performed on each access terminal in each group in the order of the data processing priority from high to low.

[0055] In this step, after entering the second cycle, that is, the period when the prediction takes effect, the cloud first processes the access end data of the high-priority group. For example, in the Internet of Vehicles scenario, the synchronous processing of the autonomous driving sensor data is first guaranteed so that the vehicle can accurately obtain the latest road conditions; after the high-priority tasks are processed, the medium-priority tasks, such as the in-vehicle navigation update, are processed; and finally, the low-priority entertainment system update playlist data is processed.

[0056] The present invention can accurately predict future loads based on the past data behavior of the access end and plan the processing order in advance. When dealing with high-load scenarios, the real-time response of key services is guaranteed by scientifically dividing the response timeliness requirements of the access end. For example, in fields such as smart transportation and industrial Internet that have strict requirements on the timeliness of data processing, system freezes or even crashes caused by data congestion can be avoided, greatly improving the stability and reliability of the entire cloud system, reducing various potential losses caused by delays, and optimizing resource allocation to improve the overall efficiency of data processing.

[0057] Optionally, the receiving a first data behavior feature of each access terminal in a first period includes:

[0058] Receiving a second data behavior characteristic of each access terminal in a first period; wherein the first period is a preset value;

[0059] Determine terminal attribute information of each access terminal, and classify the corresponding access terminal as an associated terminal or an independent terminal according to the terminal attribute information;

[0060] If the access terminal belongs to the associated terminal, optimizing the second data behavior feature to the first data behavior feature based on the first strength; if the access terminal belongs to the independent terminal, optimizing the second data behavior feature to the first data behavior feature based on the second strength;

[0061] Wherein, the first intensity is greater than the second intensity.

[0062] In this embodiment, the first period can be set to a specific time range, such as one day, one week, or one month. Taking the cloud service of an e-commerce platform as an example, during this period, the second data behavior characteristics will be received from various access terminals (such as mobile shopping apps of different users). These second data behavior characteristics may include operation information such as the user's browsing time, search keywords, and frequency of clicking on products. Assuming that the first period is one day, then during this day, the cloud will continuously receive these operation information from each user's mobile shopping app, forming a massive data set.

[0063] Terminal attribute information includes information on the access terminal's business type (such as shopping, navigation, etc.), usage scenarios, and other aspects. For shopping apps, their business is greatly affected by external factors such as promotional activities and seasonal changes. For example, in the early stage of the "Double 11" promotion, users will frequently browse products, add shopping carts, and pay attention to discount information. After the promotion begins, purchases, payments, and other operations will increase sharply. This type of terminal is an associated terminal, that is, its data behavior will be greatly affected by certain external factors; while the terminal of the automatic / assisted driving system has relatively stable data behavior characteristics, mainly continuously sending information such as the vehicle's speed, location, and road conditions detected by the sensor, and the data fluctuation is small. This type of terminal is an independent terminal, that is, its data behavior is basically not greatly affected by certain external factors. According to these characteristics, the cloud can classify the access terminal, classifying shopping apps as associated terminals and terminals of automatic / assisted driving systems as independent terminals.

[0064] For different types of terminals, the second data behavior characteristics acquired in the first period are enhanced to a suitable degree to make them closer to the actual situation in the second period. Specifically, for associated terminals, the corresponding enhancement degree is set to the first intensity, while for independent terminals, the corresponding enhancement degree is set to the smaller second intensity.

[0065] By distinguishing different types of access terminals (associated terminals, independent terminals) and adopting optimization methods of different intensities, the cloud can predict the total data processing load with higher accuracy.

[0066] It should be noted that the first intensity and the second intensity refer to the degree of expansion of data behaviors such as data access volume and data access frequency in the second data behavior feature. For example, the data access volume in the second data behavior feature is 30 times per minute, and the first intensity is used to optimize it to 50 times per minute, and the second intensity is used to optimize it to 40 times per minute. The above is only an example, and the technical solution used to limit the present invention can only be implemented based on the above example.

