Disaster recovery response processing method for multiple virtual machines in high-availability cloud computing cluster

By obtaining the current data and status level values ​​of the virtual machine, using probability analysis and priority calculation to determine the target and transfer it to the virtual machine, the problem of low efficiency of traditional disaster recovery response methods is solved, and efficient virtual machine resource utilization and business continuity are achieved. .

CN118916217APending Publication Date: 2024-11-08HUASHU INTELLIGENT RING (BEIJING) DIGITAL TECHNOLOGY CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410999828.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-24
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

Traditional virtual machine disaster recovery response methods are inefficient and cannot be flexibly adjusted, resulting in waste of backup server resources and untimely processing of critical tasks, affecting business continuity and efficiency, and underutilizing modern technologies such as artificial intelligence and data analysis.

Method used

By obtaining the current data list and status level values ​​of the virtual machine, determining the target to transfer out of the virtual machine based on the preset threshold, analyzing the target data list, and using probability analysis, priority calculation and fuzzy influence reference quantity to determine the target to transfer into the virtual machine to ensure the data Access requests are transferred to appropriate virtual machines to improve the accuracy and efficiency of disaster recovery response.

Benefits of technology

It improves the accuracy and efficiency of disaster recovery response, reduces the risks of system interruptions and data loss, ensures system stability and reliability after load transfer, and ensures the continuity of data access and business continuity.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118916217A_ABST
    Figure CN118916217A_ABST
Patent Text Reader

Abstract

The invention discloses a disaster recovery response processing method for multiple virtual machines in a high-availability cloud computing cluster, and the method comprises the steps: obtaining a current data list and a state grade value of each virtual machine, timely knowing the operation state, the load condition and the data condition of each virtual machine, and providing basic data for subsequent decision making; determining the virtual machines needing to be transferred out according to a preset virtual machine state grade numerical value threshold value, obtaining a target data list corresponding to the virtual machines, and identifying the virtual machines needing to be subjected to disaster recovery response processing to make preparations for subsequent load transfer; according to a target data list and a virtual machine state grade value, through analysis of probability analysis, priority calculation sorting, fuzzy influence reference quantity and the like, a target transfer virtual machine is determined, and it is ensured that the transferred target virtual machine can effectively process data and requests; and the data access request corresponding to the target data list is turned to the target transfer-in virtual machine, so that the continuity and stability of data access are ensured.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of disaster recovery response, and in particular to a method for processing disaster recovery response of multiple virtual machines in a high-availability cloud computing cluster. Background Art

[0002] Traditional virtual machine disaster recovery response processing methods may have the problem of low efficiency. This is mainly because traditional methods are often based on fixed rules and policies for disaster recovery response, and cannot be flexibly adjusted according to actual conditions. For example, when a disaster recovery event occurs, traditional methods may simply transfer all tasks to the backup server without considering the urgency of the task or the actual load of the server. This will lead to a waste of backup server resources and fail to ensure the timely processing of critical tasks, thus affecting business continuity and efficiency;

[0003] In addition, traditional methods are often based on static rules and simple priority divisions to handle disaster recovery response, and cannot fully utilize the advantages of modern technologies such as artificial intelligence and data analysis. For example, during the disaster recovery response process, traditional methods may not be able to intelligently adjust the disaster recovery plan according to the urgency and importance of different tasks, thus failing to maximize the processing of critical tasks and business continuity;

[0004] At the same time, when a disaster recovery event occurs, the traditional method is often to transfer all tasks to the backup server, which will lead to excessive resource utilization of the backup server, while the resources of the original server are idle. This underutilization of resources will not only increase costs, but also affect the performance and stability of the overall system, thereby reducing the efficiency and effectiveness of disaster recovery response processing. Summary of the invention

[0005] The purpose of the present invention is to provide a multi-virtual machine disaster recovery response processing method in a high-availability cloud computing cluster, which solves the above-mentioned technical problems pointed out in the prior art.

[0006] The present invention provides a method for disaster recovery response processing of multiple virtual machines in a high-availability cloud computing cluster, comprising the following steps:

[0007] Obtain the current data list and virtual machine status level value of each virtual machine corresponding to the same server cluster;

[0008] Acquire a target transfer-out virtual machine and a target data list corresponding to the target transfer-out virtual machine according to a preset virtual machine status level numerical threshold;

[0009] Analyzing the target data list according to the virtual machine state level value to obtain a target transfer virtual machine corresponding to the target data list;

[0010] The data access request corresponding to the target data list is forwarded to the target transfer-in virtual machine.

[0011] Preferably, the data list is a list of applications running on each virtual machine or stored data.

[0012] Preferably, analyzing the target data list according to the virtual machine state level value to obtain the target transfer virtual machine corresponding to the target data list includes the following steps:

[0013] Acquire virtual machine attribute information corresponding to each of the virtual machines; filter and acquire multiple virtual machines belonging to the same category according to the virtual machine attribute information; and filter and acquire multiple first transferred virtual machines according to the virtual machines belonging to the same category;

[0014] Analyze and obtain the target transfer-in virtual machine corresponding to the target data list based on the data attribute information of the target data list and the virtual machine state level value of the first transfer-in virtual machine;

[0015] Preferably, the same virtual machine is a virtual machine that belongs to the same physical machine as the target transfer-out virtual machine.

