Method and device for optimally distributing storage space of redundant disk array

Abstract

The invention discloses a storage space optimal distribution method and device for a redundant disk array, and relates to the technical field of storage device.The method comprises the steps that the stability level of a physical hard disk drive is obtained, and a data importance recognition model is established; setting an equipment monitoring group for data acquisition; establishing a network space coordinate system to supervise the safety of important areas; establishing a distribution-storage dynamic balance model of the disk array, and screening out an optimal distribution-storage work arrangement; establishing a dynamic monitoring and early warning model of the disk array equipment, and judging and predicting the running state of the disk array equipment; and establishing a smart disk array network dynamic model. By analyzing the optimal distribution mode of data distribution and storage, a distribution-storage dynamic balance model is established, an optimal distribution-storage work arrangement is screened out, a disk array equipment dynamic monitoring and early warning model is constructed, the running state of disk array equipment is judged and predicted, and a human-computer interaction visual platform is created.

Description

Technical Field

[0001] The present invention relates to the technical field of storage devices, and in particular to a storage space optimization allocation method and a device for a redundant disk array. Background Art

[0002] Disk array is composed of many independent disks, combined into a huge disk group, using the additive effect of individual disks to provide data to improve the performance of the entire disk system. Using this technology, data is cut into many segments and stored on each hard disk.

[0003] Existing technical means lack risk assessment of the current status of disk arrays, lack long-term, cumulative equipment risk assessment methods, and secondly, lack cost considerations during storage and allocation, resulting in the existing storage model being unable to fully meet various data storage needs. Summary of the invention

[0004] In order to solve the above technical problems, a method and device for optimizing the allocation of storage space of a redundant disk array are provided. The technical solution solves the problem that the existing technical means proposed in the above background technology lack risk assessment of the existing state of the disk array, lack long-term and cumulative equipment risk assessment methods, and secondly, lack cost considerations during storage and allocation, resulting in the existing storage mode being difficult to fully meet various data storage needs.

[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is:

[0006] A method for optimizing storage space allocation of a redundant disk array includes:

[0007] Acquire and aggregate at least one physical hard disk drive as a disk array;

[0008] According to the stability, the stability level of the physical hard disk drive is obtained, and a data importance recognition model is established;

[0009] According to the actual needs of disk array management, set up equipment monitoring groups for data collection;

[0010] Establish a cyberspace coordinate system, mark data movements through equipment monitoring groups, and monitor the security of important areas;

[0011] According to the actual storage requirements of the disk array, establish the disk array allocation-storage dynamic balance model to select the optimal allocation-storage work arrangement;

[0012] Based on the LSTM neural network algorithm model, a dynamic monitoring and early warning model for disk array devices is established to judge and predict the operating status of disk array devices, send abnormal information of abnormal devices and make risk alerts;

[0013] Through digital twin technology, a dynamic model of the smart disk array network is established.

[0014] Preferably, obtaining the stability level of the physical hard disk drive according to the stability and establishing the data importance identification model comprises the following steps:

[0015] Evenly divide the value range of the data loss rate of the physical hard disk drive to obtain at least one stable identification interval;

[0016] According to the midpoint of the stable identification interval, the stable identification interval is numbered from large to small, and the number of the stable identification interval is used as the stability level of the stable identification interval;

[0017] Evenly divide the value range of the allowable loss rate of the data to be stored to obtain at least one important identification interval, and the number of the important identification intervals is consistent with the number of stable identification intervals;

[0018] According to the midpoint of the important identification interval, the important identification intervals are numbered from large to small, the numbers of the important identification intervals are used as the importance levels of the important identification intervals, and the important identification intervals with the same level are paired with the stable identification intervals;

[0019] The importance levels of important identification intervals are summarized to form a data importance identification model.

[0020] Preferably, the setting of the equipment monitoring group for data collection according to the actual needs of disk array management includes the following steps:

[0021] By integrating a device monitoring group in a physical hard disk drive, the location information and dynamic storage information of the physical hard disk drive are collected;

[0022] Set a label for identifiable data and read the storage quantity of the data on the physical hard disk drive through the label.

[0023] Preferably, the establishment of a cyberspace coordinate system, marking data trends through a device monitoring group, and supervising the security of important areas includes the following steps:

[0024] According to the distribution of each physical hard disk drive of the disk array, a network space coordinate system is established, and the distribution of each physical hard disk drive of the disk array is marked in the network space coordinate system;

[0025] Based on the Internet of Things technology, the movement of data in the cyberspace coordinate system and the physical hard drive where it is located are obtained;

[0026] Physical hard disk drives with a rating higher than a preset value are aggregated into important areas;

[0027] Inputting the data stored in the physical hard disk drive into the data importance recognition model to obtain the importance level of the data;

[0028] Based on the stability level of the physical hard disk drive and the importance level of the data, determine whether there is data of insufficient level entering the important area. If so, issue a warning; if not, do nothing.