[0067] Optionally, determining the response timeliness requirement level of each access terminal includes:

[0068] Acquire a third data behavior feature of each access terminal during a data group processing stage in the cloud, wherein the third data behavior feature includes the number and frequency of repeated data request clicks and the number and frequency of page refreshes;

[0069] The third data behavior characteristics are evaluated and analyzed using a second prediction model to obtain the predicted response timeliness requirement level of the corresponding access terminal; wherein the second prediction model is constructed based on Transformer.

[0070] In this embodiment, during the data group processing stage in the cloud, the third data behavior characteristics from each access terminal are collected. Taking an online video playback platform as an example, the access terminal can be a user's mobile phone, computer or smart TV and other devices. For these access terminals, the number and frequency of repeated data request clicks by users when using the platform will be counted. For example, when a user clicks the video play button many times but fails to play it successfully, the user will click repeatedly, and the number and frequency of clicks will be recorded; at the same time, the number and frequency of page refreshes will also be counted. For example, the user may frequently refresh the page while waiting for the video to load. These data are the third data behavior characteristics. During peak hours, when the cloud predicts that its total data processing load in the next period is higher than the data processing load threshold, it will turn on the data group processing mode. At this time, some access terminals will experience a certain degree of network jamming, and users of these access terminals will have data behaviors such as repeated data requests and page refreshes.

[0071] Given that Transformer has a powerful ability to process sequence data, it can identify the characteristics of users in data request and page refresh behavior, and then predict the response timeliness requirement level of the access terminal. Therefore, the present invention constructs a second prediction model based on the Transformer architecture. The second prediction model analyzes and processes the above-mentioned third data behavior characteristics to complete the evaluation of the response timeliness requirement level of the corresponding access terminal. The higher the response timeliness requirement level, the more the user of the corresponding access terminal needs the cloud to prioritize its data request.

[0072] For the example of the online video playback platform above, the collected third data behavior features of the user are input into the second prediction model, and the second prediction model will perform complex feature extraction and analysis on these data, taking into account information such as the pattern of repeated data request clicks and the frequency trend of page refreshes. For example, if a user's repeated data request clicks and page refreshes are both very high, it may indicate that the user has high requirements for the timeliness of the current service, and the model may predict the response timeliness requirement level of the access terminal as "high"; while for users with low repeated data request clicks and page refreshes, the response timeliness requirement level is predicted to be "low".

[0073] Optionally, performing data synchronization processing on each access terminal in each group in the order of the data processing priority from high to low further includes:

[0074] The new total data processing load in the second cycle is calculated according to the third cycle. When the new total data processing load is lower than the data processing load threshold and reaches the preset data processing load, the data grouping processing is released, that is, data processing is performed in the order of data requests from each access terminal.

[0075] In this embodiment, in the group processing mode, the present invention also sets a periodic monitoring of the new total data processing load of the cloud. The monitoring period is the third period mentioned above. The duration of the third period can be determined according to the total data processing load, for example, 5 minutes or 10 minutes. The higher the total data processing load, the shorter the third period is set. In this way, the new total data processing load of the cloud can be monitored at a higher frequency to see if it is lower than the data processing load threshold, so as to release the group processing mode as soon as possible; otherwise, the third period is set longer.

[0076] When it is monitored that the real-time predicted total data processing load is lower than the data processing load threshold, it means that the current data processing demand is lower than the data processing capacity of the cloud, and the group processing mode can be considered to be released, which can improve the data response speed of all access terminals. However, due to the existence of data fluctuations, the situation where the total data processing load is lower than the data processing load threshold may only be temporary, and it may soon recover to a state higher than the data processing load threshold. In this regard, the present invention is arranged to release the data group processing only when the new total data processing load is lower than the data processing load threshold, and the degree of lower than reaches the preset data processing load.

[0077] Optionally, the preset data processing load is determined by:

[0078] Calculate the data processing load difference between the total data processing load and the new total data processing load when it is lower than the data processing load threshold value; when the data processing load difference is larger, set the preset data processing load to be larger; when the data processing load difference is smaller, set the preset data processing load to be smaller.