[0016] Preferably, the data attribute information refers to attribute information of an application running on the target roll-out virtual machine or data stored on the target roll-out virtual machine.

[0017] Preferably, the step of analyzing and acquiring the target transfer-in virtual machine corresponding to the target data list based on the data attribute information of the target data list and the virtual machine state level value of the first transfer-in virtual machine comprises the following steps:

[0018] Encoding each of the target data lists to obtain a pre-code; and encoding each of the first transferred virtual machines according to the virtual machine state level value of the first transferred virtual machine to obtain a post-code;

[0019] Performing an initial random combination according to the pre-code and the post-code to construct a plurality of first lists for transfer to codes;

[0020] Obtaining an input fitness F by calculating the fitness calculation formula according to the first list input code; determining whether the input fitness F is greater than or equal to a preset first input fitness threshold; if so, determining the first input virtual machine corresponding to the first list input code as the second input virtual machine; if not, filtering out the first input virtual machine;

[0021] A second virtual machine set is constructed based on all the second transferred virtual machines; and a target transferred virtual machine is obtained by screening the second virtual machine set through probability analysis.

[0022] Preferably, said constructing a second virtual machine set based on all the second transferred virtual machines; and obtaining a target transferred virtual machine through probability analysis and screening according to the second virtual machine set includes the following steps:

[0023] Normalize the virtual machine state level values ​​corresponding to each of the second transferred virtual machines to obtain a normalized state value s i ;

[0024] Traversing all the second transferred virtual machines in the second virtual machine set, and obtaining a transfer probability of the second transferred virtual machine according to the normalized state value calculation;

[0025] The calculation method of the transfer probability of the second transfer virtual machine is:

[0026]

[0027] In the formula, s i is the normalized state value of the i-th second-transferred virtual machine; n is the number of second-transferred virtual machines;

[0028] A target transfer-in virtual machine corresponding to the target data list is selected and acquired according to the transfer-in probability of the second transfer-in virtual machine.

[0029] Preferably, the step of selecting and acquiring a target transfer-in virtual machine corresponding to the target data list according to the transfer-in probability of the second transfer-in virtual machine comprises the following steps:

[0030] Acquire the data priority corresponding to each of the target data lists; sort each of the target data lists from high to low according to the data priority to obtain a target data list sequence set;

[0031] Sort each of the second transfer-in virtual machines from high to low according to the transfer-in probability to obtain a transfer-in sequence set;

[0032] Traversing all the target data lists in the target data list sequence set in sequence, determining the target data list ranked first in the target data list sequence set as the first data; determining the second input virtual machine ranked first in the input sequence set as the target input virtual machine of the current first data; transferring the first data into the target input virtual machine to obtain an updated second input virtual machine ranked first;

[0033] Obtain the updated second virtual machine state level value of the first-ranked second-imported virtual machine; and calculate according to the second virtual machine state level value to obtain the updated transfer probability of the first-ranked second-imported virtual machine;

[0034] Based on the updated transfer-in probability, the transfer-in sequence set is updated to obtain an updated transfer-in sequence set; after the current first data is filtered out from the target data list sequence set, the above operation is returned until the target data list sequence set is empty, and the target transfer-in virtual machines corresponding to each target data list are obtained.

[0035] Preferably, the obtaining of the data priority corresponding to each of the target data lists comprises the following steps:

[0036] Acquire the value parameter qr of the target data list according to the data attribute information;

[0037] Obtaining the fault start time d and the data loss emergency time t corresponding to each of the target data lists;

[0038] Analyze and obtain the fuzzy influencing parameter p of each target data list;

[0039] Calculate and obtain the data priority of the target data list according to the value parameter qr, the fault start time d, the data loss urgency time t, and the fuzzy influence parameter p;

[0040] The data priority is calculated as follows:

[0041]

[0042] Where f(p) is the fuzzy influencing parameter function.

[0043] Preferably, the analysis to obtain the fuzzy influencing parameter p of each target data list includes the following steps:

[0044] Traverse each of the target data lists to obtain the determined influencing parameter x corresponding to each sub-data in the target data list;

[0045] Calculate the approximate influencing parameter x' for each of the determined influencing parameters x;

[0046] The calculation method of the approximate influencing parameter is:

[0047]

[0048] Where b is the number of sub-data in the current target data list;

[0049] Calculate and obtain a fuzzy influencing parameter p according to the approximate influencing parameter x';

[0050] The calculation method of the fuzzy influencing parameter p is:

[0051]

[0052] Compared with the prior art, the embodiments of the present invention have at least the following technical advantages:

[0053] From the analysis of the above-mentioned multi-virtual machine disaster recovery response processing method in a high-availability cloud computing cluster provided by the present invention, it can be known that in specific applications, the current data list and virtual machine status level value of each virtual machine corresponding to the same server cluster are obtained; the target transfer-out virtual machine and the target data list corresponding to the target transfer-out virtual machine are obtained according to the preset virtual machine status level value threshold; the target data list is analyzed according to the virtual machine status level value to obtain the target transfer-in virtual machine corresponding to the target data list; and the data access request corresponding to the target data list is redirected to the target transfer-in virtual machine.