[0029] Preferably, the step of establishing a dynamic balance model of allocation and storage of a disk array and selecting the optimal allocation and storage work arrangement comprises the following steps:

[0030] Obtain data storage requirements at preset intervals and establish a phased storage plan list for the disk array accordingly;

[0031] According to the disk array stage storage plan list, establish the expressions of the disk array allocation cost and storage cost;

[0032] According to the phased demand, the phased data volume to be stored and the phased data stored, a phased state transfer equation for allocating disk arrays is established;

[0033] According to the disk array's phased allocation cost and storage cost expressions, a disk array phased allocation index function is established;

[0034] According to the disk array phase allocation index function, establish the storage capacity from the initial storage capacity of the first stage to the storage capacity at the end of the kth stage as S k The recursive function of the minimum total cost when ;

[0035] According to the actual situation of disk array allocation, set x k and S k The value range of is used, and the allocation-storage dynamic balance model of disk array allocation is established in a forward-pushing manner, and the optimal allocation-storage work arrangement is screened out.

[0036] Preferably, the allocation cost expression of the disk array is:

[0037]

[0038] In the formula, D(x k ) is the cost of allocating data to the disk array at the end of the kth stage, a is the fixed loss cost when the disk array does not allocate data, x k is the amount of data allocated to the disk array at the end of the kth stage, b is the fixed loss cost of the disk array when the amount of data to be stored is within the maximum allocation, α is the cost of allocating unit data to the disk array, and τ is the maximum allocation of data allocated to the disk array;

[0039] The storage cost expression of the disk array is:

[0040] C(S k )=cS k

[0041] In the formula, C(S k ) is the cost of the disk array storing data at the end of the kth stage, c is the cost of the disk array storing unit data, S k is the amount of data stored in the disk array at the end of stage k;

[0042] The disk array allocation phase state transition equation is:

[0043] S k =S k-1 +x k -d k

[0044] In the formula, S k-1 is the amount of data stored at the end of the k-1th stage, d k is the demand for data at the end of the kth stage;

[0045] The disk array phase allocation index function is:

[0046] v k (x k ,S k )=D(x k )+C(S k )

[0047] In the formula, v k (x k ,S k ) is the allocation index of the disk array at the end of the kth stage, that is, the cost of the disk array at the end of the kth stage;

[0048] The amount of storage from the initial storage of the first stage to the end of the kth stage is S k The recursive function of the minimum total cost is:

[0049]

[0050] In the formula, f k (S k ) is the initial storage amount in the first stage and the storage amount at the end of the kth stage is S k The minimum total cost, f k-1 (S k-1 ) is the initial storage amount in the first stage, and the storage amount at the end of the k-1th stage is S k-1 The minimum total cost is set when the disk array does not start phased allocation. The initial value is set to 0, that is, when k = 1, f 0 (S 0 )=0.

[0051] Preferably, the establishment of a disk array device dynamic monitoring and early warning model based on the LSTM neural network algorithm model includes the following steps:

[0052] According to the factory information of the disk array device, obtain the normal operation data of the disk array device and establish a standard library of the device operation status;

[0053] According to the type of disk array device, set the device operation status timing interval;

[0054] Based on big data and the equipment operation status standard library, the equipment operation status features are extracted, and the disk array equipment operation status input training sample set and standard training sample set are established;

[0055] According to the disk array device operation state input training sample set and the standard training sample set, based on the device operation state time series interval, the state deviation coefficient of the disk array device at different device operation state time series intervals is calculated;

[0056] According to the state deviation coefficient of the disk array device at different device operation state time intervals, combined with the disk array device operation state input training sample set, a disk array device operation state input matrix is ​​established;

[0057] According to the disk array device operation status input matrix, based on the LSTM neural network algorithm model, a disk array device dynamic monitoring and early warning model is established. According to the received device operation data, the operation status of the disk array device is judged and predicted, and the abnormal information of abnormal devices is sent and risk alerts are issued.

[0058] Preferably, the state deviation coefficient of the disk array device at different device operation state time intervals specifically includes:

[0059]

[0060] In the formula, γ is the state deviation coefficient of the disk array device at different device operation state timing intervals, δ is the equilibrium constant term of the disk array device at different device operation state timing intervals, and X i is the i-th monitoring data of the disk array device in the time interval of different device operation status, is the standard data of the disk array device operating status, and m is the number of monitoring data of the disk array device in different device operating status time intervals.