[0079] In this embodiment, the size of another threshold value of the preset data processing load is determined by the present invention based on the difference between the new total data processing load obtained based on the current data prediction and the total data processing load obtained based on the data prediction in the first cycle. Specifically, when the difference in the above data processing load is larger, it indicates that the gap between the earlier prediction and the current latest prediction is larger, indicating that the data based on the current latest prediction contains more "unexpected" data requests, which means that the volatility of its data requests is greater. At this time, the preset data processing load is set to be larger, which can reduce the impact of the situation that "the situation where the total data processing load is lower than the data processing load threshold may be only temporary, and may soon recover to a state higher than the data processing load threshold later"; and when the difference in the above data processing load is smaller, it indicates that the gap between the earlier prediction and the current latest prediction is smaller, indicating that the data based on the current latest prediction contains fewer "unexpected" data requests, and the new and old predictions are basically consistent, which means that the volatility of its data requests is smaller. At this time, the preset data processing load is set to be smaller.

[0080] A comparison data table between the above-mentioned preset data processing load and the data processing load difference may be established in advance, or a corresponding conversion function may be obtained by fitting, which is not specifically limited in the present invention.

[0081] It should be noted that since the new total data processing load is not the total data processing load within the entire second cycle, but the total data processing load between the "current moment" in the second cycle and the end moment of the second cycle, when calculating the above data processing load difference, the early predicted total data processing load can be appropriately reduced based on the ratio between the above period and the total duration of the second cycle, and then the difference with the new total data processing load is calculated.

[0082] like Figure 2 , Figure 3 As shown, an embodiment of the present invention further discloses a system for efficient processing of cloud data, which is applied to the cloud. The system includes a processing device and a storage device. The computer code stored in the storage device is called and executed by the processing device to implement the following steps:

[0083] receiving a first data behavior feature of each access terminal in a first period, performing prediction analysis on the first data behavior feature using a first prediction model, obtaining a second data behavior feature of the corresponding access terminal in a second period, and estimating a total data processing load according to each of the second data behavior features;

[0084] When the total data processing load is higher than the data processing load threshold of the cloud, determining the response timeliness requirement level of each access terminal, and grouping the access terminals according to the response timeliness requirement level, with different data processing priorities for each group;

[0085] In the second period, data synchronization processing is performed on each access terminal in each group in the order of the data processing priority from high to low.

[0086] An embodiment of the present invention further discloses an electronic device, comprising: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor executes the computer program to implement the method described in the above embodiment.

[0087] An embodiment of the present invention further discloses a computer storage medium, wherein the computer storage medium stores a computer program, and the computer program is executed by a processor to implement the method described in the above embodiment.

[0088] An embodiment of the present invention further discloses a computer program product, which includes computer code. When the computer code is executed by a processor of an electronic device, the method described in the above embodiment is implemented.

[0089] The computer-readable storage medium may be an electronic memory such as a flash memory, an EEPROM (electrically erasable programmable read-only memory), an EPROM, a hard disk, or a ROM. Optionally, the computer-readable storage medium includes a non-transitory computer-readable storage medium. The computer-readable storage medium has storage space for program codes that execute any of the method steps in the above method. These program codes can be read from or written to one or more computer program products. The program code can be compressed, for example, in an appropriate form.

[0090] It should also be noted that, in this specification, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the statement "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or device including the element.