[0054] By obtaining the current data list and status level value of each virtual machine, it can timely understand the operating status, load status and data status of each virtual machine, thus providing basic data for subsequent decision-making; according to the preset virtual machine status level value threshold, the virtual machine to be transferred out is determined, and the target data list corresponding to these virtual machines is obtained, and then the virtual machine that needs to be processed by disaster recovery response is identified, so as to prepare for the subsequent load transfer, improve the accuracy and efficiency of disaster recovery response, and be able to handle the virtual machine with problems in time, reducing the risk of system interruption and data loss; according to the target data list and the virtual machine status level value, through probability analysis, priority calculation and sorting, fuzzy influence reference, etc., analysis is performed to determine the target transfer virtual machine, ensure that the target virtual machine can effectively process data and requests after the transfer, improve the efficiency and success rate of disaster recovery response, and ensure the stability and reliability of the system after load transfer by reasonably adjusting the status of the virtual machine, reducing additional problems caused by disaster recovery response;

[0055] Finally, the data access requests corresponding to the target data list are redirected to the target virtual machine to ensure the continuity and stability of data access, reduce business interruptions caused by virtual machine failures, improve the system's disaster recovery capabilities and business continuity, and ensure the normal operation of the system and the integrity of the data. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0057] Figure 1 A schematic diagram of the overall operation steps of a method for disaster recovery response of multiple virtual machines in a high-availability cloud computing cluster provided by an embodiment of the present invention;

[0058] Figure 2 A schematic diagram of the operation steps of analyzing and transferring a target to a virtual machine in a multi-virtual machine disaster recovery response processing method in a high-availability cloud computing cluster provided by an embodiment of the present invention;

[0059] Figure 3 A schematic diagram of a simulation of an operation process of transferring a target to a virtual machine obtained by analyzing a method for disaster recovery response of multiple virtual machines in a high-availability cloud computing cluster provided by an embodiment of the present invention;

[0060] Figure 4 A schematic diagram of the operation steps of analyzing and obtaining a target transferred virtual machine in a multi-virtual machine disaster recovery response processing method in a high-availability cloud computing cluster provided by an embodiment of the present invention;

[0061] Figure 5 A schematic diagram of the operation steps of obtaining a target transfer virtual machine through probability analysis screening in a multi-virtual machine disaster recovery response processing method in a high-availability cloud computing cluster provided by an embodiment of the present invention;

[0062] Figure 6 A schematic diagram of the operation steps of selecting a target virtual machine to be transferred according to the transfer probability in a disaster recovery response processing method for multiple virtual machines in a high-availability cloud computing cluster provided by an embodiment of the present invention;

[0063] Figure 7 A schematic diagram of the operation steps of obtaining the data priority corresponding to the target data list in a multi-virtual machine disaster recovery response processing method in a high-availability cloud computing cluster provided by an embodiment of the present invention;

[0064] Figure 8 A schematic diagram of the operational steps of analyzing and obtaining fuzzy influencing parameters of each target data list in a multi-virtual machine disaster recovery response processing method in a high-availability cloud computing cluster provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0065] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0066] The present invention is further described in detail below through specific embodiments in conjunction with the accompanying drawings.

[0067] like Figure 1 As shown, the present invention proposes a multi-virtual machine disaster recovery response processing method in a high-availability cloud computing cluster, comprising the following steps:

[0068] Step S10: Obtain the current data list and virtual machine status level value of each virtual machine corresponding to the same server cluster;

[0069] It should be noted that the current data list of each virtual machine refers to the list of applications running on each virtual machine or the data stored therein; the virtual machine status level value refers to the virtual machine status level value returned by each virtual machine after a health identification instruction is sent to each virtual machine and each virtual machine responds to the health identification instruction and is calculated by comprehensively analyzing the data processing rate, load status, real-time bandwidth and data volume of each virtual machine;

[0070] The above-mentioned embodiment of the present application obtains the current data list and status level value of each virtual machine in the cluster; through this information, the operating status, load status and data status of each virtual machine can be understood, providing basic data for subsequent decision-making;

[0071] Step S20: acquiring a target transfer-out virtual machine and a target data list corresponding to the target transfer-out virtual machine according to a preset virtual machine status level numerical threshold;

[0072] It should be noted that the target transfer-out virtual machine refers to a virtual machine in the current server cluster whose virtual machine status level value is less than the virtual machine status level value threshold, and is identified as having a virtual machine operation failure and requiring disaster recovery processing; the target data list refers to a list of applications running or stored data on the target transfer-out virtual machine;

[0073] The above-mentioned embodiment of the present application determines the virtual machines that need to be transferred out and the target data list corresponding to these virtual machines according to the preset virtual machine status level numerical threshold; it helps to identify the virtual machines that need to be processed for disaster recovery response and prepare for subsequent load transfer.

[0074] Step S30: analyzing the target data list according to the virtual machine state level value to obtain the target transfer virtual machine corresponding to the target data list;

[0075] It should be noted that the above embodiment of the present application analyzes the virtual machine status level value to determine the target virtual machine to be transferred. This process can ensure that the target virtual machine can effectively process data and requests after the load transfer, thereby improving the efficiency and success rate of disaster recovery response;

[0076] Step S40: redirecting the data access request corresponding to the target data list to the target transfer-in virtual machine.

[0077] It should be noted that the above-mentioned embodiment of the present application redirects the data access request corresponding to the target data list to the target virtual machine to complete the entire disaster recovery response processing process.