[0061] Preferably, the method of establishing a dynamic model of a smart disk array network by using digital twin technology comprises the following steps:

[0062] Based on digital twin technology, a smart disk array network coordinate model is established according to the distribution markings of each area of ​​the disk array in the network space coordinate system;

[0063] Based on the intelligent disk array network coordinate model, combined with the disk array allocation-storage dynamic balance model and the disk array equipment dynamic monitoring and early warning model, the virtual and real logical interaction relationship of the intelligent disk array network dynamic model is constructed;

[0064] Based on digital twin technology, a dynamic model of the smart disk array network is established. The dynamic model of the smart disk array network monitors, displays and controls the overall allocation, storage and operation of the disk array through a network visualization display interface and human-computer interaction.

[0065] The storage space optimization allocation device of the redundant disk array is used to implement the storage space optimization allocation method of the redundant disk array, comprising:

[0066] A data acquisition module, the data acquisition module acquires and aggregates at least one physical hard disk drive as a disk array;

[0067] A level generation module, wherein the level generation module obtains the stability level of the physical hard disk drive according to the stability and establishes a data importance recognition model;

[0068] A data acquisition module, wherein the data acquisition module sets a device monitoring group for data acquisition according to actual requirements of disk array management;

[0069] A dynamic management module, which establishes a cyberspace coordinate system, marks data trends through a device monitoring group, and monitors the security of important areas;

[0070] A dynamic planning module, wherein the dynamic planning module establishes a dynamic balancing model of allocation and storage of the disk array according to the actual storage demand of the disk array, and selects the optimal allocation and storage work arrangement;

[0071] The equipment dynamic monitoring module, based on the LSTM neural network algorithm model, establishes a disk array equipment dynamic monitoring and early warning model to judge and predict the operating status of the disk array equipment, and sends abnormal information of abnormal equipment and makes risk alerts;

[0072] A visualization platform module, which uses digital twin technology to establish a dynamic model of the intelligent disk array network.

[0073] Compared with the prior art, the present invention has the following beneficial effects:

[0074] By setting up an equipment monitoring group to collect personnel dynamics, product dynamics and equipment dynamics data information, and analyzing the three-dimensional regional distribution of the disk array, a three-dimensional coordinate system is established, and the dynamic storage of data is monitored in real time through the Internet of Things technology. Secondly, by analyzing the optimal allocation method of data allocation and storage, an allocation-storage dynamic balance model is established to screen out the optimal allocation-storage work arrangement. Furthermore, the LSTM neural network algorithm model is used to establish the state deviation coefficient of the disk array operation status time interval, and a dynamic monitoring and early warning model for disk array equipment is constructed. By receiving the equipment operation data collected by the equipment monitoring group, the operation status of the disk array equipment is judged and predicted, and the abnormal information of abnormal equipment is sent and risk alerts are issued. Finally, through the digital twin technology, a dynamic model of the smart disk array network is established, and a visualization platform for human-computer interaction is created to achieve the management goals of intelligent, visualized and efficient storage. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 A schematic diagram of a flow chart of a method for optimizing storage space allocation of a redundant disk array of the present invention;

[0076] Figure 2 A schematic diagram of a flow chart of obtaining the stability level of a physical hard disk drive and establishing a data importance identification model according to the stability of the present invention;

[0077] Figure 3 A schematic diagram of a process of setting a device monitoring group for data collection according to the actual needs of disk array management of the present invention;

[0078] Figure 4 A schematic diagram of the process of establishing a cyberspace coordinate system, marking data trends through a device monitoring group, and supervising the security of important areas for the present invention;

[0079] Figure 5 A schematic diagram of a process for establishing a dynamic balancing model of allocation and storage of a disk array and screening out an optimal allocation and storage work arrangement according to the present invention;

[0080] Figure 6 A schematic diagram of a process for establishing a dynamic monitoring and early warning model for disk array equipment based on the LSTM neural network algorithm model of the present invention;

[0081] Figure 7 This is a schematic diagram of the process of establishing a dynamic model of a smart disk array network through digital twin technology according to the present invention. DETAILED DESCRIPTION

[0082] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.