[0091] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for efficient processing of cloud data, applied to the cloud, characterized in that: The method comprises the following steps: receiving a first data behavior feature of each access terminal in a first period, performing prediction analysis on the first data behavior feature using a first prediction model, obtaining a second data behavior feature of the corresponding access terminal in a second period, and estimating a total data processing load according to each of the second data behavior features; When the total data processing load is higher than the data processing load threshold of the cloud, determining the response timeliness requirement level of each access terminal, and grouping the access terminals according to the response timeliness requirement level, with different data processing priorities for each group; In the second period, according to the order of the data processing priority from high to low, the data synchronization processing is performed on each of the access terminals in each group; The receiving of the first data behavior characteristics of each access terminal in the first period includes: Receiving a second data behavior characteristic of each access terminal in a first period; wherein the first period is a preset value; Determine the terminal attribute information of each access terminal, and classify the corresponding access terminal into an associated terminal or an independent terminal according to the terminal attribute information; wherein the associated terminal refers to a terminal whose data behavior is greatly affected by certain external factors, and the independent terminal refers to a terminal whose data behavior is basically not greatly affected by certain external factors; If the access terminal belongs to the associated terminal, optimizing the second data behavior feature to the first data behavior feature based on the first strength; if the access terminal belongs to the independent terminal, optimizing the second data behavior feature to the first data behavior feature based on the second strength; Wherein, the first intensity is greater than the second intensity.

2. The method for efficient cloud data processing according to claim 1, characterized in that: The step of determining the response timeliness requirement level of each access terminal includes: Acquire a third data behavior feature of each access terminal during a data group processing stage in the cloud, wherein the third data behavior feature includes the number and frequency of repeated data request clicks and the number and frequency of page refreshes; The third data behavior characteristics are evaluated and analyzed using a second prediction model to obtain the predicted response timeliness requirement level of the corresponding access terminal; wherein the second prediction model is constructed based on Transformer.

3. The method for efficient cloud data processing according to claim 2, characterized in that: The step of performing data synchronization processing on each access terminal in each group in the order of the data processing priority from high to low also includes: The new total data processing load in the second cycle is calculated according to the third cycle. When the new total data processing load is lower than the data processing load threshold and reaches the preset data processing load, the data grouping processing is released, that is, data processing is performed in the order of data requests from each access terminal.

4. The method for efficient cloud data processing according to claim 3, characterized in that: The preset data processing load is determined by: Calculate the data processing load difference between the total data processing load and the new total data processing load when it is lower than the data processing load threshold value; when the data processing load difference is larger, set the preset data processing load to be larger; when the data processing load difference is smaller, set the preset data processing load to be smaller.

5. A system for efficient processing of cloud data, applied to the cloud, the system comprising a processing device and a storage device, characterized in that: The computer code stored in the storage device is called and executed by the processing device to implement the following steps: receiving a first data behavior feature of each access terminal in a first period, performing prediction analysis on the first data behavior feature using a first prediction model, obtaining a second data behavior feature of the corresponding access terminal in a second period, and estimating a total data processing load according to each of the second data behavior features; When the total data processing load is higher than the data processing load threshold of the cloud, determining the response timeliness requirement level of each access terminal, and grouping the access terminals according to the response timeliness requirement level, with different data processing priorities for each group; In the second period, according to the order of the data processing priority from high to low, the data synchronization processing is performed on each of the access terminals in each group; The receiving of the first data behavior characteristics of each access terminal in the first period includes: Receiving a second data behavior characteristic of each access terminal in a first period; wherein the first period is a preset value; Determine the terminal attribute information of each access terminal, and classify the corresponding access terminal into an associated terminal or an independent terminal according to the terminal attribute information; wherein the associated terminal refers to a terminal whose data behavior is greatly affected by certain external factors, and the independent terminal refers to a terminal whose data behavior is basically not greatly affected by certain external factors; If the access terminal belongs to the associated terminal, optimizing the second data behavior feature to the first data behavior feature based on the first strength; if the access terminal belongs to the independent terminal, optimizing the second data behavior feature to the first data behavior feature based on the second strength; Wherein, the first intensity is greater than the second intensity.

6. An electronic device comprising: At least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor executes the computer program to implement the method according to any one of claims 1 to 4.

7. A computer storage medium storing a computer program, characterized in that: The computer program is executed by a processor to implement the method according to any one of claims 1 to 4.

8. A computer program product, characterized in that: The computer program product includes computer codes, and when the computer codes are executed by a processor of an electronic device, the method according to any one of claims 1 to 4 is implemented.

Images