[0078] The above-mentioned embodiment of the present application can timely understand the operation status, load status and data status of each virtual machine by obtaining the current data list and status level value of each virtual machine, so as to provide basic data for subsequent decision-making; determine the virtual machine to be transferred out according to the preset virtual machine status level value threshold, and obtain the target data list corresponding to these virtual machines, and then identify the virtual machine that needs to be processed by disaster recovery response, so as to prepare for the subsequent load transfer, improve the accuracy and efficiency of disaster recovery response, and be able to handle the virtual machine with problems in time, reducing the risk of system interruption and data loss; analyze according to the target data list and the virtual machine status level value, determine the target transfer-in virtual machine, ensure that the target virtual machine can effectively process data and requests after the transfer, improve the efficiency and success rate of disaster recovery response, and ensure the stability and reliability of the system after load transfer by reasonably adjusting the status of the virtual machine, and reduce additional problems caused by disaster recovery response;

[0079] Finally, the data access request corresponding to the target data list is transferred to the target virtual machine to ensure the continuity and stability of data access, reduce business interruptions caused by virtual machine failures, improve the system's disaster recovery capabilities and business continuity, and ensure the normal operation of the system and the integrity of data;

[0080] The above-mentioned embodiments of the present application can obtain the status and data of the virtual machine in real time, which is conducive to real-time monitoring and management; can make intelligent judgments and adjustments based on preset thresholds and status level values, and improve the intelligence and adaptability of disaster recovery processing; can quickly and accurately redirect data access requests to the target virtual machine, reduce the system interruption time caused by failures, and improve the reliability and stability of the system;

[0081] The technical solution adopted in the above-mentioned embodiment of the present application can effectively improve the efficiency and success rate of disaster recovery response processing of multiple virtual machines in a high-availability cloud computing cluster, and has a positive role in ensuring the stability and continuity of the system.

[0082] Specifically, Figure 2 , Figure 3 As shown, in step S30, the target data list is analyzed according to the virtual machine state level value to obtain the target transfer virtual machine corresponding to the target data list. The following operation steps are included:

[0083] Step S31: obtaining virtual machine attribute information corresponding to each of the virtual machines; filtering and obtaining multiple virtual machines belonging to the same virtual machine according to the virtual machine attribute information; and filtering and obtaining multiple first transferred virtual machines according to the virtual machines belonging to the same virtual machine;

[0084] The same virtual machine is a virtual machine that belongs to the same physical machine as the target transfer-out virtual machine;

[0085] The above-mentioned obtaining a plurality of first transfer-in virtual machines by screening the same virtual machines is to screen out the same virtual machines from all current virtual machines, thereby obtaining all the first transfer-in virtual machines that do not belong to the physical machine corresponding to the current target transfer-out virtual machine;

[0086] It should be noted that the above-mentioned embodiment of the present application is the first screening operation to obtain the target transfer-in virtual machine, and the purpose is to screen out the first transfer-in virtual machine that is not in the same physical machine as the current target transfer-out virtual machine. This is because when the current target transfer-out virtual machine fails and disaster recovery processing is required, it may be because the corresponding physical machine is attacked or human or non-human factors cause the physical machine to be unable to normally operate the current target transfer-out virtual machine. Therefore, it is necessary to screen out each virtual machine running on the current physical machine (that is, the above-mentioned virtual machines of the same kind) to obtain multiple first transfer-in virtual machines on other physical machines.

[0087] Step S32: Analyze and obtain the target transfer-in virtual machine corresponding to the target data list based on the data attribute information of the target data list and the virtual machine state level value of the first transfer-in virtual machine;

[0088] The data attribute information refers to the attribute information of the application running on the target transfer-out virtual machine or the data stored on the target transfer-out virtual machine; the data attribute information includes data volume, reference transfer-in minimum memory size, reference transfer-in data processing rate, reference transfer-in bandwidth value, etc.;

[0089] It should be noted that the above-mentioned embodiment of the present application screens out the virtual machines on the faulty physical machine through the first screening operation of S31, so as to avoid misselecting the virtual machine of the faulty physical machine when searching for the target input virtual machine in the target data list of the target output virtual machine, thereby improving the efficiency of subsequent analysis and processing; further, through the second screening operation of S32, analysis is performed based on the data attribute information of the target data list and the virtual machine attribute information of the first input virtual machine, so as to screen out the target input virtual machine of the target data list; through the last two screening operations, the target input virtual machine can be screened out more quickly, and then the target data list can be transferred into the target input virtual machine for operation or data access, thereby reducing the data loss caused by the physical machine failure to a greater extent and improving data security.

[0090] Specifically, Figure 4 As shown, in step S32, the target transfer-in virtual machine corresponding to the target data list is obtained by analyzing the data attribute information of the target data list and the virtual machine state level value of the first transfer-in virtual machine, including the following operation steps:

[0091] Step S321: encoding each of the target data lists to obtain a pre-code; and encoding each of the first transferred virtual machines according to the virtual machine state level value of the first transferred virtual machine to obtain a post-code;

[0092] It should be noted that the encoding in the above-mentioned embodiment of the present application can use binary encoding or floating point encoding to encode the target data list to obtain the pre-encoding; wherein the pre-encoding can reflect the attribute information of the target data list after decoding;

[0093] At the same time, the post-coding obtained by encoding the virtual machine state level value of the first transferred virtual machine can reflect the virtual machine state level value of the first transferred virtual machine;

[0094] The above-mentioned pre-coding and post-coding can be randomly combined in subsequent operations to obtain an initial population. According to the initial population (i.e., the first list entry code described below), the optimal target entry virtual machine of the target data list can be determined by fitness calculation and other methods. The specific operations are shown in the following steps.