[0083] Reference Figure 1 As shown, the storage space optimization allocation method of the redundant disk array includes:

[0084] Acquire and aggregate at least one physical hard disk drive as a disk array;

[0085] According to the stability, the stability level of the physical hard disk drive is obtained, and a data importance recognition model is established;

[0086] According to the actual needs of disk array management, set up equipment monitoring groups for data collection;

[0087] Establish a cyberspace coordinate system, mark data movements through equipment monitoring groups, and monitor the security of important areas;

[0088] According to the actual storage requirements of the disk array, establish the disk array allocation-storage dynamic balance model to select the optimal allocation-storage work arrangement;

[0089] Based on the LSTM neural network algorithm model, a dynamic monitoring and early warning model for disk array devices is established to judge and predict the operating status of disk array devices, send abnormal information of abnormal devices and make risk alerts;

[0090] Through digital twin technology, a dynamic model of the smart disk array network is established.

[0091] When optimizing storage space, different data have different security requirements. Therefore, different specifications of storage are required according to their different requirements. The security risks of storing physical hard disk drives with different parameters are different. Therefore, if high-specification data is stored in a low-specification manner, the security of the data will not be guaranteed enough. In addition, the conventional operation of using disk arrays for storage is to use a uniform load method. However, due to the different positions and performances of different physical hard disk drives in the disk array, the cost of transferring a large amount of data to the physical hard disk drive is different. At the same time, the maintenance cost of storing on the physical hard disk drive is also different. Therefore, when storing, only a uniform load method cannot be used for storage, which may lead to insufficient cost control and abnormal situations during the storage process. In order to ensure the normal operation of storage, corresponding warnings are required for these situations. In this solution, corresponding algorithms are set for the various problems mentioned above to provide targeted solutions.

[0092] Reference Figure 2 As shown, according to the stability, the stability level of the physical hard disk drive is obtained, and the establishment of the data importance identification model includes the following steps:

[0093] Evenly divide the value range of the data loss rate of the physical hard disk drive to obtain at least one stable identification interval;

[0094] According to the midpoint of the stable identification interval, the stable identification interval is numbered from large to small, and the number of the stable identification interval is used as the stability level of the stable identification interval;

[0095] Evenly divide the value range of the allowable loss rate of the data to be stored to obtain at least one important identification interval, and the number of the important identification intervals is consistent with the number of stable identification intervals;

[0096] According to the midpoint of the important identification interval, the important identification intervals are numbered from large to small, the numbers of the important identification intervals are used as the importance levels of the important identification intervals, and the important identification intervals with the same level are paired with the stable identification intervals;

[0097] The importance levels of important identification intervals are summarized to form a data importance identification model.

[0098] After obtaining the stability level of the physical hard disk drive and establishing a data importance identification model, it is necessary to ensure that the physical hard disk drive stores data corresponding to its level during storage. Otherwise, the physical hard disk drive will be occupied by low-level data. When high-level data appears, there may be no space for storage. When high-level data is stored in a low-level physical hard disk drive, the security of the data cannot be fully guaranteed.

[0099] The grade of the physical hard disk drive is the grade of the stable identification interval where the data loss rate of the physical hard disk drive lies;

[0100] The level of data is the level of an important identification interval in which the data loss rate allowed by the data lies.

[0101] Reference Figure 3 As shown, according to the actual needs of disk array management, setting up a device monitoring group for data collection includes the following steps:

[0102] By integrating a device monitoring group in a physical hard disk drive, the location information and dynamic storage information of the physical hard disk drive are collected;

[0103] Set a label for identifiable data and read the storage quantity of the data on the physical hard disk drive through the label.

[0104] Reference Figure 4 As shown, establishing a cyberspace coordinate system, marking data movements through a device monitoring group, and supervising the security of important areas include the following steps:

[0105] According to the distribution of each physical hard disk drive of the disk array, a network space coordinate system is established, and the distribution of each physical hard disk drive of the disk array is marked in the network space coordinate system;

[0106] Based on the Internet of Things technology, the movement of data in the cyberspace coordinate system and the physical hard drive where it is located are obtained;

[0107] Physical hard disk drives with a rating higher than a preset value are aggregated into important areas;

[0108] Inputting the data stored in the physical hard disk drive into the data importance recognition model to obtain the importance level of the data;

[0109] Based on the stability level of the physical hard disk drive and the importance level of the data, determine whether there is data of insufficient level entering the important area. If so, issue a warning; if not, do nothing.