[0095] Step S322: performing an initial random combination according to the pre-code and the post-code to construct a plurality of first list transfer codes;

[0096] It should be noted that in the above embodiment of the present application, each pre-code is randomly combined with any post-code to obtain an initial population (i.e., the first list transfer code mentioned above);

[0097] Step S323: Calculate the transfer fitness F according to the first list transfer code by using the fitness calculation formula; determine whether the transfer fitness F is greater than or equal to the preset transfer fitness first threshold; if so, determine the first transfer virtual machine corresponding to the first list transfer code as the second transfer virtual machine; if not, filter out the first transfer virtual machine;

[0098] The transfer fitness F represents the fitness of the running state of the first transfer virtual machine corresponding to the target data list corresponding to the first list transfer code (or the fitness of the solution).

[0099] Step S324: construct a second virtual machine set based on all the second transferred virtual machines; and obtain a target transferred virtual machine by screening the second virtual machine set through probability analysis.

[0100] It should be noted that the above embodiment of the present application abstracts and represents the target data list and the virtual machine status level value by encoding, providing a data basis for subsequent random combination and fitness calculation. Such encoding method can retain the attribute information of the original data to a certain extent; multiple possible solutions are obtained by random combination, forming an initial population, providing an initial solution space for subsequent optimization algorithms; through fitness calculation, the advantages and disadvantages of each solution are evaluated, so as to screen out the appropriate transfer virtual machine. Such a screening process can improve the efficiency of virtual machine resource utilization; finally, by screening through probability analysis, the second virtual machine set is analyzed and screened to ensure reasonable resource allocation and improve the stability and performance of the system.

[0101] Specifically, Figure 5 As shown, in step S324, a second virtual machine set is constructed based on all the second transferred virtual machines; and a target transferred virtual machine is obtained by probability analysis and screening according to the second virtual machine set, including the following operation steps:

[0102] Step S3241: normalize the virtual machine state level values ​​corresponding to each of the second transferred virtual machines to obtain a normalized state value s i ;

[0103] Step S3242: traverse all the second transferred virtual machines in the second virtual machine set, and obtain the transfer probability of the second transferred virtual machine according to the normalized state value calculation;

[0104] The calculation method of the transfer probability of the second transfer virtual machine is:

[0105]

[0106] In the formula, s i is the normalized state value of the i-th second-transferred virtual machine; n is the number of second-transferred virtual machines;

[0107] Step S3243: selecting and acquiring a target transfer-in virtual machine corresponding to the target data list according to the transfer-in probability of the second transfer-in virtual machine.

[0108] It should be noted that the above-mentioned embodiment of the present application maps the state values ​​of each virtual machine into the same value range through normalization processing, eliminates the dimensional difference between the values, and makes the subsequent calculation more concise and accurate; by calculating the transfer probability, the appropriate virtual machine can be flexibly selected for resource transfer according to the state level of the virtual machine, thereby improving the intelligence and efficiency of resource allocation;

[0109] The above-mentioned embodiments of the present application calculate the transfer probability, and the scheme can intelligently select the appropriate target transfer virtual machine, making the resource transfer process more intelligent and adaptive; through normalization processing and transfer probability calculation, the scheme can better optimize the utilization of virtual machine resources, making the overall system performance more efficient and stable; the normalization processing and transfer probability calculation in the scheme can be dynamically adjusted according to the system operation status, so that the system can flexibly adapt to different environments and demand changes; the implementation of the embodiments of the present application improves the intelligence of resource allocation and optimizes resource utilization efficiency.

[0110] During the specific implementation of the above-mentioned embodiment of the present application, the technicians found that because a physical machine generally runs multiple virtual machines, when a disaster recovery event occurs in a virtual machine (i.e., when a physical machine fails), there may be multiple virtual machines that need to perform disaster recovery processing operations at the same time; therefore, it is necessary to simultaneously search for target transfers to virtual machines for these virtual machines that have experienced disaster recovery events;

[0111] However, in the above-mentioned embodiment of the present application, because the transfer probability is calculated using the virtual machine status level value of the second transfer-in virtual machine, there may be multiple virtual machines being transferred to the same second transfer-in virtual machine at the same time. At this time, the simultaneous transfer of multiple virtual machines may cause the current transferred-in second transfer-in virtual machine (i.e., the above-mentioned target transfer-in virtual machine) to run out of memory or fail. Therefore, it is also necessary to consider the attribute information of the target data list of each target transfer-out virtual machine and the implementation of load balancing, so as to transfer each target data list to multiple target transfer-in virtual machines respectively, thereby achieving rapid response to disaster recovery events and reducing the risk of data loss.

[0112] Specifically, Figure 6 As shown, in step S3243, selecting and acquiring a target transfer-in virtual machine corresponding to the target data list according to the transfer-in probability of the second transfer-in virtual machine includes the following operation steps:

[0113] Step S32431: Obtain the data priority corresponding to each of the target data lists; sort each of the target data lists from high to low according to the data priority to obtain a target data list sequence set;

[0114] Step S32432: sorting the second transfer-in virtual machines from high to low according to the transfer-in probability to obtain a transfer-in sequence set;

[0115] Step S32433: traverse all the target data lists in the target data list sequence set in sequence, determine that the target data list ranked first in the target data list sequence set is the first data; determine that the second input virtual machine ranked first in the input sequence set is the target input virtual machine of the current first data; transfer the first data into the target input virtual machine to obtain an updated second input virtual machine ranked first;

[0116] Step S32434: obtaining the updated second virtual machine state level value of the first-ranked second-imported virtual machine; calculating according to the second virtual machine state level value to obtain the updated transfer probability of the first-ranked second-imported virtual machine;

[0117] Step S32435: update the transfer-in sequence set based on the updated transfer-in probability to obtain an updated transfer-in sequence set; filter out the current first data from the target data list sequence set and return to the above step S32433 until the target data list sequence set is empty, and obtain the target transfer-in virtual machine corresponding to each target data list.

[0118] It should be noted that the above-mentioned embodiment of the present application obtains the data priority of the target data list and sorts it, and obtains the transfer sequence set according to the transfer probability, thereby ensuring that high-priority data is preferentially transferred to the second transfer virtual machine with a high state level value, thereby improving the response speed and efficiency of the system, ensuring that important data can be processed in time, and thus improving the overall performance of the system;

[0119] In specific operations, by traversing the target data list and the transfer sequence set, the target transfer virtual machine corresponding to each data list is determined; it helps to determine the best task allocation plan to ensure that each task is assigned to the most suitable virtual machine, thereby improving the overall efficiency and performance of the system; obtain the updated virtual machine state level value, and calculate the updated transfer probability based on the value; the probability of task allocation can be adjusted according to the current state of the virtual machine to adapt to the dynamic changes of the system, thereby improving the adaptability and flexibility of the system; based on the updated transfer probability, the transfer sequence set is updated to obtain the latest task allocation plan. By continuously updating the transfer sequence set, the system can dynamically adjust the task allocation according to the real-time situation, thereby better adapting to changes in the system and improving the overall performance and efficiency;

[0120] In summary, the embodiments of the present application can improve the efficiency, performance and flexibility of the system by giving priority to processing important data, reasonably allocating virtual machine resources, and dynamically adjusting task allocation probability.

[0121] Specifically, Figure 7As shown, in step S32431, obtaining the data priority corresponding to each of the target data lists includes the following steps:

[0122] Step S324311: Obtain the value parameter qr of the target data list according to the data attribute information;

[0123] It should be noted that the above value parameters refer to the value that the current target data list will bring when it is operating normally. For example, the normal operation of the target data list stored in the key data node is of great value to the entire cluster. For another example, the normal operation of the data cache or log table is not essential to the value of the entire cluster.

[0124] Step S324312: Obtain the fault start time d and the data loss emergency time t corresponding to each of the target data lists;

[0125] It should be noted that in the above-mentioned embodiment of the present application, the failure start time refers to the time point when the target transfer-out virtual machine corresponding to the current target data list fails and disaster recovery processing is required; the above-mentioned data loss emergency time refers to the time point when the risk of data loss will occur after the current target transfer-out virtual machine fails for a period of time.

[0126] Step S324313: Analyze and obtain the fuzzy influencing parameter p of each target data list;

[0127] The fuzzy influencing parameters in the above-mentioned embodiments of the present application refer to the degree of uncertainty of the fuzzy influencing factors faced when making decisions or plans; this uncertainty may come from various factors, such as the temporary access volume of the target transfer-out virtual machine, the bandwidth of the target transfer-out virtual machine, the network fluctuation of the target transfer-out virtual machine, the current and voltage of the target transfer-out virtual machine, etc. Affected by these fuzzy influencing factors, the priority of the target data list will also be greatly affected.

[0128] Step S324314: Calculate and obtain the data priority of the target data list according to the value parameter qr, the fault start time d, the data loss emergency time t, and the fuzzy influence parameter p;

[0129] The data priority is calculated as follows:

[0130]

[0131] Where f(p) is the fuzzy influencing parameter function;

[0132] It should be noted that the above-mentioned embodiment of the present application calculates the value parameter of the target data list according to the data attribute information, taking into account the importance of different data lists to the entire cluster, so as to subsequently determine the priority of the data according to its value; obtains the failure start time and data loss urgency time of the target data list, and evaluates the possible impact when the target roll-out virtual machine fails, so as to consider these factors in disaster recovery processing or decision-making; analyzes the fuzzy impact parameters of the target data list, taking into account the uncertainty factors faced in decision-making or planning, such as temporary access volume, bandwidth, network fluctuations, etc., which may affect the priority of the data list; calculates the data priority of the target data list according to the value parameter, the failure start time, the data loss urgency time and the fuzzy impact parameter, taking into account the importance of the data list, the impact degree when the failure occurs and the uncertainty factors, so as to obtain the final data priority;

[0133] In summary, by considering the failure start time and the urgency of data loss in the above-mentioned embodiments of the present application, the system can better cope with the possible impact of a virtual machine failure, thereby improving the disaster recovery processing capability of the system; by calculating the priority of the data, the system can prioritize the data according to the importance and impact of the data, thereby improving the performance and efficiency of the entire system; by analyzing fuzzy influencing parameters, the system can take into account the uncertainty factors faced in decision-making or planning, thereby more accurately evaluating the priority and impact of the data; by comprehensively considering the value, impact and uncertainty factors of the data, the system can more intelligently perform data processing and resource allocation, thereby improving the efficiency and performance of the overall system.