[0110] Reference Figure 5 As shown in the figure, establishing a dynamic balance model of disk array allocation and storage and selecting the optimal allocation and storage work arrangement includes the following steps:

[0111] Obtain data storage requirements at preset intervals and establish a phased storage plan list for the disk array accordingly;

[0112] According to the disk array stage storage plan list, establish the expressions of the disk array allocation cost and storage cost;

[0113] According to the phased demand, the phased data volume to be stored and the phased data stored, a phased state transfer equation for allocating disk arrays is established;

[0114] According to the disk array's phased allocation cost and storage cost expressions, a disk array phased allocation index function is established;

[0115] According to the disk array phase allocation index function, establish the storage capacity from the initial storage capacity of the first stage to the storage capacity at the end of the kth stage as S k The recursive function of the minimum total cost when ;

[0116] According to the actual situation of disk array allocation, set x k and S k The value range of is used, and the allocation-storage dynamic balance model of disk array allocation is established in a forward-pushing manner, and the optimal allocation-storage work arrangement is screened out.

[0117] The allocation cost expression of the disk array is:

[0118]

[0119] In the formula, D(x k ) is the cost of allocating data to the disk array at the end of the kth stage, a is the fixed loss cost when the disk array does not allocate data, x kis the amount of data allocated to the disk array at the end of the kth stage, b is the fixed loss cost of the disk array when the amount of data to be stored is within the maximum allocation, α is the cost of allocating unit data to the disk array, and τ is the maximum allocation of data allocated to the disk array;

[0120] The storage cost expression of the disk array is:

[0121] C(S k )=cS k

[0122] In the formula, C(S k ) is the cost of the disk array storing data at the end of the kth stage, c is the cost of the disk array storing unit data, S k is the amount of data stored in the disk array at the end of stage k;

[0123] The disk array allocation phase state transition equation is:

[0124] S k =S k-1 +x k -d k

[0125] In the formula, S k-1 is the amount of data stored at the end of the k-1th stage, d k is the demand for data at the end of the kth stage;

[0126] The disk array phase allocation index function is:

[0127] v k (x k ,S k )=D(x k )+C(S k )

[0128] In the formula, v k (x k ,S k ) is the allocation index of the disk array at the end of the kth stage, that is, the cost of the disk array at the end of the kth stage;

[0129] The amount of storage from the initial storage of the first stage to the end of the kth stage is S k The recursive function of the minimum total cost is:

[0130]

[0131] In the formula, f k (S k ) is the initial storage amount in the first stage and the storage amount at the end of the kth stage is S k The minimum total cost, f k-1(S k-1 ) is the initial storage amount in the first stage, and the storage amount at the end of the k-1th stage is S k-1 The minimum total cost is set when the disk array does not start phased allocation. The initial value is set to 0, that is, when k = 1, f 0 (S 0 )=0.

[0132] It can be explained that, since the disk array is composed of multiple physical hard disk drives, the locations of the physical hard disk drives are different, the cost of transferring data to the physical hard disk drives is different, and the performance of the physical hard disk drives is different, the cost of storing data is also different, including the cost of maintenance and electricity consumption. This solution analyzes the data storage demand and the supply and demand balance of the disk array, establishes an allocation-storage dynamic balance model of the disk array, and calculates the optimal allocation-storage work arrangement in a forward manner, thereby efficiently and scientifically planning the allocation-storage work arrangement of the disk array and improving the resource utilization of the disk array.

[0133] Reference Figure 6 As shown in the figure, based on the LSTM neural network algorithm model, establishing a disk array equipment dynamic monitoring and early warning model includes the following steps:

[0134] According to the factory information of the disk array device, obtain the normal operation data of the disk array device and establish a standard library of the device operation status;

[0135] According to the type of disk array device, set the device operation status timing interval;

[0136] Based on big data and the equipment operation status standard library, the equipment operation status features are extracted, and the disk array equipment operation status input training sample set and standard training sample set are established;

[0137] According to the disk array device operation state input training sample set and the standard training sample set, based on the device operation state time series interval, the state deviation coefficient of the disk array device at different device operation state time series intervals is calculated;

[0138] According to the state deviation coefficient of the disk array device at different device operation state time intervals, combined with the disk array device operation state input training sample set, a disk array device operation state input matrix is ​​established;

[0139] According to the disk array device operation status input matrix, based on the LSTM neural network algorithm model, a disk array device dynamic monitoring and early warning model is established. According to the received device operation data, the operation status of the disk array device is judged and predicted, and the abnormal information of abnormal devices is sent and risk alerts are issued.

[0140] The state deviation coefficients of disk array devices at different device operation state timing intervals specifically include:

[0141]

[0142] In the formula, γ is the state deviation coefficient of the disk array device at different device operation state timing intervals, δ is the equilibrium constant term of the disk array device at different device operation state timing intervals, and X i is the i-th monitoring data of the disk array device in the time interval of different device operation status, is the standard data of the disk array device operating status, and m is the number of monitoring data of the disk array device in different device operating status time intervals.

[0143] It can be explained that the operation of disk array devices is a long process. When analyzing the operating status of disk array devices, it is necessary to consider the cumulative deviation of the operating status of the equipment. Otherwise, it is easy to cause inaccurate and delayed judgment of abnormal information of the operating status of the disk array device, thereby causing more serious losses. This solution establishes a device operating status time series interval, and according to the device operating status time series interval, establishes a disk array device operating status input training sample set and a standard training sample set based on big data, thereby establishing the state deviation coefficient of the disk array device at different device operating status time series intervals, and establishes a disk array device operating status input matrix. Finally, through the LSTM neural network algorithm model, the operating status of the disk array device can be effectively judged and predicted, and abnormal information of abnormal devices can be sent and risk alerts can be issued.

[0144] Reference Figure 7 As shown in the figure, the establishment of a dynamic model of a smart disk array network through digital twin technology includes the following steps:

[0145] Based on digital twin technology, a smart disk array network coordinate model is established according to the distribution markings of each area of ​​the disk array in the network space coordinate system;

[0146] Based on the intelligent disk array network coordinate model, combined with the disk array allocation-storage dynamic balance model and the disk array equipment dynamic monitoring and early warning model, the virtual and real logical interaction relationship of the intelligent disk array network dynamic model is constructed;

[0147] Based on digital twin technology, a dynamic model of the smart disk array network is established. The dynamic model of the smart disk array network monitors, displays and controls the overall allocation, storage and operation of the disk array through a network visualization display interface and human-computer interaction.

[0148] Through visualization, the allocation, storage and operation status can be displayed in a timely manner. Operators can also identify and deal with possible problems through visual monitoring to ensure the rationality of the allocation.

[0149] The storage space optimization allocation device of the redundant disk array is used to implement the storage space optimization allocation method of the redundant disk array, comprising:

[0150] A data acquisition module, the data acquisition module acquires and aggregates at least one physical hard disk drive as a disk array;

[0151] A level generation module, wherein the level generation module obtains the stability level of the physical hard disk drive according to the stability and establishes a data importance recognition model;

[0152] A data acquisition module, wherein the data acquisition module sets a device monitoring group for data acquisition according to actual requirements of disk array management;

[0153] A dynamic management module, which establishes a cyberspace coordinate system, marks data trends through a device monitoring group, and monitors the security of important areas;

[0154] A dynamic planning module, wherein the dynamic planning module establishes a dynamic balancing model of allocation and storage of the disk array according to the actual storage demand of the disk array, and selects the optimal allocation and storage work arrangement;

[0155] The equipment dynamic monitoring module, based on the LSTM neural network algorithm model, establishes a disk array equipment dynamic monitoring and early warning model to judge and predict the operating status of the disk array equipment, and sends abnormal information of abnormal equipment and makes risk alerts;

[0156] A visualization platform module, which uses digital twin technology to establish a dynamic model of the intelligent disk array network.

[0157] Furthermore, the present solution also proposes a storage medium on which a computer-readable program is stored. When the computer-readable program is called, the above-mentioned storage space optimization allocation method of the redundant disk array is executed.

[0158] It is understandable that the storage medium may be a magnetic medium, such as a floppy disk, a hard disk, or a magnetic tape; an optical medium, such as a DVD; or a semiconductor medium, such as a solid state drive (SSD).

[0159] In summary, the advantages of the present invention are: by setting up an equipment monitoring group to collect personnel dynamics, product dynamics and equipment dynamics data information, and analyzing the three-dimensional regional distribution of the disk array, a three-dimensional coordinate system is established, and the dynamic storage of data is monitored in real time through the Internet of Things technology. Secondly, by analyzing the optimal allocation method of data allocation and storage, a distribution-storage dynamic balance model is established to screen out the optimal distribution-storage work arrangement. Furthermore, by using the LSTM neural network algorithm model, a state deviation coefficient under the time interval of the disk array operation state is established, and a disk array equipment dynamic monitoring and early warning model is constructed. By receiving the equipment operation data collected by the equipment monitoring group, the operation state of the disk array equipment is judged and predicted, and the abnormal information of the abnormal equipment is sent and a risk alarm is issued. Finally, through the digital twin technology, a smart disk array network dynamic model is established to create a visualization platform for human-computer interaction to achieve the management goals of intelligent, visualized and efficient storage.

[0160] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for optimizing storage space allocation of a redundant disk array, characterized in that: include: Acquire and aggregate at least one physical hard disk drive as a disk array; According to the stability, the stability level of the physical hard disk drive is obtained, and a data importance recognition model is established; According to the actual needs of disk array management, set up equipment monitoring groups for data collection; Establish a cyberspace coordinate system, mark data movements through equipment monitoring groups, and monitor the security of important areas; According to the actual storage requirements of the disk array, establish the disk array allocation-storage dynamic balance model to select the optimal allocation-storage work arrangement; Based on the LSTM neural network algorithm model, a dynamic monitoring and early warning model for disk array devices is established to judge and predict the operating status of disk array devices, send abnormal information of abnormal devices and make risk alerts; Through digital twin technology, a dynamic model of the smart disk array network is established.

2. The storage space optimization allocation method of redundant disk array according to claim 1, characterized in that: The step of obtaining the stability level of the physical hard disk drive according to the stability and establishing the data importance identification model comprises the following steps: Evenly divide the value range of the data loss rate of the physical hard disk drive to obtain at least one stable identification interval; According to the midpoint of the stable identification interval, the stable identification interval is numbered from large to small, and the number of the stable identification interval is used as the stability level of the stable identification interval; Evenly divide the value range of the allowable loss rate of the data to be stored to obtain at least one important identification interval, and the number of the important identification intervals is consistent with the number of stable identification intervals; According to the midpoint of the important identification interval, the important identification intervals are numbered from large to small, the numbers of the important identification intervals are used as the importance levels of the important identification intervals, and the important identification intervals with the same level are paired with the stable identification intervals; The importance levels of important identification intervals are summarized to form a data importance identification model.

3. The storage space optimization allocation method of redundant disk array according to claim 2, characterized in that: According to the actual needs of disk array management, setting up a device monitoring group for data collection includes the following steps: By integrating a device monitoring group in a physical hard disk drive, the location information and dynamic storage information of the physical hard disk drive are collected; Set a label for identifiable data and read the storage quantity of the data on the physical hard disk drive through the label.

4. The storage space optimization allocation method of redundant disk array according to claim 3, characterized in that: The establishment of a cyberspace coordinate system, marking data trends through a device monitoring group, and supervising the security of important areas includes the following steps: According to the distribution of each physical hard disk drive of the disk array, a network space coordinate system is established, and the distribution of each physical hard disk drive of the disk array is marked in the network space coordinate system; Based on the Internet of Things technology, the movement of data in the cyberspace coordinate system and the physical hard drive where it is located are obtained; Physical hard disk drives with a rating higher than a preset value are aggregated into important areas; Inputting the data stored in the physical hard disk drive into the data importance recognition model to obtain the importance level of the data; Based on the stability level of the physical hard disk drive and the importance level of the data, determine whether there is data of insufficient level entering the important area. If so, issue a warning; if not, do nothing.

5. The storage space optimization allocation method of redundant disk array according to claim 4, characterized in that: The establishment of the allocation-storage dynamic balance model of the disk array and the screening of the optimal allocation-storage work arrangement includes the following steps: Obtain data storage requirements at preset intervals and establish a phased storage plan list for the disk array accordingly; According to the disk array stage storage plan list, establish the expressions of the disk array allocation cost and storage cost; According to the phased demand, the phased data volume to be stored and the phased data stored, a phased state transfer equation for allocating disk arrays is established; According to the disk array's phased allocation cost and storage cost expressions, a disk array phased allocation index function is established; According to the disk array phase allocation index function, establish the storage capacity from the initial storage capacity of the first stage to the storage capacity at the end of the kth stage as S k The recursive function of the minimum total cost when ; According to the actual situation of disk array allocation, set x k and S k The value range of is used, and the allocation-storage dynamic balance model of disk array allocation is established in a forward-pushing manner, and the optimal allocation-storage work arrangement is screened out.

6. The storage space optimization allocation method of redundant disk array according to claim 5, characterized in that: The allocation cost expression of the disk array is: In the formula, D(x k ) is the cost of allocating data to the disk array at the end of the kth stage, a is the fixed loss cost when the disk array does not allocate data, x k is the amount of data allocated to the disk array at the end of the kth stage, b is the fixed loss cost of the disk array when the amount of data to be stored is within the maximum allocation, α is the cost of allocating unit data to the disk array, and τ is the maximum allocation of data allocated to the disk array; The storage cost expression of the disk array is: C(S k )=cS k In the formula, C(S k ) is the cost of the disk array storing data at the end of the kth stage, c is the cost of the disk array storing unit data, S k is the amount of data stored in the disk array at the end of stage k; The disk array allocation phase state transition equation is: S k =S k-1 +x k -d k In the formula, S k-1 is the amount of data stored at the end of the k-1th stage, d k is the demand for data at the end of the kth stage; The disk array phase allocation index function is: v k (x k ,S k )=D(x k )+C(S k ) In the formula, v k (x k ,S k ) is the allocation index of the disk array at the end of the kth stage, that is, the cost of the disk array at the end of the kth stage; The amount of storage from the initial storage of the first stage to the end of the kth stage is S k The recursive function of the minimum total cost is: In the formula, f k (S k ) is the initial storage amount in the first stage and the storage amount at the end of the kth stage is S k The minimum total cost, f k-1 (S k-1 ) is the initial storage amount in the first stage, and the storage amount at the end of the k-1th stage is S k-1 The minimum total cost is , wherein the initial value when the disk array does not start phased allocation is set to 0, that is, when k = 1, f0(S0) = 0.

7. The storage space optimization allocation method of redundant disk array according to claim 6, characterized in that: The establishment of a disk array device dynamic monitoring and early warning model based on the LSTM neural network algorithm model includes the following steps: According to the factory information of the disk array device, obtain the normal operation data of the disk array device and establish a standard library of the device operation status; According to the type of disk array device, set the device operation status timing interval; Based on big data and the equipment operation status standard library, the equipment operation status features are extracted, and the disk array equipment operation status input training sample set and standard training sample set are established; According to the disk array device operation state input training sample set and the standard training sample set, based on the device operation state time series interval, the state deviation coefficient of the disk array device at different device operation state time series intervals is calculated; According to the state deviation coefficient of the disk array device at different device operation state time intervals, combined with the disk array device operation state input training sample set, a disk array device operation state input matrix is ​​established; According to the disk array device operation status input matrix, based on the LSTM neural network algorithm model, a disk array device dynamic monitoring and early warning model is established. According to the received device operation data, the operation status of the disk array device is judged and predicted, and the abnormal information of abnormal devices is sent and risk alerts are issued.

8. The method for optimizing storage space allocation of a redundant disk array according to claim 7, characterized in that: The state deviation coefficient of the disk array device at different device operation state time intervals specifically includes: In the formula, γ is the state deviation coefficient of the disk array device at different device operation state timing intervals, δ is the equilibrium constant term of the disk array device at different device operation state timing intervals, and X i is the i-th monitoring data of the disk array device in the time interval of different device operation status, is the standard data of the disk array device operating status, and m is the number of monitoring data of the disk array device in different device operating status time intervals.

9. The method for optimizing storage space allocation of a redundant disk array according to claim 8, characterized in that: The establishment of a dynamic model of a smart disk array network through digital twin technology includes the following steps: Based on digital twin technology, a smart disk array network coordinate model is established according to the distribution markings of each area of ​​the disk array in the network space coordinate system; Based on the intelligent disk array network coordinate model, combined with the disk array allocation-storage dynamic balance model and the disk array equipment dynamic monitoring and early warning model, the virtual and real logical interaction relationship of the intelligent disk array network dynamic model is constructed; Based on digital twin technology, a dynamic model of the smart disk array network is established. The dynamic model of the smart disk array network monitors, displays and controls the overall allocation, storage and operation of the disk array through a network visualization display interface and human-computer interaction.

10. A storage space optimization allocation device for a redundant disk array, used to implement the storage space optimization allocation method for a redundant disk array as claimed in any one of claims 1 to 9, characterized in that: include: A data acquisition module, the data acquisition module acquires and aggregates at least one physical hard disk drive as a disk array; A level generation module, wherein the level generation module obtains the stability level of the physical hard disk drive according to the stability and establishes a data importance recognition model; A data acquisition module, wherein the data acquisition module sets a device monitoring group for data acquisition according to actual requirements of disk array management; A dynamic management module, which establishes a cyberspace coordinate system, marks data trends through a device monitoring group, and monitors the security of important areas; A dynamic planning module, wherein the dynamic planning module establishes a dynamic balancing model of allocation and storage of the disk array according to the actual storage demand of the disk array, and selects the optimal allocation and storage work arrangement; The equipment dynamic monitoring module, based on the LSTM neural network algorithm model, establishes a disk array equipment dynamic monitoring and early warning model to judge and predict the operating status of the disk array equipment, and sends abnormal information of abnormal equipment and makes risk alerts; A visualization platform module, which uses digital twin technology to establish a dynamic model of the intelligent disk array network.

Images