[0134] Specifically, Figure 8 As shown, in step S324313, the fuzzy influencing parameter p of each target data list is analyzed and obtained, including the following operation steps:

[0135] Step S3243131: traverse each of the target data lists to obtain the determined influencing parameter x corresponding to each sub-data in the target data list;

[0136] Step S3243132: Calculate the approximate influencing parameter x' for each of the determined influencing parameters x;

[0137] The calculation method of the approximate influencing parameter is:

[0138]

[0139] In the formula, b is the number of sub-data in the current target data list;

[0140] Step S3243133: Calculate and obtain the fuzzy influencing parameter p according to the approximate influencing parameter x';

[0141] The calculation method of the fuzzy influencing parameter p is:

[0142]

[0143] It should be noted that the above-mentioned embodiment of the present application determines the degree of influence of each sub-data on the overall system by traversing the target data list and obtaining the determined influencing parameter x corresponding to the sub-data, thereby providing basic data for subsequent calculations; the determined influencing parameter x is approximated and calculated to obtain the approximate influencing parameter x'. The approximate calculation can simplify the calculation process of the influencing parameter, improve the calculation efficiency, and may be easier to obtain and process in actual situations. The fuzzy influencing parameter p is calculated based on the approximate influencing parameter x', and the approximate influencing parameter is converted into a fuzzy influencing parameter, taking into account the influence of uncertainty factors on the influencing parameter, thereby more comprehensively evaluating the degree of influence of the data;

[0144] By comprehensively considering the influence parameters and fuzzy influence parameters, the influence of each sub-data on the entire system can be more accurately evaluated, providing more accurate data for subsequent priority calculations; by approximate calculations to determine the influence parameters and fuzzy influence parameters, the calculation process can be simplified, the calculation efficiency can be improved, and the consumption of computing resources can be reduced; by calculating the fuzzy influence parameters, the influence of uncertainty factors on the influence parameters is considered, making the evaluation more comprehensive and accurate, and better able to deal with the impact of uncertainty factors;

[0145] In summary, the technical solution adopted in the above-mentioned embodiment of the present application improves the accuracy and calculation efficiency of the impact degree of evaluation data by comprehensively utilizing definite influencing parameters and fuzzy influencing parameters, and takes into account uncertainty factors, making the evaluation more comprehensive and reliable.

[0146] In summary, the present invention proposes a method for disaster recovery response processing of multiple virtual machines in a high-availability cloud computing cluster. By obtaining the current data list and status level value of each virtual machine, the operating status, load status and data status of each virtual machine can be timely understood, thus providing basic data for subsequent decision-making; the virtual machines to be transferred out are determined according to the preset virtual machine status level value threshold, and the target data list corresponding to these virtual machines is obtained, and then the virtual machines to be processed for disaster recovery response are identified, thus preparing for the subsequent load transfer, improving the accuracy and efficiency of the disaster recovery response, and being able to handle the virtual machines with problems in a timely manner, reducing the risk of system interruption and data loss; according to the target data list and the virtual machine status level value, probability analysis, priority calculation sorting, fuzzy influence reference quantity and the like are analyzed to determine the target transfer-in virtual machine, ensure that the target virtual machine can effectively process data and requests after the transfer, improve the efficiency and success rate of the disaster recovery response, and reasonably adjust the status of the virtual machine to ensure the stability and reliability of the system after the load transfer, and reduce the additional problems caused by the disaster recovery response;

[0147] Finally, the data access request corresponding to the target data list is transferred to the target virtual machine to ensure the continuity and stability of data access, reduce business interruptions caused by virtual machine failures, improve the system's disaster recovery capabilities and business continuity, and ensure the normal operation of the system and the integrity of data;

[0148] Furthermore, when the target input virtual machine is obtained through analysis, screening and analysis are performed through the same virtual machines and probability analysis screening to improve the screening accuracy of the target output virtual machine; further, when performing probability screening analysis, the priority of the target data list is calculated according to the value parameter qr of the target data list, the fault start time d and the data loss urgency time t and the fuzzy influence parameter p, thereby realizing the screening operation of the target input virtual machine, performing data processing and resource allocation more intelligently, thereby improving the efficiency and performance of the overall system.

[0149] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. A person skilled in the art may modify the technical solutions described in the above embodiments, or replace part or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for disaster recovery response of multiple virtual machines in a high-availability cloud computing cluster. It is characterized in that The steps are as follows: Obtain the current data list and virtual machine status level value of each virtual machine corresponding to the same server cluster; Acquire a target transfer-out virtual machine and a target data list corresponding to the target transfer-out virtual machine according to a preset virtual machine status level numerical threshold; Analyzing the target data list according to the virtual machine state level value to obtain a target transfer virtual machine corresponding to the target data list; The data access request corresponding to the target data list is forwarded to the target transfer-in virtual machine.

2. According to the method for disaster recovery response of multiple virtual machines in a high-availability cloud computing cluster according to claim 1, It is characterized in that The data list is a list of applications running on each virtual machine or stored data.

3. According to claim 2, a method for disaster recovery response processing of multiple virtual machines in a high-availability cloud computing cluster, It is characterized in that The step of analyzing the target data list according to the virtual machine state level value to obtain the target transfer virtual machine corresponding to the target data list includes the following steps: Acquire virtual machine attribute information corresponding to each of the virtual machines; filter and acquire multiple virtual machines belonging to the same category according to the virtual machine attribute information; and filter and acquire multiple first transferred virtual machines according to the virtual machines belonging to the same category; The target transfer-in virtual machine corresponding to the target data list is obtained by analyzing the data attribute information of the target data list and the virtual machine state level value of the first transfer-in virtual machine.

4. According to claim 3, a method for disaster recovery response processing of multiple virtual machines in a high-availability cloud computing cluster, It is characterized in that The same virtual machine is a virtual machine that belongs to the same physical machine as the target transfer-out virtual machine.

5. According to claim 4, a method for disaster recovery response processing of multiple virtual machines in a high-availability cloud computing cluster, It is characterized in that The data attribute information refers to attribute information of an application running on the target roll-out virtual machine or data stored on the target roll-out virtual machine.

6. A method for disaster recovery response processing of multiple virtual machines in a high-availability cloud computing cluster according to claim 5, It is characterized in that The step of analyzing the data attribute information of the target data list and the virtual machine state level value of the first transferred virtual machine to obtain the target transferred virtual machine corresponding to the target data list includes the following steps: Encoding each of the target data lists to obtain a pre-code; and encoding each of the first transferred virtual machines according to the virtual machine state level value of the first transferred virtual machine to obtain a post-code; Performing an initial random combination according to the pre-code and the post-code to construct a plurality of first lists for transfer to codes; Obtaining an input fitness F by calculating the fitness calculation formula according to the first list input code; determining whether the input fitness F is greater than or equal to a preset first input fitness threshold; if so, determining the first input virtual machine corresponding to the first list input code as the second input virtual machine; if not, filtering out the first input virtual machine; A second virtual machine set is constructed based on all the second transferred virtual machines; and a target transferred virtual machine is obtained by screening the second virtual machine set through probability analysis.

7. A method for disaster recovery response processing of multiple virtual machines in a high-availability cloud computing cluster according to claim 6, It is characterized in that The step of constructing a second virtual machine set based on all the second transferred virtual machines; and obtaining a target transferred virtual machine by probability analysis and screening according to the second virtual machine set includes the following steps: Normalize the virtual machine state level values ​​corresponding to each of the second transferred virtual machines to obtain a normalized state value s i ; Traversing all the second transferred virtual machines in the second virtual machine set, and obtaining a transfer probability of the second transferred virtual machine according to the normalized state value calculation; The calculation method of the transfer probability of the second transfer virtual machine is: In the formula, s i is the normalized state value of the i-th second-transferred virtual machine; n is the number of second-transferred virtual machines; A target transfer-in virtual machine corresponding to the target data list is selected and acquired according to the transfer-in probability of the second transfer-in virtual machine.

8. A method for disaster recovery response of multiple virtual machines in a high-availability cloud computing cluster according to claim 7, It is characterized in that The step of selecting and acquiring a target transfer-in virtual machine corresponding to the target data list according to the transfer-in probability of the second transfer-in virtual machine comprises the following steps: Acquire the data priority corresponding to each of the target data lists; sort each of the target data lists from high to low according to the data priority to obtain a target data list sequence set; Sort each of the second transfer-in virtual machines from high to low according to the transfer-in probability to obtain a transfer-in sequence set; Traversing all the target data lists in the target data list sequence set in turn, and determining the target data list ranked first in the target data list sequence set as the head data; Determine the second transfer-in virtual machine ranked first in the transfer-in sequence set as the target transfer-in virtual machine of the current first data; Transferring the first data into the target transfer virtual machine to obtain an updated first-ranked second transfer virtual machine; Obtain the updated second virtual machine state level value of the first-ranked second-imported virtual machine; and calculate according to the second virtual machine state level value to obtain the updated transfer probability of the first-ranked second-imported virtual machine; Based on the updated transfer-in probability, the transfer-in sequence set is updated to obtain an updated transfer-in sequence set; after the current first data is filtered out from the target data list sequence set, the above operation is returned until the target data list sequence set is empty, and the target transfer-in virtual machines corresponding to each target data list are obtained.

9. A method for disaster recovery response processing of multiple virtual machines in a high-availability cloud computing cluster according to claim 8, It is characterized in that The step of obtaining the data priority corresponding to each target data list comprises the following steps: Acquire the value parameter qr of the target data list according to the data attribute information; Obtaining the fault start time d and the data loss emergency time t corresponding to each of the target data lists; Analyze and obtain the fuzzy influencing parameters p of each of the target data lists; Calculate and obtain the data priority of the target data list according to the value parameter qr, the fault start time d, the data loss emergency time t, and the fuzzy influence parameter p; The data priority is calculated as follows: Where f(p) is the fuzzy influencing parameter function.

10. A method for disaster recovery response of multiple virtual machines in a high-availability cloud computing cluster according to claim 9, It is characterized in that The analysis to obtain the fuzzy influencing parameter p of each target data list includes the following steps: Traverse each of the target data lists to obtain the determined influencing parameter x corresponding to each sub-data in the target data list; Calculate the approximate influencing parameter x' for each of the determined influencing parameters x; The calculation method of the approximate influencing parameter is: Wherein, b is the number of sub-data in the current target data list; the fuzzy influencing parameter p is calculated and obtained according to the approximate influencing parameter x'; The calculation method of the fuzzy influencing parameter p is: